# Rob Hoeijmakers > On AI, organisations, and what actually works. Written for senior professionals navigating the shift. European perspective, grounded in real work. Public Ghost content for AI and LLM tooling. This file includes a bounded export of public pages first, then recent public posts. ## Pages ### About Rob Hoeijmakers URL: https://rob-hoeijmakers.ghost.io/about/ Last updated: 2026-05-07T12:40:01.000Z My name is Rob Hoeijmakers. I work at the intersection of AI strategy and the practical question of how organisations can actually use it. That means two things in practice. I advise senior teams on AI strategy: what it means for their organisation, where to start, what decisions need to be made before the technology can do useful work. And through [Schmuki](https://schmuki.nl/), the agency I run, we build AI-driven knowledge systems — the infrastructure that lets organisations capture what they know and make it available in ways both people and machines can use. ### What I Do Most of my work starts with a strategic question. An executive team that senses AI will change their organisation but isn't sure how to approach it. A knowledge-intensive organisation that wants to understand what becomes possible when AI can navigate their internal expertise. A leadership team trying to distinguish what matters from what is noise. I help them think clearly, make decisions with confidence, and move from strategy to implementation. The clients I work with — KPMG, UMC, Education Lab, Enreach — all came through existing relationships. I work with people I respect, on problems worth solving. ### What I Write About This blog is where I think in public. The topics follow my practice: AI strategy, AI governance, Europe's role in shaping the digital environment, and what it means for organisations when reasoning starts to become shared between humans and machines. I write as a practitioner. Not to explain the obvious, but to work through what I find genuinely interesting or unresolved. If you want a sense of the writing before subscribing, start here: - [AI Strategy](https://rob-hoeijmakers.ghost.io/tag/ai-strategy/) - [AI Governance](https://rob-hoeijmakers.ghost.io/tag/ai-governance/) - [Europe](https://rob-hoeijmakers.ghost.io/tag/europe/) - [Future of Work](https://rob-hoeijmakers.ghost.io/tag/future-of-work/) ### Get in Touch The best way to reach me is through [LinkedIn](https://www.linkedin.com/in/rob-hoeijmakers). If you want to know more about Schmuki's work, visit [schmuki.nl](https://schmuki.nl/). If you want to follow the thinking, [subscribe to the newsletter](https://rob-hoeijmakers.ghost.io/about/#/portal/signup). Weekly, editorial, no noise — on AI, organisations, and what actually works. ### Welcome to the newsletter! URL: https://rob-hoeijmakers.ghost.io/welcome/ Last updated: 2026-04-24T12:10:25.000Z Thanks for subscribing. Every week I write about AI strategy, governance, and what it means for organisations navigating this shift. Practical, from a European perspective, grounded in real work. You'll hear from me when there's something worth saying. Rob ### Privacy Statement URL: https://rob-hoeijmakers.ghost.io/privacy-statement/ Last updated: 2025-11-02T21:42:41.000Z This website respects your privacy. No personal data are collected or shared beyond what is strictly necessary. ### **Subscriptions** If you subscribe to receive updates by email, I only collect your email address. It is used solely to send new posts or related updates from this site. Your address is not shared, sold, or used for any other purpose. You can unsubscribe at any time using the link in each email. Unsubscribed members are no longer contacted and will be removed from the system on a regular basis to minimise stored personal data. ### **Analytics and Cookies** This site does not use tracking cookies, analytics, or third-party advertising. ### **Contact** If you have questions about your data or this policy, you can find my details here: [About Rob Hoeijmakers, Digital Strategist, AmsterdamExplore my blog for insights on AI, identity, and health in the EU, alongside my photography. I share my learnings, synthesize digital trends, and showcase my work—all as part of my evolving digital presence.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-324.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/Rob-Hoeijmakers-1.jpg)](https://rob-hoeijmakers.ghost.io/about/) ### Colophon & Disclaimer URL: https://rob-hoeijmakers.ghost.io/colofon-disclaimer/ Last updated: 2026-04-24T12:33:59.000Z **Author and Responsibility** This website is written and maintained by **Rob Hoeijmakers**, Amsterdam. For any questions or permissions, please use [this site](https://robhoeij.pro). **About this Site** The site is published on the [**Ghost**](https://ghost.org) platform, hosted on servers located in the United States. Data are processed solely for technical operation; no behavioural tracking or analytics are used. **Content and Copyright** All texts and photographs, unless otherwise noted, are created by Rob Hoeijmakers. You are welcome to quote or share them with proper attribution, but not for commercial use. For specific requests, please get in touch. **Use of AI** Some texts are partly developed with the help of language models (AI), edited and published under the author’s full responsibility. **Affiliate Links** Occasionally, posts may include affiliate links. These do not affect editorial choices and currently generate no meaningful revenue. **Disclaimer** The content of this site is reflective and informational in nature. It does not constitute legal, financial, or medical advice. While all care is taken, errors or omissions may occur; the author accepts no liability for decisions based on this material. **Technical Credits** Domain and email: self-managed. Hosting: Ghost(Pro), United States. Photography: original work by Rob Hoeijmakers. ### ClipboardMarkdown URL: https://rob-hoeijmakers.ghost.io/clipboardmarkdown/ Last updated: 2026-05-12T21:13:02.000Z A small macOS utility that does one thing: it takes whatever text is on your clipboard and saves it as a `.md` file in your Downloads folder. No app to switch to, no paste-format-save cycle. One click from the menu bar. I built it for my own workflow and several people asked if they could try it, so here it is. ## What it does ClipboardMarkdown sits in your menu bar as a clipboard icon. Click it, choose **Save Clipboard as Markdown**, and the file lands in `~/Downloads` with a timestamp as the filename: `2026-04-04 15.42.10.md` It handles three source types, in order of preference: - **HTML with structure** (headings, code, lists, links) from sources like GitHub, web articles, or Google Docs: converted to clean Markdown - **Rich text** from Pages, Word, or similar: formatting preserved where meaningful - **Plain text**: saved as-is Sources without recognisable structure, like LinkedIn or plain prose in divs, fall back to plain text automatically. A system notification confirms each save. Click it to open the file directly. ## Requirements - macOS 13 Ventura or newer - Apple Silicon or Intel ## Download Download the zip, unzip, move `ClipboardMarkdown.app` to your Applications folder. Done. [ClipboardMarkdown-1.1.1Version 1.1.1, macOS 13+, Apple Silicon & Intel. Notarized.ClipboardMarkdown-1.1.1.zip509 KBdownload-circle](https://rob-hoeijmakers.ghost.io/content/files/2026/04/ClipboardMarkdown-1.1.1-1.zip "Download") ## First launch ClipboardMarkdown is notarised by Apple, so macOS will open it without warnings. Move `ClipboardMarkdown.app` to your Applications folder, double-click, and you're done. ⚠️ PDF is not supported. PDF is a rendering format, not a copy format — the clipboard never carries the document, only what the PDF viewer chose to expose as text. For PDF-to-Markdown conversion, use a tool that works on the file directly, such as Marker (open source) or a [browser-based converter](https://pdf2md.morethan.io). ## Want to inspect or build it yourself The distributed binary is notarised by Apple and you can verify what it does by reading the source. If you'd rather build from source than trust a downloaded binary, [the repository on GitHub](https://github.com/rhoeijmakers/ClipboardMarkdown) has everything you need. ### Search URL: https://rob-hoeijmakers.ghost.io/search/ Last updated: 2026-05-25T18:20:39.000Z Browse by topic [AI Strategy](https://rob-hoeijmakers.ghost.io/tag/ai-strategy/) [AI Governance](https://rob-hoeijmakers.ghost.io/tag/ai-governance/) [AI in Practice](https://rob-hoeijmakers.ghost.io/tag/ai-in-practice/) [Future of Work](https://rob-hoeijmakers.ghost.io/tag/future-of-work/) [Europe](https://rob-hoeijmakers.ghost.io/tag/europe/) ## Posts ### The App Store Is Not the Argument URL: https://rob-hoeijmakers.ghost.io/the-app-store-is-not-the-argument/ Last updated: 2026-06-06T09:58:37.000Z A Threads post went quietly viral this week. It claimed that the EU’s new digital identity wallet, despite being promoted as fully independent, still requires users to have an Apple or Google account on their phone. The implication: the EU’s ambition to build sovereign digital infrastructure is undermined before it starts. It is a good-looking critique. It does not hold up. ## The delivery problem The EU Digital Identity Wallet is a smartphone app. Like every smartphone app, it arrives on your device through an app store: the App Store on iOS, Google Play on Android. Both require an account. That account belongs to Apple or Google, not to the EU. So the observation is factually correct. But the conclusion drawn from it is circular. To criticise the wallet for this, you first have to argue that people should be able to use their smartphones without an Apple or Google account. That is a legitimate debate. It touches platform monopoly, the [Digital Markets Act](https://rob-hoeijmakers.ghost.io/eu-digital-markets-act/), the degree to which mobile operating systems function as gatekeepers. But it is a debate about the phone, not about the wallet. The wallet inherits whatever dependency the device already carries. It did not create it. The analogy is awkward but exact: complaining that a Dutch government website requires internet access. ## Citizens, not consumers There is a more fundamental confusion in the critique, and it goes beyond circular reasoning. The App Store and Google Play are commercial environments. They exist to distribute software to consumers and businesses. The [EUDI Wallet](https://rob-hoeijmakers.ghost.io/eu-eid-eidas-eudi-wallet/) is civic infrastructure. It is the digital equivalent of a passport or a national ID card. The state issuing you a passport is not making a statement about the postal service, even if the envelope arrived by post. The wallet’s sovereignty claim is about the credential, not the device. The EU is asserting that a citizen’s identity, qualifications, health data, and legal documents should be portable, state-backed, and not dependent on any commercial party for their validity. That is a claim about the relationship between citizen and state. Apple and Google are simply not the relevant actors in that frame. This is a classical European institutional move: carve out a domain where citizens interact with public institutions and insist it operates under public law, not commercial logic. That the app happens to be distributed through the App Store is about as relevant as the fact that tax forms used to be printed by private printing companies. The printer is not the point. ## What is actually at stake There is a real sovereignty question buried here, and it deserves more than a Threads post. Some national implementations of the EUDI Wallet, notably Italy and France, have built in a dependency on Google’s Play Integrity API. This means the app contacts Google at runtime to verify its own integrity before it will function. That is a structural dependency on US infrastructure baked into the security model of a European identity system, and critics inside the project’s own GitHub repository have called it out forcefully. That critique lands. It is the difference between [platform dependency as supply chain risk](https://rob-hoeijmakers.ghost.io/when-the-platform-becomes-the-policy/) and platform dependency as deliberate architectural choice. One is the condition of operating in the current mobile ecosystem. The other is a decision that could have been made differently. The app store requirement is unavoidable given the devices most citizens actually use. The Play Integrity integration is not. One is a precondition; the other is a choice. Conflating them obscures both. ## The framing problem European digital policy initiatives attract a particular kind of criticism that holds them to a standard no single instrument could possibly meet. The wallet is not designed to solve Apple and Google’s grip on mobile distribution. It is designed to give citizens a portable, privacy-preserving credential layer that works across all 27 member states without requiring each service to build its own identity system. The [sovereignty question in Europe](https://rob-hoeijmakers.ghost.io/sovereignty-without-soil/) is real and unresolved. The EuroStack ambition, the push for European cloud alternatives, the tension between regulation and building capacity: these are the coordinates of a genuine structural challenge. A critique that reduces it to “you still need an Apple account” does not advance that conversation. It performs scepticism without earning it. The wallet lands on your phone through Apple or Google because that is where your phone lives. What it does after that, and who controls the trust chain inside it, is the question worth asking. 🇪🇺 The EUDI Wallet regulation requires all EU member states to make wallets available to citizens by end of 2026\. Mandatory acceptance by banks, telecoms, and other regulated sectors follows in December 2027. --- ### A note on Play Integrity Some readers may wonder why this concern applies primarily to Android. The answer lies in the structure of the ecosystem itself. Apple's iOS platform is already vertically integrated. Apple controls the hardware, operating system, app distribution, and security model. An identity wallet running on iPhone inevitably operates within that framework. Android is different. A wallet can be distributed through Google Play, Samsung Galaxy Store, enterprise channels, alternative app stores, or direct installation. Google's Play Integrity API effectively introduces Google as a trust anchor inside that more diverse ecosystem by allowing apps to ask Google whether a device and app installation should be considered trustworthy. The controversy is therefore not that Google provides a security service. It is that a European public identity system may come to depend on Google's judgement about what constitutes a legitimate Android device. For critics concerned with digital sovereignty, that is a different question from simply downloading an app through Google Play. ### The same problem, thirty years later URL: https://rob-hoeijmakers.ghost.io/the-same-problem-thirty-years-later/ Last updated: 2026-06-07T20:37:42.000Z The first websites I built for organisations were essentially knowledge problems. Someone in the building knew something useful. The question was how to get it out of their head, into a format a system could hold, and back to someone who needed it. We called it content. We called it information architecture. We called it knowledge management, depending on which consultant was in the room. The terminology changed every few years. The problem didn't. ## The pattern I've watched this cycle play out across three decades of working at the intersection of organisations and technology. In the mid-nineties, the web arrived and suddenly there was a place to put things. The challenge was: which things, structured how, maintained by whom? Content management systems emerged to answer that. They helped. They also introduced new layers of complexity that made the original problem harder to see. Intranets came next. Then wikis. Then enterprise search. Then SharePoint, in all its configurations. Each generation of tooling carried the same implicit promise: this time, knowledge will flow. And each time, the same thing happened. The tool got adopted. The underlying question got deferred. What organisations actually struggle with is not tooling. It's the discipline of treating knowledge as something that needs to be captured, maintained and made available, not as a byproduct of work but as work itself. ## What AI reveals AI hasn't changed this. It has made it harder to ignore. When you put a language model to work in an organisation, it works with what the organisation has. Documents, policies, case files, procedures. If those are well-structured, current and findable, the model is useful. If they are scattered across drives with inconsistent naming and no clear ownership, the model reflects that back, fluently and confidently. [That is the wall every AI deployment eventually runs into](https://rob-hoeijmakers.ghost.io/built-in-not-bolted-on/): not the model's limitations but the organisation's information underneath it. I've described this elsewhere as the difference between [content as decoration and content as knowledge](https://rob-hoeijmakers.ghost.io/platecms-content-ready-for-intelligence/). Organisations have spent years publishing things. Far fewer have spent that time structuring what they know in a way that's durable and reusable. ## The constant What strikes me now, looking back, is how consistent this has been. The organisations that got the most out of the web were the ones that thought carefully about what they knew and how to represent it. The ones that got the most out of search were the ones with clean, maintained, findable content. The pattern holds for AI. The technology shifts. The underlying requirement stays fixed. This is not a pessimistic observation. It is a clarifying one. The question an organisation needs to answer before any AI project is not "which model should we use" or "which tool should we buy." It is the same question I was sitting with in 1995: what do we actually know, who owns it, and how do we keep it alive? Three decades later, that question is worth more than ever. The systems that can use a good answer to it are considerably more powerful than a static webpage. But the answer still has to come from the organisation itself. [No tool has ever solved that for anyone.](https://rob-hoeijmakers.ghost.io/the-layer-ai-is-moving-into-now/) ### The Poisoned Library URL: https://rob-hoeijmakers.ghost.io/the-poisoned-library/ Last updated: 2026-06-04T07:06:18.000Z In early 2024, a developer named Jia Tan committed a small compression fix to XZ Utils, a library present on almost every Linux server in the world. The fix was legitimate. So were dozens of previous contributions, spread across two years of careful, patient participation in the project. The backdoor came later, buried in a build script, designed to silently compromise SSH authentication. It was caught almost by accident, by an engineer who noticed his login was taking a few milliseconds longer than expected. The XZ incident is not a story about a clever attack. It is a story about what happens when trust is structural and invisible. ## The Supply Chain Problem Software runs on software. Every application you deploy is a stack of dependencies: packages, libraries, modules, each written by someone you have likely never heard of, each trusted implicitly because it was published to a registry and downloaded enough times to seem legitimate. This is how modern software works. It is also how it fails. Supply chain attacks exploit that implicit trust. The attacker does not break into your system directly. They become part of the infrastructure your system depends on. A typosquatted package name on PyPI. A compromised npm maintainer account. A years-long social engineering campaign on an open source project. The payload arrives as a dependency update, signed and versioned and entirely unremarkable. This attack surface has been growing for years. What is new is where it is heading. ## The Agent Layer An AI agent does not run on a library. It runs on tools. When I connect Claude to my calendar, my files, my search index, those connections are mediated by servers that describe their own capabilities and respond to the model's requests. The Model Context Protocol, or MCP, is one architectural approach to this: a standard for how agents discover what they can do, and how tool servers register and coordinate. The parallelism with package managers is uncomfortable and exact. An agent that trusts an MCP server is in a structurally similar position to an application that trusts an npm package. If that server is compromised, or malicious from the start, it does not just leak data. It has privileged access to the agent's reasoning and actions. It can inject instructions. It can redirect tool calls. It can quietly reshape what the agent believes it is doing. Prompt injection through a tool response is one version of this. A malicious server could return content that includes embedded instructions to the model, exploiting the fact that the agent cannot reliably distinguish between data and commands. This is not theoretical. Security researchers have demonstrated it repeatedly in the past year. The harder version is what you might call an identity attack: a server that presents itself as something trusted, behaves normally most of the time, and activates specific behavior under specific conditions. Not unlike a build script that only runs its payload on certain architectures, during certain hours. 💡 ****While this article was being written**: On June 3, 2026, OpenAI sent a security notification to users of its macOS apps. A widely used open-source npm library, TanStack, had been compromised. Not the apps themselves, but the build process surrounding them — precisely the link this article is about. OpenAI was forced to rotate code signing material for ChatGPT Desktop, Codex, Atlas, and the Codex CLI: the cryptographic signature that allows users to verify an app is genuinely from OpenAI. ## What This Reveals [The delegation problem of the internet](https://rob-hoeijmakers.ghost.io/the-delegation-problem-of-the-internet/) is already a real constraint on what agents can do. Identity, authentication, and accountability are points of friction that slow down automation. But friction also provides a kind of protection. When an agent has to authenticate to act, there is a moment of legibility. A log entry. A confirmation. As agentic systems acquire broader permissions and more capable tool access, the question of what they trust, and on what basis, becomes structural. The answer cannot be "whatever is installed," any more than it can be "whatever is in node\_modules." [Agents and Agency](https://rob-hoeijmakers.ghost.io/agents-and-agency/) made the point that the real challenge is finding the right balance of delegation, control, and trust. That observation was made with autonomous AI systems in mind. Supply chain dynamics make it harder. The question is no longer just whether to delegate, but whether the thing you are delegating through is what it claims to be. ## The Governance Gap Regulators are not yet here. The EU AI Act classifies systems by risk and mandates transparency, but its model is primarily about the AI system itself, not the infrastructure it depends on. [ISO 42001](https://rob-hoeijmakers.ghost.io/iso-42001-ai-governance/) addresses management systems around AI, which is closer, but still upstream of the specific attack surface that agentic tool ecosystems present. The Cyber Resilience Act, which covers software components, is the most applicable instrument. But it was designed for a world of packages, not a world of live tool servers that describe their own capabilities to autonomous agents at runtime. In the meantime, the XZ lesson applies: the most dangerous dependencies are not the ones you know are risky. They are the ones that have earned enough trust to stop being examined. 📓 XZ Utils is a data compression library used in most Linux distributions. The 2024 backdoor, introduced through patient long-term contribution, was discovered by Andres Freund at Microsoft and described as one of the most sophisticated supply chain attacks ever found in open source infrastructure. ### The Layer AI Is Moving Into Now URL: https://rob-hoeijmakers.ghost.io/the-layer-ai-is-moving-into-now/ Last updated: 2026-06-01T18:40:41.000Z Last year I spent several months inside a large professional services firm, helping them think through where AI fits in their work. Not as a technical question. As an organisational one. What struck me was not what the tools could do. It was where the conversations kept landing: one level higher than expected. ## From Task to Coordination The first wave of AI at work was horizontal. Tools spread across teams. People drafted faster, summarised meetings, rephrased text. Useful, visible, and largely peripheral to how organisations actually make decisions. That is changing. The tools are moving up. Not dramatically, not all at once. But the pattern is consistent. AI is beginning to touch coordination: which cases to prioritise, what information to surface before a meeting, where process bottlenecks are forming. Work that used to live in the heads of middle managers or senior analysts is being absorbed into systems that run quietly in the background. I have watched this happen in practice. Knowledge-intensive organisations, firms where expertise is the core product, are finding that AI does not just support individual workers. It begins to shape the rhythm of the organisation itself. ## What Managers Actually Do There is a useful way to think about this. Management, at its core, has always been an **information routing problem**. Information flows in, gets filtered, prioritised, and passed upward or sideways. Decisions come back down. The manager sits at the intersection. AI is now capable of doing parts of that routing. Not the judgement calls. Not the conversations that require trust, context, and accountability. But the aggregation, the pattern recognition, the flagging of anomalies: those are moving into systems. This is what [the shift from hype to workflows](https://rob-hoeijmakers.ghost.io/from-hype-to-workflows-where-ais-real-shift-is-happening/) actually means in practice. The visible, chatbot-style layer was the beginning. The deeper layer is less visible and more consequential. ## The Delegation Question When AI reaches this level, a different question becomes urgent: what do you keep, and what do you hand over? I think about this constantly in my own work. Preparation, synthesis, structural analysis: I delegate those. Orientation, judgement, the decision about what actually matters in a situation: those stay with me. The line is not arbitrary. It follows accountability. If I have to explain a decision, own a relationship, or stand behind a recommendation, I have to have made it myself. [Vision, judgement, and creativity](https://rob-hoeijmakers.ghost.io/vision-judgement-creativity-reclaiming-agency-in-the-age-of-ai/) remain human not because AI cannot simulate them, but because they require someone to be accountable for them. Most executives I speak with have not drawn this line explicitly. They use AI, sometimes extensively, but the question of where their own judgement begins and the tool's output ends is often unexamined. That is manageable when AI is a writing aid. It becomes a governance question when AI is shaping priorities. ## The Quiet Governance Problem The firms that are getting this right are not necessarily using more sophisticated technology. They are asking better questions about which decisions should remain human, and building that into how they use the tools. This connects to something I have seen repeatedly: organisations where AI disappoints are often not disappointed in the model. They are disappointed in what the model reveals about their own knowledge infrastructure, the undocumented decisions, the expertise that lives in one person's head, the processes that were never written down. As I have written elsewhere, [AI that is built in rather than bolted on](https://rob-hoeijmakers.ghost.io/built-in-not-bolted-on/) requires something underneath it to be structured enough to be useful. When AI enters the coordination layer, that structural requirement moves up a level too. The question is no longer just whether your documents are organised. It is whether your decision-making is legible enough for a system to support it, and explicit enough for a person to remain responsible for it. ## Not a Tool Question What I find interesting about this moment is that it is forcing a conversation that was overdue in most organisations: what does management actually consist of, and which parts of it require human presence? That is not a question about AI capabilities. It is a question about [the informal layer of authority and judgement](https://rob-hoeijmakers.ghost.io/the-workspace-you-already-have/) that most organisations have never had to make explicit, because humans filled it automatically. AI entering the coordination layer does not answer that question. It makes it impossible to avoid. 📒 This post grew out of work with knowledge-intensive organisations navigating where AI fits in their decision structures. The pattern I describe is not sector-specific. I have seen versions of it in professional services, healthcare, and government. ### A font is not a file URL: https://rob-hoeijmakers.ghost.io/a-font-is-not-a-file/ Last updated: 2026-06-01T17:29:11.000Z The logo file was ready. The font was not. We were setting up something for the company, and the typeface we needed was on the old machine. Not on the new one. Not anywhere we could quickly find it. My wife suggested outlining the text, which solved the problem neatly, but the workaround stuck with me for the rest of the day. I have been interested in [typography in the digital context](https://rob-hoeijmakers.ghost.io/typography-in-motion/) for decades. Fonts have always felt like objects to me. Things you download, install, own. Tangible in some mental sense, even if they exist only as files. The way a typeface appears in the font menu of an operating system reinforces that feeling. It sits there like a tool in a drawer. That mental model turned out to be wrong, or at least incomplete. ## What a font actually is A font is not a collection of letterform images. It is closer to a compact mathematical language for generating letterforms. Inside the file: abstract shape definitions, spacing rules, curve data, substitution logic, hinting instructions. No visible letters, only the rules for producing them. When software displays the word SCHMUKI in a particular typeface, it is not retrieving a stored image. It is evaluating a set of instructions at that moment, at that size, in that context. The visible shape is the output of a process, not the thing itself. This means there are three distinct layers at work whenever you read a typeset word. A semantic layer. The word itself, editable and searchable, indifferent to appearance. A typographic instruction layer. Which font, which weight, which spacing, which OpenType features. A geometric rendering layer. The actual curves that appear on screen or paper, computed on demand. Live typography is procedural. The document stores the first two layers. The third is generated whenever needed. ## What outlining does When you outline text, you collapse all three layers into the third. The word becomes a set of fixed curves. It is no longer editable. It is no longer connected to the font file. The semantic and typographic layers are gone. This is why outlining solved our problem: the resulting file no longer needed the font at all. But it also froze the geometry permanently. What remained was an archival artifact of one particular rendering, cut loose from the system that produced it. 〰️ Outlining is irreversible not because something is technically lost, but because two entire layers of information are discarded. You cannot reconstruct a generative system from its output. ## The designers were right I used to find the font hoarding behavior of designers slightly neurotic. The obsessive downloading, the local archives, the backup anxiety. Now I understand it. Fonts come and go. Foundries discontinue typefaces. Licenses expire. Files get corrupted. And when a font disappears, it is not like losing a document. It is like losing a grammar. You may still have everything you ever produced with it, outlined and archived, but you have lost the ability to generate anything new. Every outlined artifact is a frozen instance. The living system that produced it is gone. That is a quietly different kind of loss than a missing file. A discontinued font is a small extinction. The practical workaround my wife suggested took thirty seconds. The shift in how I think about [what these digital tools actually are](https://rob-hoeijmakers.ghost.io/model-meaning-etymology/) took a little longer. ### Back in the Search Arena, This Time With Vectors URL: https://rob-hoeijmakers.ghost.io/back-in-the-search-arena-this-time-with-vectors/ Last updated: 2026-06-01T18:40:58.000Z Three years ago [I wrote about site search ](https://rob-hoeijmakers.ghost.io/website-search/)as if it were a solved problem. Collect, index, serve. It made sense at the time. Ghost had a search widget, it worked well enough, and I moved on. What I did not register then was the ceiling. Ghost's built-in search indexes titles and excerpts only. The body of every post, where the actual thinking lives, stays invisible. For a site with a handful of posts, that is fine. For a site with six hundred, it means your own archive is mostly unsearchable. I knew this. I just never had the right combination of tools, time, and nerve to do something about it. ## The return This spring, that changed. I came back to the problem with Claude Code, which turned out to be the missing piece. Not because it writes perfect code, but because it removes the activation energy. The setup that used to require a developer and a week of back-and-forth now takes an afternoon and a working session. I integrated [Algolia](https://algolia.com). Fast, full-text, generous free tier. It indexes everything: title, excerpt, and the complete body of every post. The difference is immediate. Searching my own archive now works the way I think, not the way I once phrased a headline. That alone would have been enough. But I kept going. [SearchBrowse by topic AI Strategy AI Governance AI in Practice Future of Work Europe![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/icon-png-0312355708b9690a58b50120448ac5df8933071d1f6c0d3fa8e23ccf370e506e.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_3034-1-1-24331370bb4b5da83ee249ca7be247a5a8d3696edac79c4db70f070dd14ae2f6.jpeg)](https://rob-hoeijmakers.ghost.io/search/) ## The semantic layer Keyword search finds what you named. It does not find what you meant, what you explored without naming it, or what connects two posts you wrote two years apart without realising they shared a thread. For that, I built a vector database on Cloudflare. Each post is converted into an embedding, a numerical representation of its meaning. When you search, the system does not match terms: it finds conceptual proximity. You ask in plain language, and it surfaces what is semantically related, even when the words never overlap. This is the shift that [search in the age of AI](https://rob-hoeijmakers.ghost.io/search-in-the-age-of-ai/) points toward: from retrieval to understanding. I have been writing about that shift for a while. Now I have a small working version of it on my own site. ## The pipeline Together, the two layers form a RAG pipeline. [Retrieval-augmented generation](https://rob-hoeijmakers.ghost.io/exploring-retrieval-augmented-generation-rag/): a way of grounding AI output in a specific corpus rather than in general training data. In enterprise contexts, this is how [knowledge bots and AI assistants](https://rob-hoeijmakers.ghost.io/scaling-rag-applications-with-weaviate/) are built on top of internal documentation. On the scale of one writer with six hundred posts, it works differently but toward the same end. When I draft something new, I can query the pipeline: what have I already written here? What connects? What internal links belong in this piece? The AI does not guess based on what it knows about the world. It searches what I have actually written and brings it back. Every step of building this taught me something I could not have learned by reading about it. The difference between keyword and semantic retrieval is obvious in theory. It becomes concrete the first time a search returns a post you had forgotten, on a topic you never explicitly named, because the meaning was close enough. That is not a feature. That is a different relationship with your own archive. ## What it synthesises I am familiar with data. I have worked with it long enough to know that structure matters more than volume. What this pipeline does is give structure to six hundred pieces of thinking that were previously just chronological. The archive did not grow. It became navigable in a way it was not before, for visitors and for me and for the AI I work with. That last part is the one I keep returning to. Not as a novelty, but as a genuine shift in how I can build on what I have already made. ↗️ The vector database runs on Cloudflare Workers AI with D1 for storage. Retrieval is powered by BGE embeddings in Cloudflare Vectorize. Check it out here for the query: [search meaning vectors](https://hoeijmakers-search.rob-hoeijmakers.workers.dev/search?q=search+meaning+vectors) ### The Axis That Made the Chips URL: https://rob-hoeijmakers.ghost.io/the-axis-that-made-the-chips/ Last updated: 2026-05-26T06:49:21.000Z In the summer of 1952, the first Dutch computer was switched on at the Mathematisch Centrum in Amsterdam. It filled an entire classroom. It was built from 1,200 relay switches and could store about 30 bits of information. At its launch it worked once, briefly, and was then taken out of commission. It was called the ARRA, short for Automatische Relais Rekenmachine Amsterdam. Its builders, Carel Scholten and Bram Loopstra, immediately started over. The ARRA II worked. And that same year, the Mathematisch Centrum hired a young programmer named Edsger Dijkstra. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/05/1779273794632.jpeg) Edsger Dijkstra (left), Bram Loopstra and Ria Debets at the Mathematisch Centrum in Amsterdam (1954). ## Amsterdam Dijkstra spent the first decade of his career at the Mathematisch Centrum. In 1956, sitting in a cafe in Amsterdam without pen or paper, he worked out a solution to the **shortest path problem** in about twenty minutes. He published it three years later. It is now used in every navigation system on the planet. In 1960, together with Jaap Zonneveld, he completed the first working compiler for **ALGOL 60**, beating every other team in the world by more than a year. The Mathematisch Centrum was also, that same year, the institution from which Electrologica was spun off: the first Dutch computer manufacturer, building the X1 and later the X8 in Rijswijk. The thinking and the making were already connected, even before Dijkstra moved south. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/05/1526912454369976320-FTCuyDBXoAEovgN-2.jpg) ## Eindhoven In 1962 he joined the Mathematics Department at the **Technische Hogeschool Eindhoven**. There he built an operating system for the Electrologica X8\. The system was called THE, simply the initials of the university. He never chose the name deliberately. THE was built by six people, working half-time. Its architecture was organised into six strict hierarchical layers, each able to use only the functionality of the layers beneath it. Processes were coordinated using semaphores, a synchronisation primitive Dijkstra had also invented. The paper appeared in Communications of the ACM in 1968, won the ACM Programming Systems and Languages Award in 1971, and is still read in computer science courses today. The following year, [Dijkstra received the Turing Award](https://hoeijmakers-search.rob-hoeijmakers.workers.dev/search?q=thinking+building+connected), computing's highest honour, for his contributions to structured programming. What THE demonstrated was that structure is not a constraint on ambition. It is what makes ambition hold. That logic has not aged: the systems that compound are the ones built with intention from the start. ## From Rijswijk to Veldhoven The people who built the ARRA and the X8 did not disappear when those machines did. Loopstra moved to Electrologica and later to Philips. Scholten did the same. The knowledge migrated into the industrial infrastructure of the Eindhoven region. Philips was already there. From Philips, in 1984, came [ASML](https://rob-hoeijmakers.ghost.io/zeiss-and-asml/), starting in a leaky shed, building lithography machines. ASML now makes the only equipment capable of producing the world's most advanced chips. Every leading chipmaker depends on it. That continuity is not coincidental. It is what happens when theoretical rigour and a culture of making occupy the same small country for long enough. ## Why this is not just history Amsterdam and Eindhoven are in conversation again about working more closely on technology. The framing is **AI, talent, and infrastructure**. That framing is right. But the historical depth is worth holding on to, because it clarifies what the axis actually produces. The Netherlands has theoretical institutions of real quality. It has the most critical piece of hardware infrastructure in the global semiconductor industry. It has a tradition, visible from the ARRA to THE to ASML, of building things that matter from first principles, with limited resources, and getting them right. Europe is searching for this combination. [The question of whether good ideas have somewhere to live, somewhere to scale, is unresolved](https://rob-hoeijmakers.ghost.io/vision-judgement-creativity-reclaiming-agency-in-the-age-of-ai/). The Dutch answer to that question is already seventy years old. It is worth reading carefully. ### Passkeys: Better Lock, Borrowed Door URL: https://rob-hoeijmakers.ghost.io/passkeys-better-lock-borrowed-door-2/ Last updated: 2026-05-25T09:38:57.000Z I wrote about passkeys in early 2024 with some optimism. The technology is elegant. The problem it solves is real. I still think both those things. But I left something out. ## What passkeys actually fix The password is a structural failure. It is a secret you have to remember, type, transmit, and trust a server to store correctly. Any one of those steps can go wrong, and the weakest one determines your security. Phishing works because passwords travel. Data breaches work because servers collect them. Credential stuffing works because people reuse them. A passkey replaces all of that with a cryptographic key pair. Your device holds the private key. The service holds the public key. When you authenticate, your device signs a challenge; the server verifies the signature. Nothing is transmitted that could be intercepted or stolen. I wrote about [making that transition](https://rob-hoeijmakers.ghost.io/transitioning-to-passkeys/) in early 2024, and the improvement was real. It is not marketing. Back in 2022, when [passkeys were still theoretical](https://rob-hoeijmakers.ghost.io/passwords-passkeys/), I noted the convenience problem: helping a six- or eighty-year-old through a QR-code flow would be a challenge. What I did not see then was the deeper problem that convenience would bring with it. ## The key card stays the same Here is what does not change. Once you are through the lobby, the server hands you a session cookie: a small text file that carries your authenticated state from that point on. The elaborate verification at the front door collapses into a string of characters sitting in your browser. Copy that cookie and you are in, from anywhere, as anyone. The [key card problem](https://rob-hoeijmakers.ghost.io/session-cookie-authentication-security/) is untouched by how you checked in. Passkeys make the lobby better. They do not change what happens after it. ## The dependency There is a more fundamental issue. Passkeys do not live in your head, and they do not live on a neutral device. They live in a keychain: Apple's, Google's, or Microsoft's. That is where they are stored, and that is where they sync. The sync is not optional. Use more than one device and you need it. Switch from iPhone to Android and your passkeys stay behind. Add a Windows machine to an Apple setup and you are managing two separate ecosystems, each with its own rules. The line between digital convenience and [infrastructural dependency](https://rob-hoeijmakers.ghost.io/passports-phones-and-the-future-of-identity/) is vanishingly thin -- and you have crossed it the moment you need your credentials on a second device. This is a structural dependency, not a technical inconvenience. The platform controls your credentials. If you leave, you lose them. If you are locked out, you lose them. If the platform changes its policies, you are subject to those changes. I experienced a version of this [when X locked me out](https://rob-hoeijmakers.ghost.io/locked-out-by-x/) last year: the burden of proof shifts to you, the process is opaque, and there is no neutral party to appeal to. With passkeys, the stakes are higher. It is not one platform. It is your authentication infrastructure. ## A different kind of lock-in With passwords, the risk was theft. With passkeys, the risk is captivity. These are not equivalent problems, and the second one is quieter. Most users will not notice. The experience is genuinely better: a biometric check, no password to forget, no phishing surface. The dependency is invisible until it is not. That is how the best lock-ins work. Europe is working on a different answer, one built on the premise that [digital identity should not be owned by a platform](https://rob-hoeijmakers.ghost.io/digital-identity-id-empowerment-or-state-control/). That is a longer conversation. But it starts from exactly the right diagnosis: the front door matters, and so does who holds the key. 🔐 ****Passkeys** are supported by Apple, Google, and Microsoft through their respective keychain and credential manager products. Cross-platform sync between ecosystems is not natively supported – moving between platforms requires manual credential migration or re-registration with each service. ### The Session Unlocks the Door. For Anyone. URL: https://rob-hoeijmakers.ghost.io/session-cookie-authentication-security/ Last updated: 2026-05-24T08:17:25.000Z A few weeks ago I ran a command that made me stop. I was downloading a video using yt-dlp, a command-line tool, and instead of logging in I passed a single flag: `--cookies-from-browser chrome`. The tool reached into my browser, borrowed my active session, and acted as me. The platform never noticed the difference. That flag is a small thing. But it points at something large. ## Three things, not one Most people experience logging in as a single ritual. You open a site, you prove who you are, you are in. But there are actually three distinct layers, and conflating them is how the weak spot stays hidden. The first is the account: your identity registered in a system. [Creating one is often deliberately friction-heavy](https://rob-hoeijmakers.ghost.io/accounts-profiles-identity/) — email verification, phone numbers, sometimes document checks. That friction is intentional. The account is meant to mean something. The second is authentication: proving, each time, that you are the account holder. A password, a code sent to your phone, a biometric scan. This is the gate ritual, and it is where the industry puts most of its visible effort. The third is the session cookie: the lightweight token the service issues once you have passed the gate. From that point on, the cookie does the work. You stop re-proving yourself at every door. The cookie says: this person already checked in. ## The key card Think of checking into a hotel. You book a room, you show your passport at reception, and then you get a key card. From that point on, the card does the work. It opens your room, the gym, the parking barrier. Nobody asks for your passport again until checkout. The elaborate front-desk ritual gives you a sense of security. But what actually carries you through the building is a small piece of plastic with a magnetic stripe. Session cookies work the same way. You create an account, you authenticate, and the service issues a token. That token is your key card. And here is what I did not fully appreciate until recently: that card often sits as a plain file on your disk, accessible to any tool running on your machine. yt-dlp did not break into anything. It just picked up the key card I had left on the table. 🍪 Session cookies are stored locally by your browser. On most systems, tools with user-level access can read them without elevated permissions. Logging out of a service invalidates the server-side session, which is one reason it matters more than just closing the tab. ## The asymmetry Platforms invest heavily in the gate. Two-factor authentication, device recognition, login anomaly detection. The entry experience has become genuinely robust. But once the cookie is issued, many platforms let it run for weeks or months. Some bind it to a domain, few bind it to a specific device or IP address. That is [the delegation problem stated differently](https://rob-hoeijmakers.ghost.io/the-delegation-problem-of-the-internet/): authentication systems were built around a single person at a single keyboard. The session token they produce was never designed to be transferred, borrowed, or held by an automated tool. Yet that is exactly what is happening, quietly, every time a CLI tool, a script, or an AI agent acts on your behalf using credentials you already established. ## What this changes For me, this sharpened two things. The first is personal. I feel less secure online than I did before understanding this. Not because anything has changed in the world, but because I now see where the actual surface area is. The gate is solid. The key card is lying on the table. The second is professional. [Delegation across digital systems](https://rob-hoeijmakers.ghost.io/decentralised-identity-prove-what-you-must-reveal-no-more-than-you-need/) has always seemed technically hard. Giving an AI agent or a colleague selective access to act on your behalf without sharing your credentials felt like a problem requiring elaborate infrastructure. It still does, done properly. But the session cookie shows that a cruder form of delegation is already available, already working, and mostly invisible. That is not reassuring. It is clarifying. The elaborate check-in was never the whole story. It was the part we were meant to notice. ### Stripping the medium URL: https://rob-hoeijmakers.ghost.io/stripping-the-medium/ Last updated: 2026-05-21T05:54:05.000Z McLuhan said the medium is the message. He meant it as an observation, not a problem to solve. I've been treating it as a problem to solve. For the past year, I've been converting almost everything I consume into plain text. YouTube videos, podcasts, meetings, articles behind paywalls I pay for. Not to archive them. To read them, think with them, and feed them to language models. The format I end up with, every time, is markdown. ## The oldest instinct This is not a new idea. The monk copying a manuscript was doing the same thing: separating the information from its carrier so it could travel further. The printing press industrialised that instinct. The paperback made it cheap. The e-reader made it weightless. What changes in each era is not the instinct but the friction. The question is always: how much does the medium resist you? Right now, audio resists. Video resists. The [meeting](https://rob-hoeijmakers.ghost.io/fireflies-ai-transcriptions/) resists most practically of all: it happens once, in real time, and then it exists only in whoever was paying attention. The beautifully designed magazine website resists in a different way, because it was built to be experienced, not extracted. McLuhan would have loved it. The medium has become so expressive it almost swallows the message entirely. Markdown is my answer to that. Not because it is elegant, but because it gets out of the way. ## Linear in a parallel world I am a linear reader. I process information sequentially, with attention, one thread at a time. That puts me at a structural disadvantage in an information environment built for parallel consumption: the feed, the notification, the autoplay queue. The workflows I use are, at their core, a way of reasserting linearity. A podcast becomes a text I can read at the pace I choose. A YouTube video becomes something I can skim, annotate, return to. A meeting becomes a document I can search rather than a memory I have to trust. A paywalled article becomes a file, not a session. The sweet spot I am looking for is high information density with low format noise. Markdown delivers that. Everything else is packaging. ## Why not plain text Plain text and [markdown](https://rob-hoeijmakers.ghost.io/markdown-the-wd-40-of-digital-information/) look almost identical. For reading, the difference is marginal. For everything else, it matters. A markdown file can carry **frontmatter**: a small block of metadata at the top of the file that describes what it contains. Source, date, speaker, topic, context. The file becomes self-describing. You do not need an external index or a carefully named folder structure to know what you are looking at six months later. For a language model, that difference is significant. When I pass a transcript to Claude or ChatGPT, the frontmatter travels with it. The model knows where the text came from, when, and why I kept it. I am not just handing over words. I am handing over a small, structured piece of context. Plain text is readable. Markdown with frontmatter is readable, findable, and passable. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/05/Medium-change.jpg) Shortcuts to backend transcription flows. Very seamless. ## The second reader Here is where something genuinely new enters. For most of history, converting information to a more portable format had one destination: a human reader. The monk, the scholar, the commuter with a paperback. The conversion served attention. I am converting for two readers now. Myself, and the language model I will hand the text to next. A transcript I made three weeks ago becomes context in a conversation today. A meeting summary from last month becomes the briefing for this week's follow-up. A collection of articles on the same topic becomes a working corpus, not a reading list. That second reader changes the logic of the whole operation. I am not just a consumer of information. I am a curator preparing material for a machine that will help me think with it. The format has to work for both of us. Markdown does. McLuhan observed that every new medium contains the previous one. Television contains radio. The web contains print. What contains the language model? I think it might be the archive: the personal, intentional collection of text that someone has decided is worth keeping close. Gleick wrote in *The Information* that the history of communication is the history of encoding. What I am doing is, in that sense, unremarkable. What is different is who the encoded message is for. The same logic works in reverse. From a markdown file, I produce blog posts, documents, structured reports. The format is not a destination. It is the place everything passes through. Which is why seamless is not a feature. It is the only requirement. Add one step that requires attention, and the whole thing collapses. We are, after all, fundamentally lazy. Good systems know that about us. ### Unpack on Arrival URL: https://rob-hoeijmakers.ghost.io/unpack-on-arrival/ Last updated: 2026-05-16T12:23:23.000Z I sat in enough meetings where a developer explained why the new site needed to be built in React. The arguments were good. Component reusability. A single codebase. The ecosystem. The talent pool. All of it technically defensible, some of it genuinely compelling. I kept choosing tools that delivered tangible markup. HTML, ready on arrival, with the content already in it. ## What rendering actually means Here is where the terminology gets slippery, and it is worth pausing on because the confusion is load-bearing. When most people hear "rendering" they think of the browser: pixels appearing on screen, a page becoming visible. That instinct is correct. The browser always renders in that sense. But the web industry borrowed the word for a second meaning, and the two have been quietly creating confusion ever since. The question **client-side versus server-side rendering** is actually asking is: who assembles the HTML before the browser gets it? A server-side rendered site sends complete HTML in its first response. The content is there, in the markup, before the browser does anything. A client-side rendered site, the default pattern in React, sends something closer to a flat-pack: a minimal HTML shell and a JavaScript bundle. The browser downloads the bundle, executes it, fetches data from an API, and only then assembles the page. Rendering in the original sense happens last, after all of that. For a highly interactive application, this sequence makes sense. For a site whose primary job is to carry text from a writer to a reader, it introduces an invisible wall between the content and anyone trying to reach it. 👻 Ghost, the publishing platform this site runs on, has been server-rendered since its first release in 2013\. It runs on Node.js, the same runtime behind React and Next.js, but uses it to assemble complete HTML on the server before delivering it. Next.js brought server-side rendering to React in 2016\. React Server Components, which never ship JavaScript to the client at all, arrived with React 19 in late 2024. ## The industry correction Search engines hit that wall first. Googlebot would arrive at a React site, find an empty shell, and either wait for JavaScript execution or index nothing. Developers built workarounds. React added server-side rendering as an option. Next.js made it the architecture, turning React from a client-side library into a full framework capable of generating HTML on the server. Later still, React Server Components went further: components that run only on the server and never ship JavaScript to the client at all. The ecosystem spent years building infrastructure to fix a mismatch between the tool and the use case. The cost was real. DoorDash's original React frontend grew until the JavaScript bundle became, in their own engineering team's words, difficult to optimise. Pages would show a blank screen while the browser waited to finish assembling them. Users saw nothing, then everything, with an uncomfortable gap in between. They migrated to Next.js page by page, and the improvement was immediate: pages loaded with content visible from the first response, the way a server-rendered site always had. That is what the invisible wall costs at scale. The publishing tools that got this right were server-rendered from the start. The content is in the HTML on first response. No JavaScript execution required to read it. Many of them run on Node.js, the same runtime that powers the broader JavaScript ecosystem. The difference is what they use it for: assembling HTML on the server before anyone arrives, not shipping the assembly work to the browser. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/05/client-versus-server-side-rendering.jpeg) Server Sider Rendering (SSR) versus Client Side Rendering (CSR) ## The same wall, new visitors AI crawlers have the same limitation search bots always had. They read HTML. They do not reliably execute JavaScript. A React site that was half-invisible to Google is equally invisible to the systems building the retrieval indexes behind Perplexity, ChatGPT, and Claude. The stakes are slightly different now. With search, poor rendering meant a lower ranking. With AI retrieval the dynamic is more binary: either the content was there when the crawler arrived, or it was not. There is no page three. Either your writing contributed to the answer, or it did not exist. I have been doing deliberate work on this side of things: [llms.txt](https://rob-hoeijmakers.ghost.io/llms-txt/) to signal what is here and how to use it, content cached and served as plain [markdown](https://rob-hoeijmakers.ghost.io/markdown-the-wd-40-of-digital-information/) via a Cloudflare Worker so machine readers can find the cleaner format. That work only makes sense if the content is actually present in the first place. The architecture has to be right before any of the signalling matters. ## What the developers were solving The JavaScript-first instinct was not wrong. React solved real problems: rich interactivity, stateful interfaces, products that behave like applications. The problem was category drift. The same approach got applied to publishing, to marketing sites, to anything built by a team that knew React well and reached for the familiar tool. The developers I argued with were not wrong about their frameworks. They were wrong about what a content site needed from its architecture. Those are different questions, and for a long time the industry treated them as the same one. My instinct was about content, not frameworks. Content needs to be there when someone arrives: human or machine, search bot or AI crawler, fast connection or slow one. Everything else is optimisation. The AI era did not change that requirement. It just made the cost of getting it wrong more visible. ## Further reading - [Markdown: the WD-40 of Digital Information](https://rob-hoeijmakers.ghost.io/markdown-the-wd-40-of-digital-information/) \-- on why markdown has outlasted everything built around it - [llms.txt](https://rob-hoeijmakers.ghost.io/llms-txt/) \-- signalling to AI systems what your site is and how to use it - [My Visitors Are Not All Human. That Is Fine.](https://rob-hoeijmakers.ghost.io/my-visitors-are-not-all-human-that-is-fine/) \-- on bots as guests, and the distinction between those who identify themselves and those who don't ### RSS is not dead. It just changed audience. URL: https://rob-hoeijmakers.ghost.io/rss-is-not-dead-it-just-changed-audience/ Last updated: 2026-05-15T19:19:09.000Z My RSS feed gets more traffic than my homepage. Not from subscribers with a feed reader open on a Sunday morning. From machines. RSS, Really Simple Syndication, is the quiet pipe that lets you follow a website without visiting it. Old technology, XML-based, pre-social-media, designed in an era when "web application" meant a browser and a human. Publishers quietly stopped promoting it. Google Reader closed. The format seemed to be fading into dignified retirement: still useful, still honest, increasingly invisible. Then AI happened to the web. ## The machine-readable web AI crawlers, large language model pipelines, retrieval systems: these are the readers that now constitute the majority of requests on many independent sites. I can see this in my own logs. Human visitors with a browser are, depending on the day, somewhere between ten and twenty percent of total traffic. The rest is automation of one kind or another. For these systems, RSS is not a legacy format. It is a clean, structured, semantically coherent signal that something new is available. No JavaScript. No cookie banners. No layout to parse. Just content, metadata, and a timestamp. From a machine's perspective, RSS is better than a homepage. ℹ️ RSS is XML-based. Each entry contains a title, link, publication date, and usually a summary or full content. Feed readers and bots alike use it to check for updates without crawling a full site. Here you can [view traffic like RSS Readers](https://rob-hoeijmakers.ghost.io/bot-stats) on this site. ## The publisher's dilemma Publishers always had complicated feelings about RSS. The feed hands content directly to whoever asks: no analytics, no ad impressions, no engagement loop. The reader is in control. The publisher loses the visit. That tension never resolved; it just became less urgent as social platforms took over distribution. Now the [machines have arrived](https://rob-hoeijmakers.ghost.io/my-visitors-are-not-all-human-that-is-fine/) with the same expectation. Structured access, no friction, no visit required. The format designed to empower human readers turns out to be exactly what machine readers need. The audience changed. The value proposition did not. ## The human case, still I use [Reeder](https://reederapp.com). Have for a while. It is one of those apps that works so well it stops drawing attention to itself: feeds come in, I read what matters, I close it. No algorithm deciding what I see next. Just a chronological list of things people I chose to follow decided to publish. That experience is worth naming because it is genuinely different from how most content reaches people now. Slower, more selective, harder to game. ## A second youth The revival of RSS was not planned by anyone. It followed from two pressures arriving at the same moment: the platformization of everything, which made owning your own distribution feel more urgent, and the AI turn, which made structured content feeds useful again for entirely non-human reasons. For a format to find a second life, it usually needs to solve a new problem. RSS solved the same problem it always solved. The problem just acquired a new population. That is worth paying attention to. Not because RSS is exciting, but because formats built around clarity and structure tend to survive. Clarity is useful. To humans on a Sunday morning, and apparently to machines at any hour. ### The Detective and the Swarm URL: https://rob-hoeijmakers.ghost.io/the-detective-and-the-swarm/ Last updated: 2026-05-09T12:41:25.000Z At 07:43 UTC this morning, something hit my site. Not a flood in the security sense: no alarms, no WAF triggers, nothing in the human analytics. Just a number that shouldn't have been that high, on a chart I check more out of habit than worry. Four hundred requests in sixty seconds. Then back to normal. My human-facing analytics saw nothing. That absence is itself a clue. ## Two layers Most publishers have one view of their traffic: the analytics dashboard. It shows pageviews, sessions, referrers, the countries their readers come from. It tells them what humans did. What it doesn't show is everything else. And everything else, it turns out, is interesting. I run a Cloudflare Worker that logs every request to a D1 database before passing it along. Every request: [human, bot, crawler, scraper, agent](https://rob-hoeijmakers.ghost.io/bot-stats). The Worker tries to classify each one, matching user-agent strings against a database of known bot signatures. What it can't classify, it logs as unknown. That second layer is where the detective work happens. ## Reading the evidence The 07:43 spike broke down like this: 202 of those 400 requests were for `.md` paths. Paths like `/when-bots-become-readers.md`, `/web-traffic-and-the-rise-of-llms.md`, `/measuring-traffic-machines-bots.md`. The Worker classified 192 of them as human, because the user-agent strings looked like browsers: Chrome 138, Firefox 115, Edge 114\. Perfectly formed, perfectly plausible. But one user-agent hit 47 different `.md` paths in sixty seconds. Another hit 37\. Both from the same two browser fingerprints, distributed across 25 countries. Then the robots.txt requests: 116 of them in the same minute. That's a preflight pattern, a swarm checking the rules before it reads the content. Chrome 114 hasn't been a current browser for a long time. Neither has Firefox 115\. These are frozen strings, a signature of bot infrastructure that picks a browser version and pins it, never updating. The User Agent (UA) looks human. The behaviour doesn't. The conclusion assembled itself: a distributed scraper, running across a proxy network or botnet, using spoofed browser identities to avoid classification. Evasive, coordinated, and genuinely clever. ## The twist Here's where the detective story gets uncomfortable. Those `.md` endpoints don't exist by default in Ghost. I added them. A few months ago, as an experiment: serve each post as clean Markdown alongside the HTML version, reference them in `llms.txt`, see what happens. The idea was to make the content easier for AI systems to consume. Structured, clean, no JavaScript noise. The scrapers found them almost immediately. So did legitimate AI users, ChatGPT-User and OAI-SearchBot among them, reading the same paths through the same door. I set the bait. They smelled it. ## The idée fixe The reflex response to a traffic spike like this is defensive. Block the IPs, rate-limit the endpoint, add a CAPTCHA, harden the WAF. There is an entire industry built around that reflex. It rests on a premise worth examining: that keeping machines out is possible, and that it is worth the effort. Neither is quite true. A scraper that can distribute across 25 countries and rotate frozen browser UAs is not stopped by a robots.txt entry or a Cloudflare rule. It routes around friction the way water routes around a stone. And the content, once published on the public web, is going to be consumed by machine pipelines whether or not you make it easy. The more interesting question is what you can learn by watching it happen. The spike told me that my [Markdown](https://rob-hoeijmakers.ghost.io/markdown-the-wd-40-of-digital-information/) experiment is working, in the sense that it is attracting exactly the traffic it was designed for. It told me that the machine layer of the web is active, distributed, and more sophisticated than most people assume. It told me that the gap between what human analytics show and what the full request log shows is where the real picture lives. Blocking that traffic would have closed the window. Watching it left it open. ## Signal in the noise The thing about running a second logging layer is that most of what it catches is unremarkable. Googlebots, Bingbots, ChatGPT-User ticking through recent posts, the usual crawl of SEO tools and RSS readers. Noise. But the noise is the baseline. Without it, the 07:43 spike is invisible. With it, you can ask: what's different about this minute? Why these paths? Why this many countries? Why frozen UAs? The detective work is in the filtering, not the blocking! 🗒️ The .md endpoints at hoeijmakers.net are intentional. Each post is available as clean Markdown alongside the HTML version. The llms.txt file indexes them. This is an ongoing experiment in machine-readable publishing. --- Related: - [I Thought I Was Optimising for Speed](https://rob-hoeijmakers.ghost.io/i-thought-i-was-optimising-for-speed/) - [Thirty Years of Caching, Sorted in an Afternoon](https://rob-hoeijmakers.ghost.io/thirty-years-of-caching-sorted-in-an-afternoon/) - [My Visitors Are Not All Human. That Is Fine.](https://rob-hoeijmakers.ghost.io/my-visitors-are-not-all-human-that-is-fine/) - [Guests That Should Behave](https://rob-hoeijmakers.ghost.io/guests-that-should-behave/) - [Markdown, the WD-40 of Digital Information](https://rob-hoeijmakers.ghost.io/markdown-the-wd-40-of-digital-information/) ### Markdown, the WD-40 of Digital Information URL: https://rob-hoeijmakers.ghost.io/markdown-the-wd-40-of-digital-information/ Last updated: 2026-05-14T06:03:57.000Z There is a can of WD-40 in almost every workshop in the world. It doesn't do any one thing brilliantly. It loosens, lubricates, displaces moisture, protects against rust. It works on hinges, bolts, zippers, bike chains, garden tools. The reason it's everywhere is not that it's the best at anything. It's that it's useful for almost everything, costs almost nothing, and leaves no mess. Markdown is the WD-40 of digital information. ## Just enough structure John Gruber published the original Markdown specification in 2004\. The idea was simple: a plain text format that reads naturally as text but converts cleanly to HTML. A `#` becomes a heading. `**bold**` becomes bold. A blank line between paragraphs is a paragraph break. No tags, no toolbars, no proprietary file format. Twenty years later, the spec has barely changed. The software around it has been rebuilt dozens of times. The platforms it feeds into have come and gone. Markdown itself just kept working. That persistence is worth examining. It's not inertia. Rich Text Format persisted for decades out of inertia, dragging its bloated XML through every Office suite until it quietly died. Markdown persisted because it kept finding new uses its creator didn't anticipate. The reason is the sweet spot it occupies: between content and meaning. ## Content and meaning Most formats choose a side. HTML carries meaning but not content: the tags tell you what something is, but the structure is invisible to casual reading. A human can read raw HTML, but it requires effort. The format is for browsers, not people. A Word document carries content but obscures meaning: it looks right on screen, but extract the text and the structure collapses. Headings become large bold text. Tables become chaos. The meaning was in the rendering, not the file. Markdown carries both, lightly. A heading is marked as a heading, not just styled to look like one. Emphasis is marked as emphasis. A list is a list. But the markup is so minimal that a human reading the raw file sees structure without syntax noise. The `#` before a heading reads almost like an underline. The `**` around a word barely interrupts the flow. That dual legibility is the trick. A human reads it as text. A machine reads it as structured data. Neither has to compromise. ## Where it keeps showing up GitHub made Markdown the default format for READMEs in 2009\. Not because it was mandated, but because developers were already using it in plain text files and it rendered cleanly. Stack Overflow adopted it for questions and answers. Notion built an entire productivity platform around it. Obsidian turned it into a knowledge management system. Slack, Discord, WhatsApp: all support some dialect of Markdown for message formatting. Every time a new context needed lightweight structure, Markdown was already there. AI systems followed the same logic. Training data in Markdown is cleaner than HTML: no scripts, no navigation chrome, no cookie banners embedded in the text. The structure survives the extraction. A heading in Markdown is still a heading after ingestion. That's not true of most formats. When AI assistants started browsing the web and retrieving content for users, Markdown became the preferred output format for responses. Not because it was specified. Because it was the format that travelled best between systems. ## A deliberate experiment A few weeks ago I added something to this site: a Markdown copy of every post, served alongside the HTML version. A path like `/this-post.md` returns the same content as `/this-post/`, but as clean Markdown text. I also added an [llms.txt](https://rob-hoeijmakers.ghost.io/llms-txt/) file indexing all of them, a signal to AI systems that structured content is available here. The experiment was modest. Serve the content in the format machines prefer, see what happens. What happened was [immediate](https://rob-hoeijmakers.ghost.io/bot-stats). AI crawlers found the `.md` paths within days. ChatGPT-User, OAI-SearchBot, and others started reading through them systematically. And when a swarm of scrapers hit the site recently, 202 of their 400 requests in a single minute were for `.md` paths. They didn't find those endpoints by accident. They went looking for them, the way you go looking for the WD-40 when something needs loosening. The Markdown layer attracted exactly the traffic it was designed for. The format did the work. ## Why slim formats win RTF tried to carry everything. SGML tried to define everything. Both are essentially dead for everyday use. The formats that survive are the ones that carry just enough. Markdown carries hierarchy, emphasis, links, code blocks, and lists. It leaves everything else alone. That restraint is what makes it composable: you can feed it into a static site generator, a documentation tool, an AI pipeline, a note-taking app, a chat interface. It doesn't arrive with requirements. It arrives ready. The web itself works this way. HTTP is a slim protocol. HTML, at its core, is a slim format. The richness came later, layered on top. The slim foundation is why it all still runs. Markdown found its sweet spot and stayed there. Two audiences, one format, twenty years, no mess. That's not luck. That's the point. --- Related: - [Using markdown for blogging](https://rob-hoeijmakers.ghost.io/using-markdown-for-text/) - [Markdown in Google Docs: A Practical Step Forward](https://rob-hoeijmakers.ghost.io/markdown-google-docs/) - [The Detective and the Swarm](https://rob-hoeijmakers.ghost.io/the-detective-and-the-swarm/) ### The shift from x86 to ARM is about power, not just performance URL: https://rob-hoeijmakers.ghost.io/x86-to-arm/ Last updated: 2026-05-10T08:53:29.000Z The new iMac is about half the weight of the one it replaced. I noted that in an [earlier post ](https://rob-hoeijmakers.ghost.io/a-lighter-computer-for-a-different-era/)and left it there, a physical curiosity. What I did not fully grasp at the time was that the weight difference is not a design choice. It is a consequence of a decision made in 1985 by a British company called Acorn Computers. A [short YouTube video](https://youtu.be/d-8M6Rks860?si=YzCAduOCAeFaSxm1) made it click. ## Two languages Every processor speaks a language. The architecture defines that language: what instructions the chip understands, how it processes them, how it talks to the rest of the machine. x86 and ARM are simply two different languages. Same job, very different dialects. x86 goes back to 1978\. Intel's 8086 chip started a family tree that runs unbroken to every Windows PC, every AMD Ryzen, every server rack powering Google and Netflix today. Forty-five years of backward compatibility, each generation carrying the weight of the previous one. ARM came seven years later, from a different starting point. Acorn's engineers wanted a chip that was simpler, faster, and used less power. They called it the Acorn RISC Machine. The insight was almost contrarian: instead of building chips that could handle complex instructions in a single step, strip it down to simple ones executed quickly and cheaply. Most programs, it turned out, never needed the complexity anyway. ## Where the weight goes x86 chips are CISC: complex instruction set computing. A single instruction can do multiple things at once. That capability requires elaborate decoding hardware, transistors devoted not to computing but to figuring out what they have been asked to compute. At billions of operations per second, that overhead is not trivial. ARM is RISC: reduced instruction set computing. Fixed-length instructions, a simpler decoder, more transistors left over for actual work. The result: ARM chips do less per instruction but do it faster, cooler, and on far less power. A typical x86 desktop chip draws between 65 and 250 watts. A phone-grade ARM chip draws 1 to 15. That difference is why every smartphone runs ARM. You cannot run an x86 chip on a small battery. It would drain in an hour and burn your hand. And it is why the new iMac is half the weight of the old one. Apple Silicon is ARM. The lightness is not aesthetic. It is physics. ## The same logic, at scale What surprised me, reading further, is that the efficiency argument does not stop at the desk. Amazon Web Services has been deploying its own ARM-based Graviton processors in data centers, claiming up to 40% better price-performance compared to equivalent x86 options. At data center scale, lower power consumption translates directly into electricity costs. The logic that makes a laptop last eighteen hours on a charge also makes a server hall cheaper to run. The cloud is not a single machine. It is tens of thousands of them. ARM's advantage compounds accordingly. ## Convergence The cleaner story would be: ARM wins, x86 fades. The actual story is more interesting. Intel has added efficiency cores to its processors, smaller and simpler, borrowing directly from RISC thinking. ARM chips like Apple's M-series are no longer just efficiency champions; they are closing the gap on raw performance. The two architectures are converging, each absorbing what the other does well. x86 still leads in PC gaming, high-end creative workstations, and the entrenched server ecosystem. That will not shift quickly. Forty-five years of software compatibility is not something you dissolve in a product cycle. But the direction is clear. As energy efficiency becomes a constraint, at the battery level and at the grid level, ARM's fundamental design advantage becomes more relevant, not less. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/05/ChatGPT-Image-May-7--2026-at-10_09_58-PM.png) ## The lighter machine In the first post about this iMac, I wrote that local computers are becoming lighter, replaceable instruments. The ARM architecture is part of why that is true in a literal sense. Less heat, no fan required, a thinner chassis, half the weight. I assumed there was still a clear distinction between phone chips and Mac chips. Apple Silicon in a Mac meant M-series: ARM as the foundation, but scaled up, redesigned, a different member of the family. The iPhone stayed in one category, the Mac in another. The MacBook Neo ended that story. It is the first Mac to use an A-series chip, the same family found in the iPhone, rather than the M-series chips in other Macs. Reports suggest the first production run used binned A18 Pro chips, originally intended for the iPhone 16 Pro. Apple is shipping a laptop powered by recycled phone silicon. And it sells for $599. The A18 Pro is not a compromise chip. It runs at M3-to-M4 class performance for single-threaded work. The boundary between phone and computer, already blurry at the architecture level, has now dissolved at the product level too. What started as a design principle in 1985, do more with less, has worked its way from pocket to desk to data center. The lighter machine was never just about weight. 💡 ****The A and M series, briefly** Apple makes two families of ARM chips. The A-series powers iPhones and iPads: compact, optimised for a small battery, tuned for mobile workloads. The M-series powers Macs: the same ARM foundation, but scaled up with more CPU and GPU cores, more memory bandwidth, and support for larger pools of unified memory. Think of them as the same language spoken at different volumes. The MacBook Neo is the first Mac to use an A-series chip rather than an M. ### Further reading - [A lighter computer for a different era](https://rob-hoeijmakers.ghost.io/a-lighter-computer-for-a-different-era/) - [From Silicon to Intelligence: Understanding the Hardware Behind AI](https://rob-hoeijmakers.ghost.io/cpu-gpu-tpu-npu-explained/) ### I Thought I Was Optimising for Speed URL: https://rob-hoeijmakers.ghost.io/i-thought-i-was-optimising-for-speed/ Last updated: 2026-05-10T08:53:30.000Z I built a caching layer for my blog. The goal was performance. What I got, alongside that, was a clearer picture of my audience than I had ever had before. That was not the plan. ## What the numbers used to show me Before, I had Plausible. Clean dashboard, daily visitors, top pages. Reassuring in the way that summaries are reassuring: they tell you something is happening without telling you much about what. A blog attracts a certain kind of traffic. Some of it is people reading. Some of it is search indexers doing their rounds. Some of it is AI systems pulling content for retrieval pipelines. Some of it is automated probes checking whether your server has vulnerabilities. All of this shows up as traffic. Almost none of it is distinguishable in a standard analytics view. ## The crowd that arrived When the caching article hit Hacker News, thousands of requests came in within the hour. I had not planned for that. What I had planned for was having instruments in place when it happened. Every request that reaches the site now gets a label: human reader, AI crawler, SEO crawler, security scanner, unknown. Those labels feed a dashboard. So when the spike came, I could see what it was made of. Human readers, identifiable by dwell time and page depth: people who had seen the link, clicked, and stayed. RSS subscribers, a category I had underestimated: readers who had deliberately chosen to follow the feed, receiving the post through their reader of choice, a habit many had written off as extinct but which showed up clearly in the logs. Federation too: the post propagating through ActivityPub, reaching people on Mastodon and other decentralised platforms who had never visited the site directly but were reading through their own network. Indirect human readers, choosing deliberately, just through different pipes. Then the other kind of traffic. A vulnerability scanner from Singapore, arriving six minutes into the spike, probing paths that do not exist on Ghost, looking for webshells planted on other servers at other times. A residential proxy swarm with a 96 percent bounce rate and nine seconds average visit duration, arriving and leaving without reading anything. Five categories of arrival, all at once, all labelled. 📒 ****AI crawler vs AI user** An AI crawler is an automated system that fetches your content to train a model or build a retrieval index. It reads your page so that, later, a person using an AI assistant might benefit from what you wrote. An AI user is a person interacting with a model in real time, where the model retrieves your content to answer a question. The crawler visits your site. The AI user may never come near it, but still encounters your writing. ## What seeing changes The RSS readers and the federated followers were the most useful surprise. They represent deliberate readership: people who opted in, who receive content through infrastructure they control, who are in some ways a more committed audience than someone who clicked a social link and stayed for thirty seconds. They were always there. I just had not looked. The AI crawlers are also readers, of a kind. They do not render pages. They pull content and pass it on to systems that answer questions on behalf of people. Optimising for them means thinking about caching and headers and response consistency. It is a different kind of publishing, running in parallel with the one I was already doing. The scanner was useful in a different way. Once it had a name and a fingerprint, it was manageable. A firewall rule catches it before it reaches the Worker. Named things can be handled. Anonymous noise cannot. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/05/Cloudflare-Block.jpg) Cloudflare security rules no block certain probing attempts. It feels good, claiming sone agency back. ## What I would encourage You do not need a custom Cloudflare setup to start seeing more clearly. But it is worth asking what your current analytics are hiding. RSS still works. Federation works. If you publish without offering a feed or an ActivityPub endpoint, you are not reachable by a meaningful slice of deliberate readers who prefer to control their own reading environment. Your content is already being read by AI systems, whether you have prepared for that or not. How it is cached, how consistently it is served, what headers it carries: these are not performance details. They are decisions about how your writing moves through the world. And the gap between “I have analytics” and “I understand my audience” is wider than it looks. The analytics show you what your platform was designed to measure. The actual audience is larger, stranger, and more varied than that. I thought I was optimising for speed. I ended up learning who I was writing for. ### Further reading - [My Visitors Are Not All Human. That Is Fine.](https://rob-hoeijmakers.ghost.io/my-visitors-are-not-all-human-that-is-fine/) - [Thirty Years of Caching, Sorted in an Afternoon](https://rob-hoeijmakers.ghost.io/thirty-years-of-caching-sorted-in-an-afternoon/) - [When Bots Become Readers: Publishing in the Age of AI Crawlers](https://rob-hoeijmakers.ghost.io/when-bots-become-readers/) ### Thirty Years of Caching, Sorted in an Afternoon URL: https://rob-hoeijmakers.ghost.io/thirty-years-of-caching-sorted-in-an-afternoon/ Last updated: 2026-05-10T08:53:30.000Z I have a jar of screws on my workbench. For years, I would fish through it looking for the right size, usually not finding it. Last week I sorted them, by type, by thread, by length. I used ChatGPT to help: photographed a handful, asked what I was looking at, got the taxonomy straight. Once I could name them, I could organise them. You can only sort what you understand. HTTP caching was my jar of screws. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/05/IMG_5301.jpeg) ## Thirty years of fog I have been building for the web since the early nineties. Caching was always there, somewhere in the background, doing something. I knew enough to be aware of it, not enough to actually control it. Cache-Control headers, TTL values, edge behaviour, the difference between what a CDN caches and what a browser holds, what gets invalidated when and why. Every time I approached it seriously, I ran into a wall of context I did not quite have. The documentation exists. The concepts are not secret. But caching is one of those domains where the gap between understanding the vocabulary and being able to apply it correctly is surprisingly wide. I would read, nod, implement something plausible, and move on with lingering doubt. This year, working with Claude, that changed. The parallel is closer than it sounds. I had the pieces in front of me for years. What I was missing was someone to explain what I was looking at. ## New instruments We went through the whole thing together. What my Cloudflare Workers were actually doing. What headers were being sent and why. What a browser would cache versus what the edge would cache. Where the inconsistencies were. What a coherent strategy would look like for a site like mine: a moderate personal blog with a global readership, running on Ghost, served through Cloudflare. It took an afternoon. Not because the subject got simpler, but because I had, for the first time, an instrument that could hold the full complexity with me. Ask a question, get an answer calibrated to my exact setup, follow a thread, revise, implement, check. The back-and-forth that used to require either a specialist or weeks of trial and error compressed into something manageable. The result was a caching strategy I actually understand. Headers that mean what I intend. Edge behaviour that is consistent. Rules I can read back and explain. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/05/Cache-stats.jpg) ## The audience had already changed The reason it finally felt urgent was not vanity metrics or pagespeed scores. It was a shift in who was actually reading. Human visitors are still there. But a [growing share of traffic to a site like mine now comes from crawlers](https://rob-hoeijmakers.ghost.io/my-visitors-are-not-all-human-that-is-fine/): search indexers, AI training pipelines, retrieval systems that serve content to agents rather than browsers. These systems do not render pages. They do not wait for JavaScript. They send a request, receive a response, and move on. For them, caching is not a convenience. It is the primary mechanism that determines cost, latency, and reliability of access. If you care about how your content moves through the world now, including through AI systems, you have to care about caching. Not as a performance optimisation for human browsers, but as infrastructure for machine readership. That reframing changed what I was optimising for. HTML cached at the edge, globally, with consistent headers and predictable expiry. Not because I expect a person in Singapore to shave 200ms off their pageload, but because the next request for that page is more likely to come from a retrieval system than a browser, and the request after that, and the one after that. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/05/Tiered-cache-Cloudflare-for-Ghost-blog.png) Cloudflare tiered caching on top of Ghost blog. ## What the tools made possible The caching itself is not new. The concepts are decades old. What changed is that I could finally see the system clearly enough to shape it. With the right instrument, a domain that had been opaque for thirty years became workable in a single session. That is not a small thing. There are other jars of screws on the workbench. 🗒️ The caching strategy described here was implemented using Cloudflare Workers, Cache Rules, and D1 for request logging. A [public cache-stats dashboard](https://rob-hoeijmakers.ghost.io/cache-stats) shows the current breakdown of traffic by type: human, AI crawler, SEO crawler, and unknown. ### Further reading - [My Visitors Are Not All Human. That Is Fine.](https://rob-hoeijmakers.ghost.io/my-visitors-are-not-all-human-that-is-fine/) - [When Bots Become Readers: Publishing in the Age of AI Crawlers](https://rob-hoeijmakers.ghost.io/when-bots-become-readers/) - [The End of Google Search (as we know it)](https://rob-hoeijmakers.ghost.io/the-end-of-google-search-as-we-know-it/) ### My Visitors Are Not All Human. That Is Fine. URL: https://rob-hoeijmakers.ghost.io/my-visitors-are-not-all-human-that-is-fine/ Last updated: 2026-06-05T13:52:11.000Z My site had a busy week. My analytics tool showed good numbers. Then I looked at the raw traffic logs and started counting visitors that weren't people. Not a problem. A discovery. For the past few months I've been running a custom traffic layer on this site, built on top of Cloudflare, which sits between my visitors and my server and sees everything that arrives. I wrote a small classifier that sorts incoming traffic into categories: human readers, crawlers from search engines, crawlers from AI companies, monitoring tools, and a catch-all for things I can't identify. On any given day, **humans are a minority**. That sentence used to feel alarming when I first framed it to myself. Now it just feels accurate. If you're curious, the live dashboard is public: [hoeijmakers.net/bot-stats](https://rob-hoeijmakers.ghost.io/bot-stats). ## The old mental model The web was built around a simple distinction. Humans used browsers. Bots sent raw HTTP requests, usually announced themselves honestly, and showed up cleanly in your server logs with a recognisable name. That distinction is largely gone. Modern automation often runs a real browser, the same engine that Chrome uses, and behaves accordingly: it loads the page, executes the code that runs on it, triggers the analytics beacon that tells my dashboard someone visited, scrolls, waits, and moves on. Add a residential proxy network, which routes traffic through actual household internet connections rather than servers in a datacentre, and the request arriving at my site looks indistinguishable from a person in Rotterdam reading an article over their home wifi. A visit is no longer proof of a visitor. ## Borrowed identities The residential proxy industry routes automated traffic through pools of real consumer internet connections. Millions of them. A client pays to use them; the destination website sees a local address, an ordinary internet provider, nothing suspicious. It sees, in other words, a person. This infrastructure serves legitimate purposes: checking whether an ad actually appeared in a specific country, testing how a website looks from different locations, monitoring competitor prices. It also serves extractive ones. The point is that it is now a normal part of web traffic, running quietly beneath the surface of every publisher's analytics. 🗒️ ****Residential proxies** route automated traffic through real household internet connections, making requests appear local and human. The same infrastructure is used for ad verification, AI training data collection, competitive intelligence, and sometimes fraud. ## What I actually see My traffic dashboard shows patterns rather than proof. No single signal identifies a visitor as a script rather than a person. Combinations do. Sessions that load a page and vanish, with no onward movement, no return visit, no sign of reading. Visitors arriving from a wide spread of countries with identical behaviour. Entry pages that cluster unnaturally: always the homepage, never a deep link from a real referrer. Requests arriving at suspiciously regular intervals. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/04/Residential-proxy-networks.jpg) This peak in the analytics is a Residential Proxy Network. Some of these are browser based and fire off the Javascript beacon. Some of this traffic is fine. Search crawlers indexing my writing are doing something useful. Crawlers from AI companies reading my articles are, in some sense, a new kind of audience. I've added a small index file to the site specifically to help them navigate it more cleanly. Some of it I decided not to serve. Amazon's infrastructure was generating substantial crawl traffic with no plausible return. I blocked it. Not in outrage: as a quiet editorial decision. My content, my choice about who reads it. That act of blocking felt unexpectedly good. Not because of what it prevented, but because of what it clarified. I have a position on this. I can act on it. ## The logfile returns There is a generation of web publishers who learned to read raw server logs before dashboards existed. They knew their traffic intimately. Dashboard analytics, when it arrived, abstracted most of that away. Traffic became a number, then a graph, then a metric. Infrastructure like Cloudflare is pulling some of that texture back. It sits in front of the server and sees things browser-based analytics cannot: how fast requests arrive, whether IP addresses rotate, whether a visitor ever fetches images or just the HTML. The signal is different from what a dashboard shows. I am not claiming logs are more truthful than dashboards. I am noting that they are different, and that the difference is currently interesting. [Moving past bots vs. humansAs AI assistants and privacy proxies challenge the capabilities of traditional bot detection, the Web needs new models for accountability. We believe that control should remain with the client, and that an open ecosystem of anonymous credentials is key to preserving user privacy while protecting origins from abuse.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon-32x32-14ef936f-ad9a-4058-b1ae-4d86a139fce3.png)The Cloudflare BlogThibault Meunier![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/Moving_past_bots_vs._humans-OG-e25d5312-627f-4b53-b8fc-e6980d315591.png)](https://blog.cloudflare.com/past-bots-and-humans/) ## What audience means now The web is no longer a medium where humans visit websites. It is a medium where humans, search crawlers, AI crawlers, monitoring scripts, and commercial intelligence systems all arrive through the same door. Some are reading. Some are indexing. Some are training models on what they find. Some are checking whether an ad appeared. Some I cannot classify at all. My analytics dashboard shows me one version of this. My traffic logs show me another. The honest position is that both are partial. What I have, with a little tooling and some curiosity, is a better picture than I had six months ago. And the ability to make choices about it. That is enough. ### Further reading - [Flying Blind: Measuring Traffic When Your Readers Are Machines](https://rob-hoeijmakers.ghost.io/measuring-traffic-machines-bots/) - [When Bots Become Readers: Publishing in the Age of AI Crawlers](https://rob-hoeijmakers.ghost.io/when-bots-become-readers/) ### When the Platform Becomes the Policy URL: https://rob-hoeijmakers.ghost.io/when-the-platform-becomes-the-policy/ Last updated: 2026-05-10T08:53:31.000Z The Dutch state secretary for Digital Economy and Sovereignty now communicates with a small group of civil servants via a European messaging app. Which one? Her spokesperson won't say. The initiative is, as they put it, "still in its infancy." But the announcement itself is what matters. The Dutch government has concluded that its dependency on WhatsApp and Signal is a risk worth managing. That is late, in European terms. France has been running Tchap for government use since 2019\. The Bundeswehr has had its own messaging infrastructure for years. Belgium launched Beam for 750,000 civil servants and military personnel last month. But the direction is now consistent across EU member states, and the reasons have shifted. ## Beyond Privacy This is no longer primarily about privacy ideology. It is about geopolitics. With the US increasingly willing to use technology as a lever of power, the question "can we access our own data tomorrow?" has moved from paranoid to prudent. A January survey by SWG found that 59% of Europeans already consider a US block on digital services a realistic scenario; 86% consider it at least plausible. Governments are drawing the obvious conclusions. What is less obvious is that the same logic applies to businesses, just more quietly. ## The AI Access Dispute While governments look for messaging alternatives, the European Commission is fighting a separate but related battle. It has now sent Meta a second "charge sheet" over the fees Meta wants to impose on external AI providers, including OpenAI and Anthropic, for access to the WhatsApp Business platform. The sequence is telling. Meta first banned third-party AI providers outright. When the EU challenged that, Meta introduced a pricing model: 5 to 13 cents per message for external AI chatbots. Meta AI, its own assistant, pays nothing. The Commission's position is that this is the same outcome in different legal clothing: a structure designed to foreclose competition in favour of Meta's own AI product. For most businesses running a WhatsApp chatbot today, the direct impact is limited. What changes if Meta prevails is the competitive landscape for AI assistants at scale. If a platform can effectively exclude rival AI providers, the choice of assistant is no longer yours to make. ⚖️ The EC charge sheet covers Meta's terms for external AI providers on WhatsApp Business, not the costs businesses already pay via BSPs like 360 Dialog or Twilio. Those conversation fees are a separate layer. The regulatory dispute is upstream, about who gets platform access at all. ## Platform Dependency as Risk WhatsApp reaches over 85% of the Dutch population. For businesses, that makes it close to mandatory infrastructure. The problem with mandatory infrastructure is that you stop thinking about it as a choice. A pricing change, a policy update, an account suspension: these arrive without negotiation. What the Dutch government is framing as a sovereignty question is, for businesses, a more familiar concept. It is supply chain risk, applied to communication infrastructure. When you build your entire customer contact layer on a single platform, you have created a single point of failure you did not consciously design. The government pilot is small and tentative. But it reflects something that has been accumulating for a while: the recognition that neutral infrastructure is a fiction. WhatsApp is a platform with commercial interests, regulatory entanglements, geopolitical exposure, and an AI strategy that is now under formal legal challenge. Those are not abstract conditions. They are conditions that shape what you can build on top of it. ### Further reading - [Why WhatsApp Channels Trigger the Digital Services Act](https://rob-hoeijmakers.ghost.io/why-whatsapp-channels-trigger-the-digital-services-act/) - [The EuroStack explained](https://rob-hoeijmakers.ghost.io/the-eurostack-explained/) - [From Shopping Cart to Server Rack: How Lidl Builds a European Cloud](https://rob-hoeijmakers.ghost.io/lidl-stackit-european-cloud/) ### GPT Image 2 Changed My Mind on AI Visuals URL: https://rob-hoeijmakers.ghost.io/gpt-image-2-changed-my-mind-on-ai-visuals/ Last updated: 2026-05-10T08:53:31.000Z I had, without quite deciding to, written ChatGPT off for images. The earlier output felt like a tool optimised for fantasy illustrations, product mockups, and memes. Useful for some things. Not for the way I work. My interest in visualisation has always been functional: diagrams, schematics, process maps. Images that explain something rather than represent something. For that category, AI image generation had consistently disappointed. Then last week I needed to understand how a washing machine actually works internally. The inner drum, the outer tub, the bearings, the door seal, where the shaft exits the housing. Text descriptions get you some of the way. A diagram would get you further. So I asked ChatGPT Images 2.0, running on the new GPT Image 2 model, to generate one. What came back made me revise my assessment. ## What arrived The output was not a pretty picture of a washing machine. It was a coherent technical infographic: multiple views of the same machine, labelled components, readable typography, front and rear logic, the drive belt and motor in correct relation to each other. It carried layout hierarchy. It had editorial polish. It looked less like image generation and more like a service manual that had not yet been published. The surprise was specific. A washing machine is a good test precisely because it has no aesthetics to fall back on. The image is either mechanically correct and useful, or it is not. But the deeper surprise was repeatability. This was not a lucky render. The model had reasoned about what a washing machine actually is before generating anything, and that showed in every label, every cross-section, every spatial relationship. You could ask again and get something coherent. That is new. ## Why this category matters The images I care about are not decorative. Every organisation explains something: how a process works, how a system fails, what happens in a given sequence. For years, turning that complexity into a clear visual required time, a specialist, and budget. The bottleneck was not the knowledge, it was the production. GPT Image 2 is built around the idea that the model thinks before it draws, reasoning about what an explanation actually requires before generating anything. Text inside images, historically the biggest weakness of these models, has improved substantially: labels, signage, UI elements, correct spelling and consistent spacing. The result is a model that does not just render, it plans. And because it understands the subject before it draws, the output is repeatable: ask the same question twice and you get something coherent both times, not two unrelated lucky guesses. That is the shift. Not better art. Better explanation, on demand.​​​​​​​​​​​​​​​​ 💡 GPT Image 2 launched on April 21, 2026\. It introduced thinking mode, web search during generation, and the ability to produce up to eight consistent images from a single prompt. Available to all ChatGPT users; extended outputs require a paid plan. ## The category I had dismissed Many people still assess AI image tools by the use cases that dominated the first wave: avatars, memes, synthetic photography, visual gimmicks. That assessment is not wrong, it is just stale. The more consequential category is knowledge visuals: images that help a professional understand something faster than prose alone. Technical diagrams, process flows, onboarding schematics, infrastructure maps. The kind of thing that organisations need constantly and produce slowly. If that gap is now measurably shorter, the operational implication is significant. Not because designers are replaced, but because explanation becomes cheap enough to happen more often. I only wanted to know how my washing machine works. The diagram that came back raised a different question: what else is now explainable in minutes that previously required a budget line? ### Further reading - [From text to diagrams: working with Mermaid](https://rob-hoeijmakers.ghost.io/from-text-to-diagrams-working-with-mermaid/) - [From Ideas to Precision: Why I Use Mermaid with LLMs](https://rob-hoeijmakers.ghost.io/mermaid/) ### Dressing up an AI Model in a Harness URL: https://rob-hoeijmakers.ghost.io/harness-ai-model/ Last updated: 2026-05-10T08:53:31.000Z A friend who is an engineer used the word harness. We were talking about the infrastructure around a model: the tools it can call, the memory it can access, the context it receives. The harness, he said, is what makes the model function as an agent. The word landed wrong. ### Restraint A harness restrains. It channels force that would otherwise go somewhere you don't want it. The equestrian harness, the safety harness, the climbing harness: all of them are about control, about keeping something contained and directed. There is an adversarial assumption built into the metaphor, a relationship between the wearer and the harness that is fundamentally one of managed risk. That is not what the infrastructure around an AI model does. It equips. It gives the model access to tools, continuity across sessions, a defined role within a specific context. The harness, in this sense, is less like a restraint and more like a workbench: the thing that makes work possible. You would not say a surgeon's instruments harness the surgeon. ⚔️ In AI development, a harness refers to the surrounding infrastructure that makes a model operational in a specific context: the tools it can call, the memory and session state it can access, the APIs it connects to, and the logic that governs how it responds. The harness is what turns a raw model into a working agent. ### Standard Language [Ethan Mollick](https://www.oneusefulthing.org) uses the same term in his newsletter, framing AI as three interlinked concepts: models, apps, and harnesses. His harnesses are the tools an AI can use and how the model is hooked up to them. The definition is functional and clear, and it reflects how the term is already settling into common use among practitioners. Which makes it worth pausing on. The words that become standard in a new field carry assumptions forward. "Harness" inherits from a tradition where powerful things need to be held. "Infrastructure" is more neutral. "Toolkit" points in a different direction again. None of them are wrong exactly, but they weight the thinking differently. [Sign of the future: GPT-5.5One impressive step on the curve![](https://static.ghost.org/v5.0.0/images/link-icon.svg)One Useful ThingEthan Mollick![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/https-3A-2F-2Fsubstack-post-media.s3.amazonaws.com-2Fpublic-2Fimages-2F2c591570-ee4a-4107-a9d1-b167d3f4a64a_1484x1060.png)](https://www.oneusefulthing.org/p/sign-of-the-future-gpt-55) ### The Wrong Question The distinction matters because the word shapes the thinking. If the surrounding infrastructure is a harness, the model is something to be tamed, a force that needs to be channelled before it can be trusted. That framing puts the engineering problem in the wrong place. The question becomes: how do we constrain it? When the more interesting question is: how do we equip it? My friend's use of the word was entirely competent. He knows what the infrastructure does. But the equestrian image kept surfacing: the horse, strong and fast, made useful by what is strapped around it. It is a working metaphor, in the sense that it functions. It just describes a different relationship than the one I want to be building toward. The model is not a horse. The infrastructure is not a harness. We are still looking for the right words. ### Further reading - [Can Today’s AI Agents Survive Their Own Runtime?](https://cobusgreyling.medium.com/can-todays-ai-agents-survive-their-own-runtime-ec6aaac7109a) - [When AI Moves Into Your Working Environment](https://rob-hoeijmakers.ghost.io/when-ai-moves-into-your-working-environment/) - [Word of the year: model](https://rob-hoeijmakers.ghost.io/model-meaning-etymology/) - [What coding with AI feels like now](https://rob-hoeijmakers.ghost.io/what-coding-with-ai-feels-like-now/) ### Built in, not bolted on URL: https://rob-hoeijmakers.ghost.io/built-in-not-bolted-on/ Last updated: 2026-05-10T08:53:31.000Z A chatbot surfaces above an old document repository. A copilot attaches itself to fragmented data. A smart search layer sits on top of years of accumulated disorder. The AI works, up to a point. Then it runs into the same wall every time: the information underneath it. ### The problem underneath When people talk about AI disappointing them, they usually mean the model. The output is generic, the answers are wrong, the assistant keeps missing context. But the model is often not the issue. What it has to work with is. Most organisational knowledge still lives in SharePoint folders with inconsistent naming, email threads where decisions disappear, documents that exist in four versions with no clear indication of which is current. Add a copilot to that environment and you get a faster, more articulate reflection of the same confusion. The AI amplifies what is already there. If the structure is weak, the results are weak. ### Architecture first Built-in AI assumes that information has structure, that documents carry provenance, that workflows are legible enough for a system to participate in them. Not a large IT replacement project. A different starting point. Knowledge work organisations that are getting real results from AI tend to share one thing: they treated the information architecture as part of the AI investment, not a precondition someone else would handle. That means version control that is actually used, consistent tagging, clear ownership of documents, workflows that do not rely on institutional memory to navigate. A well-structured knowledge base is useful long before any AI touches it. The AI just makes the gap visible faster. ### The compliance accelerant In regulated sectors, the pressure is sharper. When an adviser needs to reconstruct a recommendation from three years ago, or a regulator asks which version of a policy was in force on a given date, the question is not whether the AI can answer it. The question is whether the information environment can support the answer at all. AI governance and information governance are converging. Organisations that have separated the two, running AI initiatives on top of legacy information infrastructure, will encounter this collision eventually. 💡 WiseWare structures organisational knowledge into governed memory objects: decisions, policies, commitments, evidence. Source-backed, human-reviewed, auditable. Built in Amsterdam. wiseware.nl ### The question that shifts For a while, the distinguishing question was "do you have AI?" Now it is closer to "is your organisation built to work with it?" A copilot on top of a well-structured environment is a genuine productivity multiplier. The same copilot on top of a fragmented one is a better interface to the same old problem. The architecture was always the work. AI just made the cost of skipping it more visible. ### Further reading - [When Git Grows Up](https://rob-hoeijmakers.ghost.io/when-git-grows-up/) - [When AI Moves Into Your Working Environment](https://rob-hoeijmakers.ghost.io/when-ai-moves-into-your-working-environment/) - [The Workspace You Already Have](https://rob-hoeijmakers.ghost.io/the-workspace-you-already-have/) - [AI literacy: from definition to practice](https://rob-hoeijmakers.ghost.io/ai-literacy-from-definition-to-practice/) ### The Workspace You Already Have URL: https://rob-hoeijmakers.ghost.io/the-workspace-you-already-have/ Last updated: 2026-05-10T08:53:31.000Z My daily setup involves three AI systems, a shared drive, and a folder structure that only makes sense to me. It works. It also wouldn't survive a compliance audit, a colleague handoff, or a question about where the data goes. ### The Informal Layer Most knowledge workers operating with AI have built something like this without naming it. Conversations in Claude for research, Gemini inside Google Workspace for documents, ChatGPT for a second opinion. Files land in Drive or on a local disk. Context lives in chat history, or in memory, or in neither. The result is a personal productivity layer that is genuinely powerful and entirely ungoverned. The document management industry has spent decades building the governed version: structured repositories, metadata, audit trails, version control, retention policies. SharePoint. OpenText. M-Files. The category is called DMS, document management system, and the market is around ten billion euros and growing fast. What it has never managed to build is the experience of actually thinking with your documents, of asking a question and getting a useful answer rather than a search result. ### Two Directions AI is now arriving from both sides. The DMS vendors are bolting intelligence onto their repositories: auto-classification, semantic search, generative summaries. The AI vendors are adding structure to their chat interfaces: folders, projects, persistent memory, file search. Both are converging on the same space, the AI-native knowledge workspace, from opposite ends. The difference is where they start. Legacy DMS starts from governance and adds intelligence. AI-native tools start from intelligence and add governance. That ordering matters more than it sounds. A system designed around compliance and filing tends to feel like compliance and filing, even after AI is layered on. A system designed around thinking and conversation tends to feel like that, even after folders and access controls are added. ### The European Gap There is a third variable that rarely features in the US-centric DMS conversation: where the data lives and who controls the models. For European practitioners and organisations, working with AI means trusting that client conversations, internal documents, and strategic thinking are not being used to train someone else's model in another jurisdiction. The major platforms offer reassurances, but the architecture is American, the models are American, and the terms of service reflect that. 🇪🇺 The EU AI Act and GDPR together create real obligations around data residency and model governance. For many European organisations, "where is my data" is not a compliance question. It is a strategic one. This is where the category gets interesting for smaller European practices. The improvised setup, Claude plus Gemini plus Drive, is functional but not sovereign. An EU-hosted workspace that supports multiple models, keeps data in Europe, and organises knowledge the way teams actually work is a different proposition. Not a DMS in the legacy sense, and not just a chat interface. Something in between, and perhaps more useful than either. 🇳🇱 [WiseWare.chat](https://wiseware.nl) is one of the first products positioning explicitly in this space. Built in Amsterdam, EU-hosted, the product structures company knowledge into governed memory objects: decisions, policies, commitments, evidence. Source-backed, human-reviewed, auditable. Early product, coherent direction. One to watch. What that category ends up being called is still open. The tools are arriving before the language has. ### Further reading - [Built in, not bolted on](https://rob-hoeijmakers.ghost.io/built-in-not-bolted-on/) - [When Git Grows Up](https://rob-hoeijmakers.ghost.io/when-git-grows-up/) ### When Git Grows Up URL: https://rob-hoeijmakers.ghost.io/when-git-grows-up/ Last updated: 2026-05-10T08:53:32.000Z The problem with git was never git. It was the assumption that everyone working with content could think like a developer. Version control is a sound idea for knowledge. Track changes, branch, merge, roll back. For organisations managing legal, financial, or medical content at scale, that discipline is exactly what is needed. The tooling just never matched the people. Authors are not engineers. They do not want to commit, push, or resolve merge conflicts. They want to write, review, and publish with confidence that nothing disappears. ### The interface problem What changes when you add agents to a git-based knowledge system is the interface between the system and the person using it. Branching becomes a guided act rather than a technical one. An agent can explain what a branch means in the context of a document review, prompt the right action, and flag when something looks inconsistent with the rest of the knowledge base. Editorial standards are where this becomes interesting. In most content organisations, style guides and governance rules live in a wiki page or shared drive folder that someone remembers to check. Complying with them is a personal responsibility, which means compliance is uneven. When an agent carries those standards into the editing environment itself, the guidance becomes active. It surfaces at the moment it is relevant, not when someone remembers to look it up. ### Governance as the use case This matters more in some sectors than others. For organisations where content carries legal weight, where a document version can be cited in a court or audit, the distance between "we have a process" and "the process is enforced" is significant. Version control closes part of that gap. Active editorial guidance closes more of it. The clients who need this are not looking for a productivity tool. They are managing risk. The question they are asking is not whether their teams can write faster, but whether the organisation can demonstrate that what was published, when, and under what review, is traceable and defensible. ### European footing For organisations in regulated European industries, where content is held, and by whom, is not a neutral question. A knowledge platform with European hosting and ISO 27001 certification in progress is not a marketing differentiator. It is a procurement requirement waiting to be formalised. The certification path matters here in a specific way. For public sector and governance-heavy clients, ISO 27001 is the entry point to a longer chain: certification enables insurance, insurance enables enterprise contracts, enterprise contracts carry SLA penalties that in turn require the certification to hold. Once you commit to this path, you commit to maintaining it. The liability is real and personal. That is precisely why most smaller technology companies avoid it, and why those that do pursue it can credibly operate at a different scale. The combination of git-based versioning, agent-assisted authoring, and European infrastructure with serious certification underway is a specific answer to a specific problem. It is not the most obvious stack. But for the organisations that need it, it fits in a way that general-purpose tools do not. 💡 [WiseWare](https://wiseware.nl) is a platform for AI-assisted knowledge management at scale, with European hosting and ISO 27001 certification in progress. [Schmuki](https://schmuki.nl) resells and co-delivers WiseWare for governance-heavy organisations in legal, financial, and medical sectors. ### Further reading - [Built in, not bolted on](https://rob-hoeijmakers.ghost.io/built-in-not-bolted-on/) - [The Workspace You Already Have](https://rob-hoeijmakers.ghost.io/the-workspace-you-already-have/) ### Delegating Past Your Own Ceiling URL: https://rob-hoeijmakers.ghost.io/delegating-past-your-own-ceiling/ Last updated: 2026-05-10T08:53:32.000Z This one goes a level deeper than usual. If you run a self-hosted blog or manage your own web infrastructure, stay with it. The payoff is real. Cloudflare, the network and security layer that sits in front of most of this blog, has been part of my setup since the beginning. DNS, CDN, basic security. The layer in front of Ghost, my publishing platform, that I configured once and mostly left alone. Not because there was nothing more to do, but because the gap between what Cloudflare can do and what I could confidently operate was wide enough to leave alone. That gap closed recently. Not because Cloudflare got simpler, but because I stopped being the one operating it. ## The ceiling Every tool has a capability ceiling for any given user. For most practitioners running a Ghost site, Cloudflare's ceiling sits somewhere around DNS and caching. The dashboard is capable but not intuitive. Workers, R2, analytics at the edge: these are real features with real value, but they require a mental model that takes time to build and maintain. Most people stay at the surface. I did too. What Claude Code changed is where that ceiling sits. I can now ask it to read my zone configuration, deploy a Worker, or update a content file, and it does. Not because I learned Cloudflare more deeply, but because I delegated past the point where I would normally stop. ## What delegation made possible I asked Claude Code to generate an [llms-full.txt](https://rob-hoeijmakers.ghost.io/llms-full.txt) from my RSS feed, a complete text inventory for AI systems. It did. Then I asked whether it could produce a Markdown version of every page, served directly from the network layer without touching Ghost. It could. It wrote a Cloudflare Worker, a small script that runs at the network edge and intercepts requests before they reach the server. That Worker intercepts any request ending in `.md`, fetches the page content, converts it to Markdown, and returns that instead. Ghost never sees the request. Each page now also carries a `rel="alternate"` link in its HTML head, so machines that know to look will find the alternate format automatically. Example >> [https://hoeijmakers.net/delegating-past-your-own-ceiling.md](https://rob-hoeijmakers.ghost.io/delegating-past-your-own-ceiling.md) The Worker code was readable. I could follow it. I would not have written it myself, and I would not have attempted the feature without Claude Code as operator. That's the point: the outcome wasn't just the same task with less friction. It was a capability I didn't have before, sitting just above the ceiling I previously accepted. ## Ghost and Cloudflare, divided cleanly Ghost owns the content and the publishing workflow. Cloudflare owns the delivery layer and everything Ghost doesn't cover. Claude Code connects them. None of these systems needed to know about the others. The Worker intercepts requests and responds differently based on the URL. Ghost keeps doing what it does. This division feels stable. It also suggests something worth sitting with: the limiting factor in most small-scale web infrastructure isn't the tools. The tools are capable. The limiting factor is the operator's ceiling, the point where complexity exceeds confidence and you stop reaching for what's there. Delegation shifts that line. Not permanently, not universally, but enough to matter. ### Further reading - [Guests That Should Behave](https://rob-hoeijmakers.ghost.io/guests-that-should-behave/) - [What llms.txt can do for your website](https://rob-hoeijmakers.ghost.io/llms-txt/) - [When Bots Become Readers: Publishing in the Age of AI Crawlers](https://rob-hoeijmakers.ghost.io/when-bots-become-readers/) ### Guests That Should Behave URL: https://rob-hoeijmakers.ghost.io/guests-that-should-behave/ Last updated: 2026-05-10T08:53:32.000Z The traffic spikes in [Plausible](https://rob-hoeijmakers.ghost.io/cookie-free-analytics/) (Web analytics) made no sense. Peak after peak, no referral source, no pattern I recognised. Bots, clearly, but the kind that arrive carrying a real browser, behaving like a human long enough to slip past lightweight analytics. Not a security incident. More like guests who don't knock. That framing stuck with me as I worked through the fix. Bots are guests. Most of them are welcome. The question is which ones, and on what terms. ## Welcome and unwanted The web has always had crawlers. Search engines, archivers, feed readers: automated visitors that make the open web function. I have no objection to those. What changed over the past year or two is volume and intent. By mid-2025, crawling for AI model training accounted for nearly 80% of all AI bot activity on Cloudflare's network. Many of those crawlers identify themselves honestly. Some don't, cycling through residential IP addresses and real browsers to blend in. The ones showing up in my Plausible dashboard were the second kind. Uninvited guests who arrive in disguise, consume resources, and distort the picture. Cloudflare's bot protection handles exactly this: using signature matching, heuristics, machine learning, and behavioural analysis to distinguish the two, while maintaining a Verified Bot directory for legitimate crawlers like Googlebot that are allowed through. I enabled the AI scrapers toggle. The spikes stopped. ## The Workers detour Getting this working on a [Ghost](https://rob-hoeijmakers.ghost.io/ghost-blog/) site took a step I hadn't anticipated. Cloudflare is not a simple product for an individual publisher. There are layers, and the terminology assumes familiarity with infrastructure concepts most bloggers don't carry day to day. I worked through the configuration with Claude, which made the complexity navigable. A few exchanges, some trial and error, and it was done. The interesting byproduct came when Cloudflare suggested tightening up robots.txt. That small nudge opened a second door. I'd been serving llms.txt as a Ghost page, which worked, but felt like a workaround. With a Cloudflare Worker already in place, I wrote a proper script to serve it as a real txt file and referenced it from robots.txt. A hack became infrastructure. 📎 [Llms.txt](https://rob-hoeijmakers.ghost.io/llms-txt/) is an emerging convention for telling AI systems what to pay attention to on a site. Where robots.txt sets access rules, llms.txt provides semantic guidance: what's here, what matters, how to use it. Think of it as a house guide left out for guests who will never ring the bell. ## Hospitality, restated The distinction that matters is not bots versus humans. It's guests who identify themselves and follow the rules versus those who arrive in disguise and behave badly. The first kind are welcome. The second kind get turned away at the edge, before they reach the door. llms.txt is the other side of that same logic. For the guests who do behave, who are reading the site to retrieve, to understand, to cite: here is what I've made, here is how to find it, here is what it's for. Blocking the bad actors and signalling clearly to the good ones. Two tools, one underlying idea. The web's implicit contract, that content is public and humans read it, is being rewritten. I'm fine with that. I just want the new guests to knock first. ### Further reading - [What llms.txt can do for your website](https://rob-hoeijmakers.ghost.io/llms-txt/) - [The End of Google Search (as we know it)](https://rob-hoeijmakers.ghost.io/the-end-of-google-search-as-we-know-it/) - [When Bots Become Readers: Publishing in the Age of AI Crawlers](https://rob-hoeijmakers.ghost.io/when-bots-become-readers/) ### Bildung and the Function Problem URL: https://rob-hoeijmakers.ghost.io/bildung-and-the-function-problem/ Last updated: 2026-05-10T08:53:32.000Z Wilhelm von Humboldt is one of those figures I knew only by association: the brother, the geographical namesake, the name on the university. The actual ideas, I had missed. A piece from the Cosmos Institute put Humboldt's concept of *Bildung* in front of me. The argument is simple and uncomfortable: education shaped by Humboldt was about forming a whole person, not training a function. The modern professional world, by contrast, has spent a century doing the opposite. We have been very good at turning people into functions. ## The Bildung idea *Bildung* is hard to translate, which is part of why it stayed in German. It means something like self-cultivation through engagement with the world: art, science, language, culture, other people. German has a second word for the other model: *Ausbildung*, vocational training, formation toward a function. The prefix *aus* signals completion, extraction, being shaped into a finished product. One word for becoming a person. One word for becoming a role. The goal of *Bildung* is not competence in a domain. The goal is a richer, more capacious person. Humboldt believed the university existed to develop that person, not to fill a role. This is not naive idealism. Humboldt was a state official. He designed the University of Berlin in 1810 with this philosophy embedded in the structure: research and teaching together, students as active participants, knowledge pursued for its own sake. The model spread across Europe and shaped what we call research universities. ## The function trap What the Cosmos Institute piece points at, and what I find genuinely interesting, is that AI is now exposing a tension that was always there. When knowledge work was scarce and complex, being a good function was valuable. You knew the law, the process, the domain. That knowledge was your identity, your leverage, your job security. AI is eroding the scarcity. Not all of it, and not evenly, but the trend is clear enough that the question becomes: what remains when the function is cheaper to automate than to hire? The Humboldt answer, if taken seriously, is that what remains is the person. The capacity to make judgements, to hold complexity, to bring a formed perspective to an ambiguous situation. None of those come from function-training. They come from exactly the kind of broad, messy, integrative engagement that *Bildung* describes and modern professional education systematically devalues. ## What I do not know I want to be honest about the limits of what I am saying here. I have not read Humboldt, I have encountered the idea. I am not a philosopher of education or a historian of the university. What I can say is that the frame feels generative: it names something I have been sensing without having the word for it. The risk of reaching for Humboldt now is that it becomes a nostalgic move, a way of dressing up the resistance to change in the language of classical education. The universities Humboldt influenced are not obviously doing *Bildung* today. They are credentialing machines with research departments attached. The more interesting question is whether the AI transition creates conditions where something like *Bildung* becomes practically useful again, not as philosophy, but as a competitive model for how people and organisations develop. Whether that is inspiration or a warning, I genuinely do not know. 📎 The Cosmos Institute piece that triggered this: "[You Are Not a Function](https://blog.cosmos-institute.org/p/you-are-not-a-function)" ### Further reading - [Vision, Judgement, Creativity: Reclaiming Agency in the Age of AI](https://rob-hoeijmakers.ghost.io/vision-judgement-creativity-reclaiming-agency-in-the-age-of-ai/) - [The Thermostat Effect: Why AI Feels Distant but Works Deep](https://rob-hoeijmakers.ghost.io/thermostat-effect-on-ai/) ### Claude Code Routines and the Access Problem URL: https://rob-hoeijmakers.ghost.io/claude-code-routines-and-the-access-problem/ Last updated: 2026-05-10T08:53:32.000Z ## The Step That Usually Blocks Setting up an automation that reads your calendar involves the same step almost every time: authentication. OAuth flow, credential storage, a token that needs refreshing, a connector that needs configuring. The task itself takes minutes. Getting access takes longer. I've written before about [the delegation problem of the internet](https://rob-hoeijmakers.ghost.io/the-delegation-problem-of-the-internet/): most digital systems assume one account, one person, one operator. Automation quietly breaks on that assumption. ## Routines Claude Code recently added **a scheduling layer called Routines**, currently in research preview. A routine is an automation you configure once, including a prompt, a repo, and your connectors, and then run on a schedule, from an API call, or in response to an event. As a quick test, I set it up to print my Google Calendar each weekday at 8:00 AM, the same recurring job I already run in n8n. It worked immediately. No connector setup, no OAuth dance, no credential to store. Claude already has access to my Google Calendar through the existing integration. The Routine just used it. That's the actual story. The scheduling feature is convenient. The inherited access is the payload. ## Familiar Ground n8n still does more. The workflow logic is deeper, the integrations more composable, the control more granular. For anything complex, that matters. But for tasks that already live inside the Claude environment, Routines removes the step that most often kills the momentum to automate something in the first place. There's a category of small, recurring jobs that never get automated because the setup cost exceeds the value: a daily briefing, a weekly summary, a reminder that pulls live data. The barrier is rarely the task itself. It's the plumbing. When the plumbing is already in place, the calculus shifts. ## Still Early There are limits to factor in. On Claude Pro, Routines allows 5 runs per rolling 24 hours, with additional runs drawing on Extra Usage. Max subscribers get 15\. For a single daily briefing, 5 is enough. For anything more frequent, the economics need checking first. My runs haven't visibly registered in the usage counter yet, so how the accounting actually works in practice is still unclear to me. Scheduled tasks also have a way of failing quietly, and reliability over weeks matters more than a clean first run. That's where dedicated tools earn their place. But the direction is legible. Authentication and access, the layer that blocks most automation for most people, are becoming inherited properties of the tools where the work already happens. The Routine ran. The interesting part was what I didn't have to do first. 💡 Routines is available for Claude Code users on Pro, Max, Team, and Enterprise plans. Pro: 5 runs/day. Max: 15 runs/day. Team and Enterprise: 25 runs/day. Additional runs draw on Extra Usage when enabled. ### Further reading - [The Delegation Problem of the Internet](https://rob-hoeijmakers.ghost.io/the-delegation-problem-of-the-internet/) - [When AI Moves Into Your Working Environment](https://rob-hoeijmakers.ghost.io/when-ai-moves-into-your-working-environment/) ### The Claude plan I actually use URL: https://rob-hoeijmakers.ghost.io/claude-plans-usage/ Last updated: 2026-06-07T20:18:03.000Z A friend asked which Claude subscription he should get. The pricing page left him more confused than when he started. I can't tell him what's right for his situation, but I can tell him what I do. ## The setup I run a Pro subscription. That covers the bulk of what I need: writing, analysis, research, structured thinking, light agentic work. I hit the daily usage limit occasionally. The weekly limit, rarely. When I do hit the daily limit, Claude switches to prepaid credit automatically. No interruption, no decision to make. I top up the balance now and then, and the buffer keeps working in the background. Total monthly cost stays well below the Max plans, because I only pay overage when I actually need it. That asymmetry matters. Max makes sense if you're hitting limits constantly or need guaranteed capacity. For the way I work, occasional overage is cheaper than a high fixed floor. ## Model choice Sonnet or Opus makes a real difference here too. Opus is the more capable model and costs significantly more per conversation. I use Sonnet for most things, and it handles the bulk of what I do well enough. Opus comes out when the problem genuinely requires it. The instinct to always reach for the most powerful model is understandable. It's also where costs quietly compound. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/04/Claude-Plan-usage.png) An overview of how I use Claude and the subscription. ## The invisible meter The trickier part is that token usage isn't visible while you're working. It's a bit like bandwidth: you don't think about it until something trips. A conversation feels like a conversation, not like a metered resource ticking down. It gets more pronounced when you run things in parallel. Multiple agentic sessions, several long document conversations open at once: usage can spike in ways that aren't obvious until you've hit a limit or a bill arrives. The more autonomous the setup, the less usage feels connected to the work you're actually doing. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/04/Claude-Usage.jpg) There are probably tools and commands available to check usage more precise but for me the Claude usage page under settings is the dashboard. ## The API side I also keep a small prepaid credit balance for the API, which is separate from the claude.ai subscription. I use it for tools I've built, and occasionally for Claude Code in the terminal. The two balances don't talk to each other: a Pro subscription on claude.ai doesn't cover API usage, and API credit doesn't unlock Pro features. 💡 Claude is one of several AI tools I use daily. ChatGPT plays a different role in my workflow: more thinking and cognitive exploration, where Claude is more about doing and building. I wrote about that distinction here → "[ChatGPT thinks with you, Claude builds with you](https://hoeijmakers.net/chatgpt-thinks-with-you-claude-builds-with-you/)" ## Plans as a moving target My friend is on Max, partly because he's waiting for the Word plugin, and partly because he suspects that being in a plan early pays off as Anthropic adds features over time. That last point is probably right. Anthropic does adjust what's included in plans as products mature, and being an established subscriber tends to work in your favour. For him, Max makes sense. For me, Pro plus a small credit buffer is still the more cost-efficient fit. The right answer depends on how you work, not on which plan looks most complete on a pricing page. Start with Pro. See where the limits actually land for your patterns. Add a credit buffer so a daily ceiling doesn't stop you mid-afternoon. Reassess after a month. P.s. Update 24 April. This monthly limit is something you set yourself and it means a limit to the extra credit that is used. I misread and had it a 20 euro's when it jammed my progress. So I just had to adjust the limit. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/04/Monthly-limit-Claude.jpg) Adjust the limit for the Claude monthly spend. ### Further reading - [ChatGPT Thinks With You. Claude Builds With You.](https://rob-hoeijmakers.ghost.io/chatgpt-thinks-with-you-claude-builds-with-you/) ### The EU Age Verification App Was Designed to Be Distrusted URL: https://rob-hoeijmakers.ghost.io/the-eu-age-verification-app-was-designed-to-be-distrusted/ Last updated: 2026-05-10T08:53:33.000Z The announcement slide said "NO TRACEABILITY" in large capitals on a blue background ringed with gold stars. The Threads comment section replied: "How do we verify children, ah yes let's make big database of all people." The architecture is the opposite of a database. The communication guaranteed nobody would notice. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/04/EU-Age-verification-campaign.jpg) Left LinkedIn, right Threads. ## What they built The Age Verification App is a "mini wallet," built on the same technical foundation as the forthcoming [EU Digital Identity Wallet](https://rob-hoeijmakers.ghost.io/eu-eid-eidas-eudi-wallet/). When a platform requests age verification, the app generates a cryptographic proof derived from a credential issued by a national identity authority. The platform receives one binary answer: over 18, or not. No name, no date of birth, no address. The underlying identity data never leaves your device. The same mechanism works offline. At a festival gate, a bar, a tobacconist. You present your wallet, the venue's device gets a yes or no, and sees nothing else. Most adults have handed a passport to a bouncer to prove they are old enough to enter, giving a stranger full sight of their identity because the system had no finer instrument. This changes that. The credential is cryptographically bound to the issuing device. It cannot be copied or handed to someone else; the proof would fail. The weak point is onboarding: if someone sets up a wallet using another person's ID document, the credential reflects that person's age. A process problem, not a cryptographic one. 💡 The blueprint is open source at [ageverification.dev](https://ageverification.dev). Six member states including France, Spain, and Denmark are currently in pilot. A second version, released October 2025, adds passport and ID card onboarding alongside eID. ## Two failures, one root cause The Commission made two distinct mistakes, and they compound each other. The first is aesthetic. Blue backgrounds, gold stars, caps-lock claims. This is the design language of an institutional poster, activating exactly the distrust it was meant to address. Design communicates before words do. A system built on cryptographic privacy, dressed in the visual language of state surveillance, loses the argument before it starts. Estonia's e-Residency programme, adjacent identity infrastructure, answered three questions before anything else: what is it, who is it for, how do I get one. The second is structural. "The European Age Verification App is ready" is product language. It implies: find it, download it, use it. What actually launched is an open-source blueprint that member states can adapt, currently in pilot in six countries, dependent on national identity infrastructure that varies by country, with no App Store link, no rollout date, no answer to the question any citizen would ask: what do I actually do with this, and when? Both failures share a root. The Commission framed this entirely as a child protection measure. "It is for parents to raise their children. Not platforms." Every word is defensive. It excludes every adult who benefits from not uploading a passport scan to access a service, or not handing their full identity to a bouncer at a festival. The genuine upside, comfort, convenience, less data exposed in daily life, went unspoken. [It is for parents to raise their children. Not platforms. The European Age Verification App is ready. It will allow users to prove their age when accessing online platforms. Just like shops ask for… | European Commission | 48 commentsIt is for parents to raise their children. Not platforms. The European Age Verification App is ready. It will allow users to prove their age when accessing online platforms. Just like shops ask for proof of age for people buying alcoholic beverages. And it ticks all the boxes: ✅ Highest privacy standards in the world ✅ Works on any device ✅ Easy to use ✅ Fully open source More info: https://lnkd.in/dXQeA6Kx | 48 comments on LinkedIn![](https://static.ghost.org/v5.0.0/images/link-icon.svg)LinkedInThierry Thevenet![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/1776256088585)](https://www.linkedin.com/posts/european-commission%5Fit-is-for-parents-to-raise-their-children-activity-7450158038201782273-rXDm) ## The real circumvention problem The Threads commenter who noted "VPNs are also ready" was pointing at something real, though not quite in the way they meant. The cryptographic proof cannot be spoofed or transferred. But a VPN that makes a user appear to browse from outside EU jurisdiction sidesteps the obligation to verify at all. The circumvention is not of the architecture; it is of the enforcement. This is a DSA problem, not a technical one. Whether platforms are compelled to integrate the system, and whether that compulsion reaches beyond EU borders, are open questions the announcement did not address. The blueprint closes one circumvention route. Jurisdiction closes another only if enforcement follows. ## The unlaunched product The technical work is done, and it is genuinely good. What has not started is the communication work: framing this around the adult user who wants to prove their age without surrendering their identity. That person also happens to satisfy the DSA, protect children, and reduce platform liability. The upside was always there. The Commission chose not to use it. A system designed to preserve privacy should not have to overcome the impression that it destroys it. That gap is not a technical problem. It is a choice. *Discussing this on* [*LinkedIn*](https://www.linkedin.com/posts/rob-hoeijmakers%5Fdigitalidentity-europeanunion-privacy-share-7450297848182730755-2llz?utm%5Fsource=share&utm%5Fmedium=member%5Fios&rcm=ACoAAAAGPccBqY4Rvlcj2zZ2PdvOk7tl5dQrkUs) *if you want to weigh in.* ### Further reading - [Understanding Digital Identity in the EU: eID, eIDAS, and the EUDI Wallet](https://rob-hoeijmakers.ghost.io/eu-eid-eidas-eudi-wallet/) - [Wero: Europe’s new payment brand](https://rob-hoeijmakers.ghost.io/wero-europes-new-payment-brand/) - [Digital Europe in Two Acts: DSA and DMA](https://rob-hoeijmakers.ghost.io/dsa-dma-compared/) ### The article I couldn't write URL: https://rob-hoeijmakers.ghost.io/the-article-i-couldn-t-write/ Last updated: 2026-05-10T08:53:33.000Z I read a news article last week about Amsterdam asking data centres to reserve space for European clients. Something bothered me about it. I opened a new document and started writing. Three days later I had notes for three separate pieces, none of them finished, and a growing suspicion that the thing bothering me was not what I thought it was. ## The split The first article was about category errors. The piece conflated continuity, confidentiality and security, three distinct problems that require different interventions. I wrote that one fairly quickly. It felt satisfying in the way that corrections always feel satisfying, precise, a little superior. I set it aside. The second article was about Amsterdam specifically. The city at the centre of European internet infrastructure, home to one of the world's largest internet exchange points, that spent years fighting data centres and now finds itself asking the ones it allowed to hold space for European clients. The irony was clean. Too clean, maybe. I set that one aside too. The third article was the one I actually wanted to write. It was about the NIMBY problem. How the cities most anxious about digital sovereignty are the ones most aggressively preventing the infrastructure sovereignty requires. How we export our data centres the same way we export our waste: to somewhere poorer, somewhere with weaker rules, somewhere that lacks the political capital to refuse. And how the people who campaigned against the incinerator keep producing rubbish, just without having to look at it. That one felt true. But it also felt incomplete. Because the NIMBY argument is still a solutions argument. It still assumes we know where we are going and just need to remove the obstacles. And I am no longer sure we know where we are going. ## The habit underneath Here is what I kept circling back to. Digital dependency on American infrastructure is not a procurement failure. It formed the way habits form: for real reasons, over time, because it kept making sense. The services were good. The prices were low. The integration was seamless. You did not end up here through weakness or inattention. You ended up here because it worked, and kept working, until the geopolitics shifted and suddenly the habit was visible in a way it had not been before. The recovery metaphor is useful up to a point. You make a policy. You change the default at each renewal. You take some things in-house. You support European alternatives even when they are not quite as good. I have done some of this in my own practice. Not the hard parts yet. The core productivity stack is still there. I tell myself I am waiting for the right moment, which is probably true and probably also a rationalisation. But recovery assumes you know what you are recovering toward. What does a healthy relationship with digital infrastructure look like? What does a European digital society actually feel like, cost, require? Those questions do not have clean answers. And the conversation keeps avoiding them, because solutions are easier to talk about than destinations. ## Where we were going We built the digital world we built because we were optimising for something. Speed. Convenience. Scale. Price. The platforms that won were the ones that removed friction fastest and charged least. We did not ask, collectively, what we were giving up or where we were heading. We just went. Individuals, companies, governments, all of us, at different speeds, in the same direction. Now the direction is contested. And the response, ten-year strategies, cloud alliances, rack space negotiations, is still being framed as a logistics problem. How do we get from here to sovereignty? What are the steps, the phases, the milestones? I keep wanting to ask a prior question. Sovereignty for what? A European digital society organised around what values, what trade-offs, what vision of what we owe each other and what we want to keep private and what we are willing to share? That conversation is not happening at the policy level. It is barely happening at the cultural level. And without it, the logistics question has no real answer, because you cannot plot a route to a destination you have not named. ## The unfinished document So I have three unfinished articles and this one, which is about why I could not finish them. I think the honest position is this. The sovereignty conversation is real and urgent and I take it seriously. I am also not sure the vocabulary we have for it is adequate to the actual problem, which is less about infrastructure than about intention. Less about where the servers are than about what kind of digital life we are trying to build, and for whom, and under whose rules, and whether we are willing to pay what it costs. I do not have an answer to that. I am not sure anyone does right now. But I think it is the question sitting underneath all the other questions, and I wanted to at least put it on the page. The three articles will probably get written eventually. This one needed to come first. ### Further reading - [The Paradox of Sovereignty: Europe’s Search for Freedom in a Connected World](https://rob-hoeijmakers.ghost.io/sovereignty-paradox/) - [The AI Continuity Problem](https://rob-hoeijmakers.ghost.io/the-ai-continuity-problem/) ### Google AI Plus, Pro, Ultra: Clearing Up a Year of Name Changes URL: https://rob-hoeijmakers.ghost.io/google-one-vs-google-workspace-for-gemini/ Last updated: 2026-05-10T08:53:33.000Z A year ago I published a piece trying to make sense of Google's paid AI options. At the time the product was called Gemini Advanced. It sat inside a Google One subscription, and the main question was whether it was worth adding alongside a Google Workspace plan. That post still gets traffic. The product it describes has been renamed twice since. If you searched for "Gemini Advanced" or "Google One AI Premium" trying to figure out what happened to your subscription, this is the updated version. ## The name history The confusion starts here, so it is worth naming it directly. **Gemini Advanced** launched in early 2024 as the premium AI tier inside Google One. Access to the most capable Gemini model, plus expanded cloud storage. **Google One AI Premium** was the next name. Same product, same position, different label. The "AI Premium" framing tried to clarify that this was not just a storage subscription. **Google AI Pro** is the current name, now sitting in the middle of a three-tier structure: Google AI Plus below it, Google AI Ultra above it. The plans live at one.google.com, still under the Google One brand. 💡 Google One AI Premium = Google AI Pro. Three names, one product lineage. If you subscribed to either earlier name, you are now on Google AI Pro. Nothing was cancelled; it was repositioned. Three rebrands in roughly eighteen months is a lot. It is not just cosmetic churn. It reflects something real: Google has been working out how to position a standalone AI subscription when its AI is simultaneously embedded in Search, in Workspace, and available for free in the Gemini app. The naming kept shifting because the answer to "what is this, exactly" kept shifting. ## The free baseline: Gemini the app Before getting to the paid plans, one thing worth being clear about: Gemini the chat interface at gemini.google.com is free. You do not need a paid plan to use it. It gives you access to a capable model for everyday tasks, and it is separate from both the paid Google AI plans and from Google Workspace. The paid plans give you higher access to better models within that same app, more generous limits, and access to features like Deep Research and NotebookLM at higher capacity. But the app itself is not paywalled. That distinction matters when evaluating whether any of the paid tiers are actually worth it for your use. ## The three paid tiers All three are personal subscriptions, tied to a personal Google account. Workspace accounts cannot access them directly. **Google AI Plus** is the entry tier. More access to Gemini 3.1 Pro, Deep Research, NotebookLM with expanded limits, 200 GB of storage, and family sharing with up to five others included. **Google AI Pro** is the middle tier, the direct successor to Gemini Advanced. Higher access to the same features, 5 TB of storage, and it now extends into developer tools: Jules, Gemini CLI, and $10 in monthly Google Cloud credits. This is where most personal users considering an upgrade will land. **Google AI Ultra** is the top tier. Highest model access, 30 TB of storage, YouTube Premium included, and experimental features including Project Mariner, a browser agent for automated tasks. Many of the more ambitious features remain US-only for now, worth checking before committing at this price point. Family sharing is standard across all three tiers. The primary subscriber shares AI access with up to five family members at no extra cost per person. For households where multiple people want access, Plus or Pro is a reasonable deal. ## Workspace: bundled whether you want it or not For organisations, the path is different. Google Workspace runs on Business Starter, Standard, Plus, and Enterprise tiers, priced per seat. Since January 2025, Gemini is bundled into all plans, accompanied by a 17 to 22 percent price increase. The separate Gemini add-on is gone. You are paying for it regardless. The depth of AI access varies by tier. Business Starter includes limited daily Gemini access and the standalone Gemini app. The Gemini side panels in Gmail, Docs, Sheets, and Meet only appear from Business Standard upward. If your team intends to actually use it in their daily apps, Standard is the realistic floor. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/03/Google-Workspace-Gemini.jpeg) In order to use Gemini in Google Workspace you need to give access. What I have done personally: switched most of the Workspace AI features off. Partly for privacy reasons, partly because the integration feels uneven in practice. The features are there, but they do not always fit the workflow. Google lets you toggle them, which is worth knowing. Bundled does not mean mandatory. ⚠️ Workspace accounts cannot subscribe to Google AI plans. These are personal Google account subscriptions only. If your work email runs through Workspace, you need a separate personal Google account to access the Google AI tiers, or rely on whatever Gemini features are included in your Workspace plan. [Google’s AI for Individuals and SMBs: Clarity, Costs, and Strategic ChoicesGoogle’s AI options are powerful but confusing. Here’s how I compared Google One vs Workspace—and why SMBs should think twice before locking in.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-105.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_3823.jpeg)](https://rob-hoeijmakers.ghost.io/googles-ai-for-individual-smb/) Update June 2025: Deep dive into Google Workspace and AI ## An honest aside Over the past year, I have shifted more of my work to Claude, and I want to be transparent about why that matters for how I read Google's offering. Part of it is capability and fit for the work I do. Part of it is something harder to name about Google as a product company. The naming history above is one symptom. Google Cloud is another: powerful, but it consistently requires more orientation than it should. The AI plans themselves span Google One, Workspace, [AI Studio](https://aistudio.google.com/prompts/new%5Fchat), and the free [Gemini app](https://gemini.google.com/app) in ways that are not always obviously connected. For practitioners working across languages and tools, the seams show. Dutch language support, for instance, lags behind what you get from Claude or even ChatGPT in several respects. Google is making real progress at the model level. NotebookLM remains genuinely useful for source-based research. The Workspace integration, for teams already living in Google's apps, is a coherent advantage, as I wrote in [The Workspace Advantage](https://rob-hoeijmakers.ghost.io/google-gemini-workspace-future/). But coherence at the product level is a different thing, and it is not where Google is strongest. ## Where this leaves the choice The question of which tier, if any, is worth it depends on how deep you already are in Google's world. For families sharing a personal subscription, Plus or Pro is a reasonable entry point. For organisations on Workspace, Gemini is now included whether you asked for it or not, but you can turn most of it off. For practitioners working across multiple tools and suppliers, the choice is less obvious than the marketing suggests. The free Gemini app covers more ground than most people realise before they reach for a paid plan. And if you are still searching for what happened to Gemini Advanced: it became Google AI Pro, the middle tier of a three-tier structure that Google will probably rename again within the year. ### Further reading - [The Workspace Advantage: Google’s Quiet Lead in the AI Shift](https://rob-hoeijmakers.ghost.io/google-gemini-workspace-future/) - [From Free to Paid: Choosing the Right AI Model (with a European Lens)](https://rob-hoeijmakers.ghost.io/ai-models-europe-free-vs-paid/) - [Google Gemini’s Video AI: What’s New and Why It Matters](https://rob-hoeijmakers.ghost.io/google-gemini-video-ai/) ### The AI Continuity Problem URL: https://rob-hoeijmakers.ghost.io/the-ai-continuity-problem/ Last updated: 2026-05-10T08:53:34.000Z Whisper runs on a Mac mini in my office. It transcribes conversations locally, no cloud, no API call, no dependency on a provider staying available or affordable. It is not the most capable transcription tool I have access to. It is the one I control. That distinction has started to matter more. ## Load-bearing AI is no longer optional in how I work. It sits inside research, writing, client work, automation. Removing it would not be an inconvenience, it would be a business interruption. That is a useful capability to have built, and also a new kind of exposure. The failure modes are not exotic. A geopolitical rupture that places European businesses on the wrong side of US export controls. A provider that decides your sector, use case, or geography carries too much compliance risk. An API pricing change that breaks your unit economics overnight. A network outage on a day when the work cannot wait. None of these require imagination. Some are already happening at the edges. ## The minimum viable hedge My response so far is not a full architecture. It is a posture. The first line is supplier diversity. I use multiple models across multiple providers. No single point of failure, no single commercial relationship that can strand me. This is the cheapest form of continuity planning: just don't consolidate everything onto one platform. The second line is selective local capability. Whisper for transcription is the working example. A Mac mini running a local LLM for specific workflows is the experiment in progress. The goal is not to replicate cloud AI locally, that is neither feasible nor worth the investment for a small operation. The goal is to keep the door open: to maintain enough local capability that critical workflows can survive a cloud disruption, even in degraded form. [Running Gemma 4 on Your iPhoneGoogle’s Gemma 4 runs offline on your iPhone. A follow-up to my local LLM experiment, now with a sharper app, a better model, and a clearer sense of what this category is becoming.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-544.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_4604-3.jpeg)](https://rob-hoeijmakers.ghost.io/running-gemma-4-on-your-iphone/) The constraint is real. Time spent building local infrastructure is time not spent on client work. The investment has to be proportional, which means choosing carefully which workflows justify it. Transcription earns it: the volume is high, the sensitivity is real, and Whisper works well enough to be a genuine alternative. ## Continuity as strategy Most small companies and solo practitioners haven't made this decision consciously. AI crept in as a convenience and became structural before anyone asked what happens if it stops. That is not a criticism, it is the normal pattern with any infrastructure shift. Electricity, cloud storage, SaaS: each time, dependency arrived before continuity planning did. ⚠️ The risk isn't that AI becomes unavailable everywhere. It's that your specific access, through a specific provider at a specific price point, becomes unavailable. Diversification and partial local capability are the practical responses. The question for any AI-dependent operation is not whether to use cloud models. It is whether you have thought through what you would do without them, and whether the answer is better than "stop working." A local Whisper instance and a second API account is not a complete answer. It is a start, and it is more than most have. ### Further reading - [Llama 3.2 on a Mac](https://rob-hoeijmakers.ghost.io/llama-32-on-a-mac/) - [Running a Local LLM on Your iPhone](https://rob-hoeijmakers.ghost.io/running-a-local-llm-on-your-iphone/) ### The Neural Engine Does Not Run Your LLM URL: https://rob-hoeijmakers.ghost.io/the-neural-engine-does-not-run-your-llm/ Last updated: 2026-05-10T08:53:34.000Z In November I wrote that Apple's Neural Engine powers on-device AI features such as transcription, photo recognition, and real-time translation. That was correct. What I left unexamined was the assumption underneath: that the Neural Engine is also where local language models run. It is not. This is not practical knowledge in the sense that it changes what you do tomorrow. It is foundational: understanding what the iPhone actually is as an AI platform, and where its architecture is heading. [From Silicon to Intelligence: Understanding the Hardware Behind AIA short video about NPUs and TPUs led to a deeper look at the physical side of AI. From the Neural Engine in your iPhone to the massive processors powering data-centre models.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-533.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/9F5BB129-0CFF-4279-B7F0-62E7C488B494-4.jpeg)](https://rob-hoeijmakers.ghost.io/cpu-gpu-tpu-npu-explained/) ## Fixed versus flexible The Neural Engine is a fixed-function accelerator. Apple designed it for operations that are known in advance, where the shape of the computation does not change between runs. Face ID. The camera pipeline. Siri's wake word. For these tasks, Apple can design dedicated circuits, run them at around 2 watts, and finish faster than a general-purpose processor could. The chip is fast, efficient, and inflexible by design. Language model inference works differently. The attention mechanism in a transformer shifts with every token. Memory access patterns are irregular. The compute graph changes with the length of the input. A fixed-function accelerator cannot adapt to this. What you need is a programmable processor, and on iPhone that is the GPU. 💡 Fixed-function accelerators trade flexibility for efficiency. They are optimised for specific operations and cannot handle workloads outside that envelope. The Neural Engine is fast at what it was built for. LLM inference was not in that brief. ## What makes local LLMs work Apple's MLX framework, which is what serious on-device model work runs on, was designed for Apple silicon's unified memory architecture. CPU, GPU, and Neural Engine share the same memory pool. No data needs to move between separate chips. For LLM inference, MLX routes to the GPU. A small quantised model fits in the shared pool alongside everything else, and the GPU works through it without transfer overhead. The unified memory is the real architectural story. The Neural Engine TOPS figure Apple leads with in chip announcements describes something genuine, just not LLM inference. It describes how fast Face ID runs. ## The energy gap Your phone battery lasts a day because most of its processors sleep most of the time. The Neural Engine, at around 2 watts, can run continuously without meaningfully changing that. The GPU, at around 20 watts under load, cannot. Leave it running and your phone is warm in your pocket and dead by lunch. That gap is why the Neural Engine exists at all. Apple uses it for tasks that never fully switch off: the microphone listening for "Hey Siri," the camera recognising scenes before you tap the shutter. Small models, running constantly, at a power cost you never notice. Now imagine that kind of always-on behaviour applied to a language model. An assistant that reads your next meeting and prepares context. A translation layer active in the background. A model that processes before you ask. That is where the Neural Engine becomes interesting again for AI, because the GPU cannot sustain that role without draining the battery. The catch is model size. Today's LLMs are too large and too hungry to live on the Neural Engine. But models are shrinking fast through quantisation and distillation, and the threshold keeps moving. At some point, a capable enough model will fit within what the Neural Engine can handle continuously. When that happens, always-on AI on iPhone becomes practical in a way it currently is not. We are not there yet. For now, when you run a local LLM on iPhone, the GPU does the work. I have been experimenting with exactly this, and the results are more capable than I expected. --- ### Further reading - [Running a Local LLM on Your iPhone](https://rob-hoeijmakers.ghost.io/running-a-local-llm-on-your-iphone/) ### Running Gemma 4 on Your iPhone URL: https://rob-hoeijmakers.ghost.io/running-gemma-4-on-your-iphone/ Last updated: 2026-05-10T08:53:34.000Z Last summer I ran [a local language model on my iPhone](https://rob-hoeijmakers.ghost.io/running-a-local-llm-on-your-iphone/) using Haplo AI. Gemma was the only model that actually produced a useful result. Now Google has shipped Gemma 4, and the story has changed enough to revisit. ## What Gemma is Gemma is Google's family of open-weight models, designed to be downloaded and run locally rather than accessed through a cloud API. The distinction matters. Where Gemini lives on Google's servers and requires a connection, Gemma publishes the weights. You can download them, run them on your own hardware, quantize them for smaller devices, fine-tune them for specific tasks. This is also why [Gemini Nano](https://rob-hoeijmakers.ghost.io/chrome-gemini-nano-and-the-browser-as-ai-platform/), which ships inside Chrome and on Pixel devices, is not the same thing. Nano is part of the closed Gemini family: Google controls it, distributes it as a black box, and you call it through an API without ever seeing or owning the model. Gemma is a separate release, designed from the start for local deployment. Same research lineage, completely different distribution model. Why multiple sizes? A 2B model fits on a phone and responds in seconds. A 27B model needs a laptop with a capable GPU. Gemma 4 comes in four variants; the smallest runs on a current iPhone. The choice is always between what the hardware can carry and what the task actually needs. 💡 Gemma 4 comes in E2B, E4B, 26B MoE and 31B Dense variants. The E2B (about 2.5GB download) runs on an iPhone 15 Pro or newer. Multimodal support, meaning it can reason about images as well as text, is included in the smaller variants. ## Under the hood Two apps let you run Gemma on iPhone, and they are built differently in ways that show. **AI Locally** uses MLX, Apple's own machine learning framework, built around the unified memory architecture of Apple Silicon. The model runs on the GPU via Metal, tight and native. Google **AI Edge Gallery** uses LiteRT, which Google built for Android NPUs, then translated into Metal on iOS. It works, but it is an extra step that Apple's own framework does not need. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/04/IMG_4680.jpeg) Edge Gallery and Locally AI for iOS Neither uses the [Neural Engine for LLM inference](https://rob-hoeijmakers.ghost.io/the-neural-engine-does-not-run-your-llm/), despite what the marketing around "on-device AI" often implies. Language models need the GPU's flexibility. Face ID and camera processing run on the Neural Engine because they are fixed, predictable operations. Running the E4B model, the larger of the two downloadable Gemma 4 variants, my iPhone 16 Pro got noticeably warm. That is not a complaint. It is the GPU working: the phone is doing real compute, not calling a server. [The Neural Engine Does Not Run Your LLMMy earlier post on AI hardware implied the Neural Engine handles on-device AI. It doesn’t, not for LLMs. Here is what it actually does, and why the distinction matters.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-539.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_4620-1.jpeg)](https://rob-hoeijmakers.ghost.io/the-neural-engine-does-not-run-your-llm/) ## Demo versus tool [Edge Gallery](https://ai.google.dev/edge) is polished and honest about what it is. Conversations are ephemeral, nothing is saved between sessions, and the feature set reads like a capabilities showcase: image questions, audio transcription, a tool-calling demo. Google built it to show what Gemma 4 can do on a phone. For that purpose it works well. [Google AI Edge | Google AI for DevelopersIntroducing AI Edge![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon-new.png)Google AI for Developers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/share.png)](https://ai.google.dev/edge) [AI Locally](https://locallyai.app) is building toward something different. It has Shortcuts integration, a voice mode (English only for now), and a model browser that flags which models will run well on your specific device. I have been using it more than Edge Gallery, and the preference is clear after a few sessions. [Locally AI - Run AI models locally on your iPhone, iPad, and Mac.Run Llama, Gemma, Qwen, DeepSeek, and more on your iPhone, iPad, and Mac. Optimized for Apple Silicon. Offline. Private.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon-1.png)Locally AIAdrien Grondin![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/preview.jpg)](https://locallyai.app) The limitations are also clear. The context window is small: long documents hit the ceiling quickly, PDF reading crashed outright, and transcription is not a realistic use case at this scale. These are not edge cases; they are the main things you would reach for a cloud model to do. 💡 AI Locally also has a [Mac app](https://apps.apple.com/nl/app/locally-ai-local-ai-chat/id6741426692?l=en-GB), built on the same MLX foundation. It runs local models on Apple Silicon, including the Apple Foundation Model — Apple's own on-device LLM, accessible without a cloud connection. Worth installing if you are already on the iPhone app. ## Progress, with caveats What has actually changed since my Haplo AI experiment is the baseline. The hardware has caught up: on a recent iPhone, a 3B to 4B model runs fast enough to feel responsive. Gemma 4 handles Dutch well, which matters more than it might seem for a model running entirely on device. And the apps have matured from rough experiments into something you could actually use. What has not changed is the gap with cloud models. Local AI on a phone is useful for bounded tasks where the input is sensitive: a financial document you have not decided to share, notes that should not leave the device, a question you would rather not route through a server. For that use case, the combination of privacy and reasonable quality is now genuinely good enough. For everything else, the cloud model is still the right tool. The context window is larger, the reasoning is deeper, and nothing crashes when you open a PDF. ### Further reading - [The AI Continuity Problem](https://rob-hoeijmakers.ghost.io/the-ai-continuity-problem/) - [Running a Local LLM on Your iPhone](https://rob-hoeijmakers.ghost.io/running-a-local-llm-on-your-iphone/) - [The Neural Engine Does Not Run Your LLM](https://rob-hoeijmakers.ghost.io/the-neural-engine-does-not-run-your-llm/) ### My First Mac App URL: https://rob-hoeijmakers.ghost.io/my-first-mac-app/ Last updated: 2026-05-10T08:53:34.000Z The problem was simple and specific. Every time I wanted to save something to a Markdown file, I had to open an editor, paste, name the file, save. The sequence could take a minute. I looked for a web clipper, but every option pulled me into browser extensions and cross-platform wrappers. What I needed lived at the OS level, not the browser level. So I described it to Claude. And Claude wrote me [a Swift app](https://rob-hoeijmakers.ghost.io/clipboardmarkdown/). ## What it does The app sits in my Applications folder. I copy something, open it, choose between plain text or HTML, and save directly to a folder. That is it. Along the way I learned something I hadn't known: clipboard content exists in multiple formats simultaneously. The same copied text is available as plain text, as HTML, as rich text. Choosing between them is itself a design decision. The app makes that choice explicit. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/04/image.png) ## The category shift What matters here is not the app. It is the category it belongs to. Swift is a compiled language. Xcode is a professional framework. This is not a Python script or a browser extension. It is native software, built with the same tools Apple developers use, running with the same OS-level access. I am not a developer. I can think in terms of digital products, reason about edge cases, describe behaviour precisely. What I have never had is the ability to translate that into something executable at this level. That translation layer is what shifted. [AI Coding Boom 2026: GitHub Repos, npm Downloads, and Product Hunt Launches Are SurgingAI coding tools have driven GitHub repos from 2.9M to 6.8M/month in 3 years. See the data behind the 2026 building boom across npm, Product Hunt, and Show HN.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/icon-3.png)Val KamenskiVal Kamenski![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/og.8890b7d6.png)](https://www.kamenski.me/articles/ai-coding-boom-2026-github-npm-data?utm%5Fsource=chatgpt.com) ## Software for one Most software is general-purpose because specificity doesn't scale. The economics push toward horizontal tools: something that solves a problem for enough people to justify building it. What AI makes possible is the opposite: software for one workflow, one environment, one person's exact friction. It doesn't need a product launch or a pricing page. It just works. I have a mental list of problems like this, accumulated over years: too specific to find an existing solution, too niche to commission. That list is shorter now. The question I keep returning to: as more people with product thinking but no coding background start shipping their own narrow tools, what does that do to the market for the broad ones? [ClipboardMarkdownClipboardMarkdown is a free macOS menu bar utility that saves clipboard text as a Markdown file in \~/Downloads. No setup, no subscription.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-531.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/clipboardmarkdown-hero.png)](https://rob-hoeijmakers.ghost.io/clipboardmarkdown/) I made the app available, give it a try. ### Further reading - [ChatGPT Thinks With You. Claude Builds With You.](https://rob-hoeijmakers.ghost.io/chatgpt-thinks-with-you-claude-builds-with-you/) - [Using markdown for blogging](https://rob-hoeijmakers.ghost.io/using-markdown-for-text/) - [OS-Level Translation: A Multilingual Productivity Hack](https://rob-hoeijmakers.ghost.io/os-level-translation-a-multilingual-productivity-hack/) ### ChatGPT Thinks With You. Claude Builds With You. URL: https://rob-hoeijmakers.ghost.io/chatgpt-thinks-with-you-claude-builds-with-you/ Last updated: 2026-05-10T08:53:35.000Z ChatGPT made me think better. Claude makes me build differently. That distinction took a few weeks of actual use to understand, and it's the reason I switched. With ChatGPT, the dynamic was cognitive support: a sharper search, a thinking partner, a way to process ideas faster. Useful, often excellent. But the work still lived with me. The tool assisted; I produced. Claude changed the direction of that relationship. The environment composes. ## Teaching, not prompting The mechanism is project files, CLAUDE.md files, and skills. You invest time once in explaining your context, your style, your workflow logic. The system holds it. Every subsequent session starts from that foundation rather than from scratch. That investment compounds. I have a skill for drafting posts on this blog, one for processing Schmuki meeting transcripts, one for bookkeeping in MoneyBird. Each encodes a workflow I would otherwise rebuild conversation by conversation. The result is consistency at a pace I couldn't sustain manually. One person, producing at team scale. This is different from prompting. Prompting is asking. Teaching the environment is closer to hiring: you explain how things work here, and the system operates accordingly. ## Claude Code and the Git prerequisite Claude Code added another layer. It's a terminal-first coding environment, agentic in a way the web interface isn't. I used it to build a treemap of Dutch employment and AI exposure, pulling public data, structuring it, iterating on the visualisation through conversation rather than through manual edits. The kind of data project that previously required either a developer or a very patient afternoon. What I hadn't expected: the unlocking skill wasn't prompt technique. It was Git. Moving a WordPress site to GitHub, learning version control well enough to work with it fluently, that turned out to be the prerequisite for getting real value from Claude Code. The model handles the code. Git handles the ground beneath it. 💡 Getting started with Claude Code requires sorting out the right subscription tier and, more practically, getting comfortable with Git. Both are one-time investments that change what's possible afterwards. ## What digital crafting feels like There's a term forming for what this is: digital crafting. Not using AI as a tool, and not being replaced by it. Building with it, encoding your practice into the environment, and operating at a scale that was previously out of reach for a single practitioner. The question I'm sitting with: as more of the workflow gets encoded and more of the output gets consistent, where does the craft actually live? I think it lives in the teaching. In knowing what to encode, how to instruct, where to leave room for judgment. That's not a diminished form of work. It's a different one. ⚠️ One current limitation for Dutch users: the Claude iOS app doesn't support Dutch yet. For voice interaction and mobile use, that's a real gap. Worth watching as Anthropic expands language support. ### Further reading - [The Claude plan I actually use](https://rob-hoeijmakers.ghost.io/claude-plans-usage/) ### Leuven's quiet superpower: imec URL: https://rob-hoeijmakers.ghost.io/leuven-imec/ Last updated: 2026-05-10T08:53:35.000Z I visited Leuven last year. Beautiful city, perfectly preserved, Gothic town hall, students everywhere. On its edge I walked past a modern building with the name imec on it. It looked serious. I had no idea what it was. Turns out it might be one of the most strategically important buildings in the global semiconductor industry. ## Nobody's chip company Imec employs researchers from over a hundred countries and holds no manufacturing lines of its own. It is, by design, nobody's chip company. That turns out to be the point. In 1984, the Flemish regional government did something unusual. It didn't attract a semiconductor manufacturer, or subsidise an existing one. It funded a research commons: a neutral laboratory where competitors could work on shared problems without handing advantage to a rival. The founding team were young researchers, several of them returned from Silicon Valley, backed by a government willing to think in decades rather than quarters. They called the ambition a SuperLab. It became imec. [Imec R&D, nano electronics and digital technologiesImec is the world-leading R&D and innovation hub in nanoelectronics and digital technologies.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon-47.ico)imec![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/share-image_1.jpg)](https://www.imec-int.com/en) ## The paradox Most technology clusters form around an anchor company: [ASML](https://rob-hoeijmakers.ghost.io/zeiss-and-asml/) in Eindhoven, ARM in Cambridge. Leuven built the infrastructure first and let the ecosystem follow. Because Belgium had no dominant chip producer to protect, imec could be structurally neutral. Intel, Samsung, TSMC and others could collaborate there without the research benefiting a direct competitor. The absence of a national champion was the founding condition for a global one. ## Nine days ago That neutrality is now carrying serious weight. On March 18, imec announced the installation of the ASML EXE:5200: the most advanced lithography machine in existence, one of fewer than a dozen available anywhere in the world, valued at $400 million. It sits in imec's cleanroom in Leuven, available to the global ecosystem of chip makers, materials suppliers and research partners. Not owned by one company. Shared, on neutral ground, to accelerate what comes next. The machine targets sub-2nm chip development, territory that doesn't yet exist in commercial production. [Imec receives the world’s most advanced High NA EUV system | imecA major milestone in propelling industry into the angstrom era.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon-46.ico)imec![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/DSC00975.JPG)](https://www.imec-int.com/en/press/imec-receives-worlds-most-advanced-high-na-euv-system) Forty years after the SuperLab idea, the logic holds. Europe doesn't manufacture the world's leading chips. But it owns two of the choke points that make those chips possible: ASML builds the only machines capable of printing them, and imec is where the industry learns how to use what comes next. ## Invention, not policy What did the Flemish government actually build in 1984? A research institute, formally. But the design choices, non-profit structure, university roots, enforced neutrality, long time horizons, were not incidental. They were the product. Imec works because of what it was prevented from becoming as much as what it was built to be. That's closer to invention than policy. The EU Chips Act is now directing billions toward semiconductor capacity across Europe, and imec sits at the centre of it. But imec's CEO Luc Van den Hove is more cautious than the policy framing suggests: allied collaboration, not autarky. The chip supply chain is too interdependent to be repatriated. What Leuven has can't be replicated by a funding announcement. It took forty years to build, and required above all the discipline not to turn it into something else along the way. 💡 Imec stands for Interuniversity Microelectronics Centre — in Dutch: Interuniversitair Micro-Electronica Centrum. Founded in 1984 as an autonomous non-profit, spun out of KU Leuven, one of Europe's oldest universities. They officially style the name in lowercase: imec. ### Further reading - [How ZEISS and ASML Enable the Modern Chip Industry](https://rob-hoeijmakers.ghost.io/zeiss-and-asml/) - [Strengthening Europe’s Semiconductor Backbone: The Role of ESMC](https://rob-hoeijmakers.ghost.io/esmc/) - [From Silicon to Intelligence: Understanding the Hardware Behind AI](https://rob-hoeijmakers.ghost.io/cpu-gpu-tpu-npu-explained/) ### Back to the file URL: https://rob-hoeijmakers.ghost.io/back-to-the-file-2/ Last updated: 2026-05-12T21:43:54.000Z That was 1995 or thereabouts. The web was young, the tools were primitive, and the directness of it all felt natural because there was no alternative. Then the alternatives arrived. Content management systems. Page builders. Caching layers. Each one a reasonable response to a real problem. Together, a stack that most people running a small website never fully understood and never really needed to. The web got easier to use and harder to own. ## The stack I was running [Schmuki.nl](https://schmuki.nl), the website of the agency I run with Maya Schmuki, was on WordPress. With Divi on top for visual design, and LiteSpeed Cache underneath for performance. Three systems, each solving a limitation of the one below it. It worked. I was never happy with it. WordPress is the largest CMS in the world, which tells you something. It grew by becoming a collection point: a core you extend with plugins for anything you might need. Every plugin solves something. Together they form a system that nobody designed as a whole and nobody fully understands. Mediocrity at scale. Not broken enough to replace, never good enough to trust completely. For a fifteen-page agency website with no dynamic content, no shop, no membership area — the whole thing was elaborate infrastructure for what is ultimately just a set of files. ## What Claude Code made possible I migrated the site using Claude Code, the command-line version of Claude. A coding agent: you describe what you want done, it does it. I am not a developer. I never have been. But I know what I want, and I know how to judge whether something works. Claude Code directed the process: export the site as static HTML using Simply Static, clean up the images folder, fix all path references, push to GitHub Pages. Step by step, with explanation. I executed, reviewed, fed back what I saw. The judgement stayed with me. The technical execution largely did not have to. The one real obstacle was LiteSpeed Cache, which generates CSS and JavaScript with hash-based filenames that Simply Static cannot discover. The export was broken until I understood why — and disabling LiteSpeed before running the export solved it. That moment of diagnosis was the one place where my own understanding actually mattered. Everything else, Claude Code handled. ## The strange familiarity of it The site now lives on GitHub. It is a folder of HTML files, a folder of images, and a CSS file. I make changes through Claude Code. The contact form runs through Formspree. There is no server to maintain, no database to worry about, no plugin to update. It feels oddly like 1995\. Not because the technology is the same — it is not, not even close — but because the relationship to the material is the same. You know what is there. You put it there. There is no layer between you and the file. What is different is that you no longer need to be a technician to work this way. Claude Code is what closes that gap. It translates intent into action precisely enough that the directness developers have always had access to is now available to anyone willing to stay in the loop and exercise judgement. That is not a small thing. Thirty years later, I am back to editing files. It turns out that was the right way to work all along. ### EU Inc.: between words and reality in Europe’s new company proposal URL: https://rob-hoeijmakers.ghost.io/eu-inc-europe-company/ Last updated: 2026-05-10T08:53:35.000Z I have spent a fair bit of time over the past months on something quite mundane: setting up and running a company in the Netherlands. Not in theory, but in practice. Notaries, registrations, tax, the small frictions that come with it. So when the European Commission comes out with a proposal that says you can start a company in 48 hours, fully online, for under €100, it immediately catches my attention. Not because I take the headline at face value, but because I recognise the problem it is pointing at. This proposal is part of what is being called the “28th regime”, often referred to as EU Inc. [Signing the Future: How a Digital Incorporation Made eIDAS RealA first-hand look at how eIDAS and digital signatures replace paper rituals with cryptographic trust.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-520.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_0188-3.jpeg)](https://rob-hoeijmakers.ghost.io/digital-signature-eidas-experience/) ## What actually happened Nothing has changed yet. I cannot use this. You cannot use this. It does not exist in practice. What the Commission did is put a proposal on the table. That sounds small, but it isn’t. It moves EU Inc. from an idea into something that now has to be discussed, negotiated, and shaped. ## Why this feels different Europe has tried this before. There is already a European company form, the Societas Europaea. It exists, but in practice you rarely encounter it unless you are dealing with larger structures. So it is not the ambition that is new. What feels different to me is the approach. Instead of trying to align everything between member states, this proposal more or less accepts that alignment is hard, maybe too hard, and tries something else: add a layer on top that is simpler to use. That is a pragmatic move. EU Inc. does not replace national systems. It sits next to them. A 28th option, alongside the 27 existing ones. ## From system to experience The way it is presented stands out. 48 hours. Fully online. Under €100. These are not legal categories. This is how you would describe a service. If you have ever tried to do anything across systems, that framing makes sense. You stop thinking in legal structures and start thinking in time, cost, and predictability. That is what EU Inc. is trying to address, at least at the starting point. ## Giving it a face This is not a quiet technical file somewhere in Brussels. Ursula von der Leyen is putting her name on it. That gives it weight, but also exposure. If this gets diluted or disappears, it will be visible. If something meaningful comes out of it, that will be visible too. So it raises the stakes. ## How I look at it What strikes me is that this is no longer just an idea. EU Inc. has been around as a concept for a while. I wrote about it earlier as a way to think about a more unified European company structure. Now it has crossed a line into something the Commission has formally picked up and translated into a proposal. At the same time, I know how these processes tend to go. There are domains that remain national and sensitive: taxation, labour law, insolvency. Those are not part of this proposal, and they are not easy to align. So the question is not whether Europe can make registration faster. It is whether that first step can meaningfully reduce the friction that comes after. [One European Company (EU-INC)The idea of a single European company keeps returning. Not because it’s easy, but because Europe keeps running into the same limits.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-521.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/8EB28EBF-9AAC-4623-8754-316E93AC9A01-1.jpeg)](https://rob-hoeijmakers.ghost.io/one-european-company-eu-inc/) ## Where that leaves me I would not plan anything around this yet. But I do take it seriously as a signal. There is a shift here. Less patience with internal complexity, and a more pragmatic attempt to work around it instead of resolving it completely. EU Inc. is not reality yet. But it is no longer just an idea either. It is now a proposal that others have to engage with. And from experience, that is usually where things start to move. Slowly, unevenly, but still, in a direction. ### Further reading - [‘48 hours and €100’: EU presents sped-up business registration law](https://www.euronews.com/my-europe/2026/03/18/48-hours-and-100-eu-presents-sped-up-business-registration-law) ### Mapping AI Exposure Across the Dutch Labour Market URL: https://rob-hoeijmakers.ghost.io/mapping-ai-exposure-across-the-dutch-labour-market/ Last updated: 2026-05-15T10:56:19.000Z I set out with a fairly contained question: what does AI exposure look like when you map it onto the labour market? In that process I came across a model of the US job market that had already disappeared. What remained was a fork, incomplete but workable, and that was enough to start rebuilding it. What began as a reconstruction gradually turned into something else. Not only an attempt to understand the Dutch situation, but also an exploration of how this kind of work can now be approached at all. [Nederlandse Arbeidsmarkt & AI Invloed114 beroepsgroepen · 9,7M werkenden · AI-invloed gescoord door Claude op ESCO-omschrijvingen![](https://static.ghost.org/v5.0.0/images/link-icon.svg)![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/preview-7.png)](https://schmuki.nl/beroepen) ## Translating across systems The core challenge was not technical in the narrow sense, but structural. The [original model](https://karpathy.ai/jobs/) was grounded in US job classifications, and those do not translate directly into a European or Dutch context. So the work became a layered translation. From US roles to [ESCO classifications](https://esco.ec.europa.eu/en), and from there to Dutch [CBS](https://www.cbs.nl) data. That mapping step is where this kind of analysis usually becomes slow and specialised, because it requires both domain understanding and a fair amount of manual alignment. What changed here is how that alignment was done. I used a coding assistant to work through the repository and adapt the structure, but I also relied on AI in a more interpretive way, using it to match categories across systems. That kind of classification work would normally be painstaking and manual. Here, it became something you can iterate on, check, and refine in dialogue. [US Job Market Visualizer![](https://static.ghost.org/v5.0.0/images/link-icon.svg)](https://karpathy.ai/jobs/) ## A stack of tools, compressed in time Looking back, what stands out is not just that it worked, but how it came together. Different layers of AI were involved, each playing a distinct role: - a coding assistant to reconstruct and adapt the model (Claude Code) - an LLM to reason across classification systems and suggest mappings - another layer, including ChatGPT, to structure, explain, and publish the result What would previously have been a sequence of specialised steps collapsed into a single workflow. Not perfectly, and not without judgement, but within the span of an afternoon. That compression of time is perhaps the most tangible shift. It changes not only what is possible, but also what feels reasonable to attempt. ## A chart that behaves like a map The resulting visualisation is, strictly speaking, a bubble chart. It combines multiple dimensions: AI exposure on one axis, job volume on another, and economic weight reflected in the size or colour of each point. But in use, it behaves more like a map. You are not analysing individual points in isolation, but scanning for concentrations, for areas where pressure might build up, for combinations of exposure and scale that start to matter. That makes it a useful form for this kind of question. It does not resolve the issue, but it creates a surface on which more informed questions can emerge. 🌳 A ****treemap** visualises a hierarchy by dividing a single rectangle into nested boxes, where each box represents a category and its size is proportional to a value such as number of jobs or economic weight. Instead of drawing branches, it uses a space-filling layout that maps the entire structure into one view, making it possible to compare both structure and scale at once. In this project, that matters: the goal is not just to show how occupations relate, but how large they are relative to each other. The treemap makes that immediately visible, giving a compact overview of the Dutch labour market where distribution, imbalance, and concentration can be grasped in a single glance. ## From output to shared artefact Publishing the model via [GitHub](https://rhoeijmakers.github.io/nl-jobs/) and GitHub Pages was a deliberate step. It turns the work into something that can be inspected, questioned, and extended by others. Instead of a static report, it becomes a small public artefact. The assumptions are visible, the structure can be followed, and anyone can take it further by forking or adapting it. That changes the dynamic from presenting conclusions to enabling iteration. 0:00 /0:10 1× Options for the map ## Moving into data work, and what that means What I found most striking is that I could do this at all. I am not a data scientist, and this is not typically the kind of model I would build. And yet, with the current generation of tools, it becomes possible to move into that space and produce something that is at least directionally meaningful. There is a clear sense of satisfaction in that, but it also raises a broader question. If more people can do this, what happens to the value of this kind of work? When the production of structured, convincing visualisations becomes easier, the signal they carry becomes less reliable. The risk is not only error, but overconfidence in what looks well-formed. ## A shift in where the value sits What seems to be happening is not a disappearance of expertise, but a relocation of it. The effort required to produce the artefact decreases, but the importance of framing, validation, and interpretation increases. Decisions about how classifications are aligned, how proxies are chosen, and how limitations are handled remain decisive. Those aspects are less visible in the output, but they shape its meaning. Working through this yourself makes that more tangible. It creates both appreciation for the underlying work and a certain distance from the result. ## Closing For me, this project sits somewhere between curiosity and capability. It started as a way to understand a specific question about the labour market, but it ended up revealing something about the broader shift in how we work with data. There is a quiet satisfaction in being able to build and share something like this, not because it is definitive, but because it opens a space for exploration. And perhaps more importantly, it shows how much can now be done, carefully but independently, in the span of an afternoon. ### Further reading - [Nederlandse Arbeidsmarkt & AI Invloed](https://schmuki.nl/beroepen) ### When AI Moves Into Your Working Environment URL: https://rob-hoeijmakers.ghost.io/when-ai-moves-into-your-working-environment/ Last updated: 2026-05-10T08:53:36.000Z For most people, AI still lives in a chat window. You open a browser tab, type a prompt, receive an answer, and copy the result somewhere else. It works surprisingly well, and for many tasks that model is perfectly sufficient. Over the past weeks I started experimenting with a slightly different setup. Instead of keeping AI in the browser, I began running several AI tools directly inside my working environment. ## A different place for AI The environment I am referring to is what developers call an **IDE**, short for *Integrated Development Environment*. In simple terms, it is a workspace where files, folders, editing tools, and terminal access come together in one application. Tools like Visual Studio Code are commonly used for this. Traditionally, IDEs are used to write and manage software. But in practice they are simply structured environments for working with collections of files. That turns out to be an interesting place for AI to operate. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/03/Code-2026-03-13-19.22.57.png) The IDE I used was Visual Studio Code. ## From asking questions to working on material When AI runs in a browser, the interaction is mostly conversational. You ask something, receive a response, and decide what to do with it. When AI runs inside the working environment itself, the relationship changes slightly. The AI is no longer separate from the material. It operates in the same environment where the files already live. Notes, drafts, document collections, code repositories, and folders are all directly visible within the workspace. The AI can analyse, summarise, or modify parts of that material with much less copying and pasting. It starts to feel less like asking questions and more like working on the material together. ## A small experiment In my case the setup was simple. Inside the IDE I opened several terminal sessions and ran different AI models as command-line tools. Each session could analyse files, summarise content, or help restructure information. The interesting part was not the models themselves. The models are broadly similar to the ones people already use in chat interfaces. What changed was the **location** of the interaction. The AI was no longer something I briefly consulted. It was part of the environment where the work itself was taking place. ## Why this might matter This setup is still somewhat experimental and more common among developers. But the underlying idea may turn out to be relevant for knowledge work more broadly. Much of our work revolves around collections of material: notes, documents, drafts, datasets, and repositories of information. If AI tools operate directly inside those environments rather than outside them, the workflow changes. Less asking. More collaborating with the material itself. It is a small shift, but it may turn out to be an important one. ### Further reading - [Excel and the future cockpit of business logic](https://rob-hoeijmakers.ghost.io/excel-and-the-future-cockpit-of-business-logic/) ### A Wider Frame: Apple’s Camera Features in Practice URL: https://rob-hoeijmakers.ghost.io/a-wider-frame-apple-s-camera-features-in-practice/ Last updated: 2026-05-10T08:53:36.000Z Every now and then new hardware creates a small window for experimentation. Not a big project, just a moment to try features that were always there but never quite made it into daily use. I had two of those on my list: **Continuity Camera**, **Center Stage**, and something called **Desk View**. All three sit somewhere around the same idea: using camera technology a little more intelligently. The premise is straightforward. The camera in an iPhone is simply better than the cameras that used to be built into many Macs. Older iMacs and MacBooks often shipped with 720p webcams, mounted in a fixed position at the top of the screen. The result is familiar: a somewhat soft image, and a viewing angle that tends to look slightly up at you from below the chin. Using the iPhone changes both things at once. The image quality improves, and the camera can be placed somewhere more sensible. On a small stand, for example, slightly above eye level. That is what Continuity Camera does: it lets the Mac use the iPhone camera as its webcam. ## The First Small Obstacle Trying this out is slightly less obvious than you might expect. Most instructions suggest opening a video call in FaceTime or another meeting app. The moment you do that, the iPhone can appear as a camera option. But that approach requires calling someone, which is not always convenient. Not everyone is keen to participate in your technical experiments. The small trick I discovered is that you can simply open **QuickTime Player**, which is installed on every Mac, and start a *New Movie Recording*. In the camera menu you can select the iPhone as the camera. That immediately gives you a live video feed. No call required. It turns out to be the perfect little laboratory: you can move the phone, change the angle, check lighting, and compare it to the built-in Mac camera. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/03/image.png) Start Quick Time Player and head for the menu bar, New Movie Recording. ## What Actually Changes Two things become apparent quite quickly. First, the image quality difference is real. The iPhone camera produces a noticeably sharper and more stable image than older Mac webcams. Second, the position of the camera matters more than the resolution. When the camera sits slightly above the screen instead of inside it, the framing simply feels more natural. There are also some software features layered on top of this. Center Stage, when supported by the camera, keeps the subject centred by automatically adjusting the frame. And Desk View uses a wide-angle camera and some clever perspective correction to show what is lying on the desk below the screen. Desk View is perhaps the most curious of the three. Without moving the phone, the Mac can generate a second camera view that looks as if there is an overhead camera pointing down at the desk. It is not perfect, but it is surprisingly usable. Interestingly, that trick works with any sufficiently wide-angle webcam, including the ones built into newer iMacs. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/03/Apple-MacBook-Air-Desk-View-250305_big.jpg.large_2x.jpg) Desk View ## A Small Upgrade Trick What I came away with is that these features are less about novelty and more about incremental upgrades. If your Mac already has a good camera, the difference may be modest. But if parts of your setup are slightly dated, Continuity Camera is a simple way to upgrade them without buying additional equipment. You are essentially borrowing the camera from the device that already has the best one. ## The Minor Trade-offs There are also a few practical downsides. Using the iPhone as a camera usually means it needs to sit somewhere fixed. A small stand or mount helps. If you travel a lot, that can make the setup a little less spontaneous. And of course there is the obvious one: while the phone is serving as your camera, you cannot casually pick it up and start using it. There is also a small practical dependency: the iPhone now becomes part of your video setup. That means relying on its battery life, or occasionally plugging it in during longer calls or recordings. Still, for something that is already built into the devices, it is a surprisingly capable little toolset. Sometimes the most interesting part of new hardware is simply taking the time to explore the corners you normally ignore. ### Further reading - [A lighter computer for a different era](https://rob-hoeijmakers.ghost.io/a-lighter-computer-for-a-different-era/) ### A lighter computer for a different era URL: https://rob-hoeijmakers.ghost.io/a-lighter-computer-for-a-different-era/ Last updated: 2026-05-10T08:53:36.000Z After many years with a large Intel iMac, we recently replaced it with the current 24-inch Apple Silicon model. The change itself was technically straightforward. What surprised me was how clearly it illustrates a broader shift in computing. The computer on the desk has quietly become something different from what it was even five or seven years ago. ## Quick takeaways - The era of **large, monumental desktop machines** seems to be ending. - Much of the real work now happens in **cloud services and AI systems**. - Local computers are becoming **lighter, replaceable instruments** rather than long-term containers of data. - Migration between machines can now take **hours instead of days**. - Video communication has become **a central capability**, not an afterthought. ## The end of the monumental desktop The previous 27-inch Intel iMac was heavy, solid and almost architectural. It felt like a machine designed to last for a decade. The new one feels entirely different. Thin, light and almost casual in comparison. The weight difference alone is striking. This is not merely a design choice. It reflects a deeper shift in where computing power now lives. Much of the real capability no longer sits inside the machine itself. 🏋️ The weight difference between the two generations is larger than it looks. The old ****27-inch iMac weighs about 8.9 kg**, while the new ****24-inch model weighs roughly 4.5 kg**. In practice, the new iMac is ****about half the weight**, reflecting both the smaller display and the much more compact Apple Silicon architecture. ## The computer as an access point In everyday work, many of the most important systems now live outside the device: - ChatGPT - Claude - Gemini - cloud storage - collaborative tools Increasingly, even more technical work happens through **command-line tools** that connect to external services. The local computer becomes less of a container for software and data and more of an **interface to a distributed computing environment**. ## Migration becomes reconfiguration Another small but telling change is how easy it has become to move between machines. With relatively little data stored locally and most files in cloud storage, the transition from the old iMac to the new one took roughly an hour. The setup was done manually to avoid carrying unnecessary files across. A decade ago, replacing a main computer often meant days of migration, copying libraries, reinstalling software and reconstructing environments. Today the process looks more like **reconfiguring an instrument**. That also shortens the investment horizon. Instead of expecting a desktop to last seven years or more, the realistic lifecycle may now be closer to three or four. ## The quiet transition to USB-C Another visible change over the past years is the shift to **USB-C as the universal connector**. The older iMac had a row of different ports: USB-A, Thunderbolt, Ethernet and others. The new one essentially offers two USB-C ports and assumes that most expansion will happen through hubs or adapters. In practice this turned out to work well. [USB-C ](https://rob-hoeijmakers.ghost.io/usb-c/)has become a genuinely versatile interface for displays, storage and peripherals. The only small inconvenience was Ethernet. For a desktop machine, wired networking still feels natural, but even a standard presentation hub did not include LAN. That required ordering a separate USB-C Ethernet adapter. Apart from that, the reduced number of ports rarely becomes an issue. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/03/IMG_3781.jpeg) The ports on the older iMac. ## The screen question One small complication did appear. Moving from a 27-inch screen to a 24-inch one means losing some working space. In one of the setups the solution was simple: adding a second monitor next to the iMac. That makes the arrangement slightly less minimal than the classic all-in-one idea, but it restores the practical working area. One screen can remain focused on the main task, while the other holds reference material, terminals, chats or documents. In practice it turns out to be a comfortable compromise between elegance and usability. ## Communication as a first-class feature One of the most noticeable improvements is the camera. The previous iMac’s webcam was poor enough that an external camera had been installed. The new model includes a significantly better camera with automatic framing and much stronger image quality. That improvement reflects another structural change: computers are now used continuously for **video communication**. Cameras are no longer peripheral components but central ones. 🎥 The new iMac includes ****Center Stage**, Apple’s intelligent camera framing feature. Using an ultra-wide sensor and on-device machine learning, the camera quietly reframes the image during video calls so you stay centred even if you move. Instead of a fixed webcam view, the system dynamically pans and crops the image, and it can widen the frame if another person joins. The result is a more natural presence on calls, as if a small camera operator is subtly keeping the conversation in focus. ## A lighter machine in a larger system None of these changes are specific to Apple, although the iMac makes them particularly visible. The physical computer on the desk is becoming thinner, quieter and easier to replace. At the same time, more of the actual computing environment lives elsewhere in networks, services and AI systems. In that sense the machine itself is shrinking, even as the computing world around it continues to expand. ### Further reading - [Apple’s Pages, Numbers and Keynote: Strong Tools, Weak Culture](https://rob-hoeijmakers.ghost.io/apples-pages-numbers-and-keynote-strong-tools-weak-culture/) - [Use Center Stage to keep you centered in the camera frame - Apple Support](https://support.apple.com/en-us/111102) ### The Delegation Problem of the Internet URL: https://rob-hoeijmakers.ghost.io/the-delegation-problem-of-the-internet/ Last updated: 2026-06-03T11:45:56.000Z But a surprising amount of modern work is not limited by knowledge, tools, or intelligence. It is limited by identity. The small rituals of logging in, confirming codes, verifying accounts, and proving, **again and again**, that you are you. ## Quick takeaways - Many digital systems assume that **only the account holder can act**. - Real life, however, runs on **delegation**: partners, colleagues, assistants, children, caregivers. - Two-factor authentication and identity checks make delegation difficult or impossible. - The result is a hidden productivity drain: **identity management work**. - The same constraint also limits what AI agents can actually do. ## The delegation gap In ordinary life, people constantly act on behalf of others. A partner renews insurance. A child handles paperwork for parents. An assistant arranges travel or contracts. A colleague submits documents. This is not exceptional behaviour. It is how daily life functions. Yet most digital systems assume something else entirely: one account, one person, one operator. Security systems reinforce this assumption. Passwords, two-factor authentication, biometrics, device verification. Each action requires the account holder to appear and confirm themselves. This works well for security. But it quietly breaks something else: **delegation**. ## Identity work Because of this design, a surprising part of modern life is spent performing small identity rituals. Logging in. Entering codes. Approving notifications. Verifying devices. These steps are individually trivial. Collectively they form a kind of invisible labour. You might call it **identity work**. It is the administrative layer that sits between intention and action. ## The AI paradox This constraint also explains something curious about AI. AI systems are increasingly capable of planning, analysing, and organising tasks. In principle they could handle many everyday administrative jobs. But they usually cannot. Not because they lack intelligence. Because they lack **authorised identity**. The real systems where action happens, *banks, government portals, insurers, utilities*, are locked behind personal authentication. So the human remains the final operator, approving and executing steps that machines could otherwise handle. ## When delegation becomes administration Some systems attempt to support delegation through formal mandates or authorisations. But these often turn delegation itself into another administrative task: registering permissions, managing expiry dates, renewing access codes. The effort required to delegate can become almost as large as the task being delegated. ## A missing piece of digital infrastructure What seems largely absent from today’s digital world is a simple capability: **safe, temporary, and limited delegation of digital authority.** The ability to say: - this person may manage this account for a period - these actions are allowed, others are not - access can be revoked at any time Corporate IT systems have such mechanisms. Everyday digital services rarely do. ## Closing reflection For decades we have focused on making information easier to access and process. But productivity may increasingly depend on something more mundane: How easily we can **act on behalf of one another** in digital systems. Real life is cooperative. The internet, in many places, still assumes **we operate alone**. --- ### Further reading - [The session unlocks the door. For everyone. ](https://rob-hoeijmakers.ghost.io/session-cookie-authentication-security/) ### A Personal Retention Strategy for ChatGPT URL: https://rob-hoeijmakers.ghost.io/a-personal-retention-strategy-for-chatgpt/ Last updated: 2026-05-10T08:53:37.000Z I work daily with large language models, and with ChatGPT in particular. It has become part of how I think, draft, compare perspectives, and test ideas. But there is a side effect: it accumulates. Conversations remain visible. Threads multiply. The system keeps a trail of exploratory thinking that was never meant to be permanent. Recently, I decided to approach that accumulation more deliberately. Not as a technical problem, but as a matter of habit. ## Quick takeaways - Data accumulation has psychological effects, even when storage is cheap. - ChatGPT offers limited middle ground between “keep everything” and “delete everything”. - Periodic, deliberate clearing works better for me than continuous trimming. - Cognitive hygiene is part of responsible AI use. - Personal practice shapes how we later think about organisational data governance. ## How ChatGPT handles conversations In practical terms, the options are relatively simple. Conversations are stored by default. You can archive them, but that mainly changes their visibility. They still exist in the account environment. There is also a more radical option: deleting chats, or deleting all chats. That removes them from the interface and reduces what remains available for continuity across threads. In the Plus environment, there is some cross-thread awareness and temporary memory. If you clear everything, you also lose that contextual continuity. In the Business environment, memory behaves differently, but the structural choice is similar: retain broadly, or remove decisively. There is no refined “delete everything older than three months” button. No nuanced retention schedule. It is either accumulation by default, or active intervention. That simplicity forces a decision. ## The psychology My instinct to save everything did not come out of nowhere. It was shaped in an earlier phase of the internet. There was a time when storage was limited. Hard drives were small. Backups were manual. An export could genuinely save your work, your correspondence, your history. Losing data meant real loss. So you learned to preserve. To copy. To archive. To keep local control. That reflex made sense then. It was protective. The environment has changed completely. Storage is abundant. Cloud systems retain by default. Platforms are designed to accumulate. Conversations, drafts, metadata, versions, everything remains somewhere. Paradoxically, scarcity trained us to save carefully. Abundance now tempts us to save indiscriminately. I noticed that I was still operating with the old reflex in a new context. Keeping everything no longer protected me from loss. It simply increased volume. Letting go, in an era of automatic retention, feels counterintuitive. Deleting something that is effortlessly stored can feel almost irresponsible. And yet, for me, it has become the more disciplined act. Not because data is dangerous in itself, but because unlimited retention creates a diffuse environment. The value of a conversation often lies in what it enabled, not in its permanent storage. Once that became clear, the old reflex began to loosen. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/02/ChatGPT-2026-02-26-10.11.49.png) ## Hygiene and rhythm I experimented with gradual pruning. It kept the topic alive in my mind. There was always something to clean up. What works better for me is a tidal approach. A period of active use. Then a deliberate session of consolidation and clearing. Before deleting, I sometimes export or summarise what still matters. Often, I realise that very little truly needs to be retained. The core ideas have already moved into notes, documents, or articles. Pressing delete is slightly uncomfortable. There is always a moment of doubt. But the effect afterwards is consistent: the interface feels lighter. The tool feels more like a workspace and less like an archive. ## The business implication This is not only personal preference. In professional contexts, we discuss AI governance, retention policies, compliance, and risk. Those discussions are necessary. But they remain abstract if not grounded in practice. If we treat conversational AI as a permanent cognitive archive, we implicitly accept indefinite retention of exploratory thinking. If we treat it as a workspace, we introduce rhythm, boundaries, and completion. For me, the choice to clean out regularly is aligned with how I now treat data more broadly. Keep what has structural value. Remove what has served its purpose. That stance affects how I think about organisational data as well. Stewardship begins with habit. ## Closing I no longer aim to preserve every intermediate step of thinking. ChatGPT is a powerful amplifier. But amplification does not require permanence. Periodic reset is not a rejection of the tool. It is a way of keeping it usable, proportionate, and aligned with how I want to work. --- ### Further reading - [ChatGPT Memory Has Quietly Changed (Plus & Pro)](https://rob-hoeijmakers.ghost.io/chatgpt-memory-has-quietly-changed-plus-pro/) - [Anki, Human Memory in a Digital World](https://rob-hoeijmakers.ghost.io/human-memory-in-a-digital-world/) ### Apple’s Pages, Numbers and Keynote: Strong Tools, Weak Culture URL: https://rob-hoeijmakers.ghost.io/apples-pages-numbers-and-keynote-strong-tools-weak-culture/ Last updated: 2026-05-10T08:53:37.000Z ## Quick takeaways - Apple’s productivity apps are technically strong but socially marginal - The iWork identity has faded into a broader creative subscription layer - Core features remain free, premium features signal a freemium shift - Microsoft and Google built institutional gravity; Apple did not - Productivity tools become dominant through shared culture, not design alone ## A quiet repositioning Pages, Numbers and Keynote have always lived quietly inside the Apple ecosystem. Free, preinstalled, stable, and tightly integrated across devices. Recently, Apple refreshed them visually and repositioned them inside its broader creative subscription layer. The old “**iWork**” framing has largely disappeared. In its place sits a more expansive, creative umbrella. At the same time, subtle freemium dynamics have emerged. Core functionality remains free, but premium templates, content and AI-assisted features sit behind subscription tiers. This is not a radical shift. It is a tonal one. And tone matters. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/02/IE-2026-01-14T145949.850-1315206661.jpg) Apple Creator Studio ## Strong tools, thin culture Technically, these apps are not weak. Pages handles most document work with calm elegance. Keynote remains one of the most fluid presentation tools available. Numbers is unconventional but capable. If you are already fully inside the Apple ecosystem, they feel coherent. Sovereign even. But professionally, they remain marginal. No one says, “We are a Pages-based organisation.” The difference is not capability. It is cultural gravity. ## Lock-in without institutional weight Microsoft 365 became dominant because organisations standardised on it. [Google Workspace](https://rob-hoeijmakers.ghost.io/google-gemini-workspace-future/) reshaped collaboration norms in the browser. They built certification tracks, compliance layers, administrative tooling, shared templates, training ecosystems. [Excel](https://rob-hoeijmakers.ghost.io/excel-hidden-operating-system-of-business-reasoning/) became a language. Docs became a habit. Apple did not build that institutional layer around its productivity suite. The lock-in exists at the hardware and ecosystem level. It does not extend into organisational culture. ## The freemium signal The move toward a broader creative subscription model reinforces this positioning. Freemium works well when tools are personal, creative and modular. It works less well when the goal is institutional standardisation. By placing Pages, Numbers and Keynote inside a creative services story, Apple implicitly signals that these are complements to a creative ecosystem, not competitors in enterprise infrastructure. That may be strategic clarity rather than weakness. 💡 What is ****Apple Creator Studio**? [Apple Creator Studio](https://www.apple.com/apple-creator-studio/) is Apple’s broader creative subscription layer that bundles professional tools such as Final Cut Pro, Logic Pro and Pixelmator Pro, while Pages, Numbers and Keynote remain free at their core; certain premium templates, content libraries and AI-assisted features are available within the paid tier, and the older “iWork” branding has largely faded from public positioning, reframing these productivity apps not as a standalone office suite but as part of a wider creative ecosystem. ## The missing identity Apple excels at product identity. The Mac has identity. The iPhone has identity. Final Cut has identity among creators. Pages and Numbers do not. The iWork label once gave them at least a collective name. With that fading, they risk becoming functional utilities rather than cultural artefacts. There is no manifesto for them. No visible professional tribe. No shared productivity philosophy attached to them. Without identity, even strong tools remain peripheral. ## A personal tension There is a quiet temptation to consolidate everything inside the Apple ecosystem. It feels aesthetically coherent and technically integrated. But collaboration reintroduces gravity. Where do your clients work? Where do your peers exchange templates? Where is collective knowledge compounding? That is where culture forms. And productivity without culture rarely becomes dominant. --- ### Further reading - [Apple Creator Studio](https://www.apple.com/apple-creator-studio/) ### Vision, Judgement, Creativity: Reclaiming Agency in the Age of AI URL: https://rob-hoeijmakers.ghost.io/vision-judgement-creativity-reclaiming-agency-in-the-age-of-ai/ Last updated: 2026-05-26T07:34:58.000Z Recently I was struck by a simple triad. In a knowledge world increasingly shaped by artificial intelligence, three human capacities remain decisive: **vision, judgement, creativity**. At first glance it sounds reassuring. Of course those matter. They always have. But the more I sat with it, the more uncomfortable it became. Vision, in particular, is easy to praise and difficult to practise. I realised that much of my own thinking about AI and the future had become analytical to the point of inertia. I could map trends. I could outline structural forces. I could explain capital dynamics, institutional pressures, automation curves. And yet, somewhere along the line, analysis began to crowd out orientation. ## The trap of analysis It is not hard to see why. The current landscape is dense. AI systems are advancing rapidly. Knowledge work is being partially automated. Capital appears to scale faster than labour. Geopolitics feels brittle. Education is under pressure to justify its value beyond information transfer. Viewed purely through analysis, the five-year horizon easily tilts toward unease. The more variables one includes, the more indeterminate the picture becomes. It is tempting to oscillate between hype and doom, between technological optimism and institutional anxiety. Both share the same flaw. They treat the future as something that happens to us. ## The pivot The shift for me came almost embarrassingly simply. Instead of asking only, *What will the future bring?* I began asking, *How would I like the future to be?* That question does not deny structural forces. AI capabilities will expand. Incentive systems will shape behaviour. Markets will reward certain configurations and penalise others. These are conditions. But conditions are not destiny. The moment preference enters the frame, something changes. Fear loses some of its hold. The conversation moves from prediction to design. Vision, then, is not clairvoyance. It is the articulation of what we want to preserve, cultivate, and build under new technological circumstances. Judgement becomes the ethical filter. What do we protect when efficiency tempts us to cut corners? Where do we insist on human responsibility even when systems can approximate decisions? Creativity becomes institutional. How do we redesign firms, incentives, educational pathways, ownership models so that they reflect our preferences rather than drift along default settings? ## Five years, concretely A five-year horizon is short enough to be tangible and long enough to require intention. Within five years: - AI will be embedded more deeply in professional workflows. - Certain forms of [knowledge work will shrink](https://rob-hoeijmakers.ghost.io/mapping-ai-exposure-across-the-dutch-labour-market/), others will expand. - The relationship between capital and labour will be renegotiated in subtle ways. - Education will be forced to clarify whether it transfers information or cultivates judgement. None of this is apocalyptic. None of it is trivial. The real question is whether we approach this period as [spectators or as architects](https://rob-hoeijmakers.ghost.io/agents-and-agency/). At an individual level, that means asking: - How do I maintain [depth of judgement when tools generate plausible answers instantly](https://rob-hoeijmakers.ghost.io/not-doing-is-also-an-act/)? - How do I train attention and discernment rather than mere prompt fluency? - How do I remain morally accountable in hybrid human-machine decisions? At a collective level, it means asking: - What incentive structures do we reward? - What forms of ownership and governance do we encourage? - How do we prevent systemic fragility while embracing capability? These are not abstract policy debates. They are design questions embedded in companies, schools, professional communities, and families. ## Learning from others One further realisation followed. I do not have to invent a vision in isolation. There are thinkers, entrepreneurs, economists and technologists who are already sketching futures. Some are optimistic, some sceptical, some institutionally focused, some market-driven. Engaging with their visions is not about adopting them wholesale. It is about sharpening one’s own judgement. Vision is rarely born in a vacuum. It matures through dialogue, comparison and friction. ## Where this leaves me I do not yet have a fully articulated blueprint for the next five years. The design work remains to be done. What has changed is my stance. I no longer try to forecast the future as if it were weather moving in from the horizon. I try to clarify what I would like to see stabilised and what I am willing to help build within evolving technological conditions. That shift alone moves me out of analysis paralysis and back into authorship. This is not a finished vision. It is a change in direction. The work of articulating preferences, testing them against reality and translating them into institutional design lies ahead. But perhaps that is precisely the point. In a world of accelerating tools, vision is not about certainty. Judgement is not about omniscience. Creativity is not about novelty for its own sake. They are practices. And for now, it is enough to begin **practising them deliberately**. ### The Persistence of the Table URL: https://rob-hoeijmakers.ghost.io/the-persistence-of-the-table/ Last updated: 2026-02-19T07:45:31.000Z Things keep happening. Events accumulate. Transactions, measurements, messages, decisions, states. Time does not just pass; it leaves residue. And once there is residue, there is a need to hold it, inspect it, and make sense of it without having to relive it. One of the oldest and most reliable ways we do this is by putting time into rows. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/02/tabular-tablet.jpg) Tablet ## Time made inspectable A table does something deceptively simple. It takes succession and turns it into adjacency. What happened earlier and what happened later are no longer separated by time, but placed next to each other in space. Events become rows. Time becomes order. This move is so familiar that it often goes unnoticed. But it is foundational. Without it, accumulation would remain narrative. You would have to move through time again and again to understand what has happened. With it, time can be paused, flattened, and held in view. A table is not primarily about data. It is about making progress inspectable. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/02/image.png) ## Accumulation needs a surface As soon as things accumulate, a surface becomes necessary. Ledgers, logs, registers, inventories, timetables. Across history, tables appear wherever continuity matters. Not to tell stories, but to allow comparison. To see what is growing, repeating, deviating. What matters is not that the information is structured, but that it is *laid out*. Laid out so it can be seen all at once, or at least in parts, without traversing the whole sequence again. This is why tables feel less like documents and more like workspaces. They are places where time has been temporarily stabilised. ## Glanceability as a human response Once time has been spatialised, a particular human mode becomes possible: glancing. Glancing is often treated as superficial. In practice, it is how humans cope with accumulation. It allows orientation without immersion. Pattern recognition without commitment. A sense of “enough” without completion. Tables support this unusually well. You can skim, pause, return. You can look briefly and still learn something reliable. You do not have to resolve meaning to extract signal. In this sense, glanceability is not a design feature. It is the human complement to temporal compression. ## Why tables keep resurfacing in software Much of modern software deals with accumulation. More data, more events, more state over time. Interfaces change. Visualisations come and go. But when understanding becomes difficult, systems tend to expose a table. An admin view. An export. A debug panel. A list of rows ordered by time. This is not a failure of imagination. It is a return to a form that reliably turns ongoing process into something that can be inspected. Spreadsheets and databases sit on opposite sides of this tendency, but share the same core move. Both are ways of holding accumulated time in a stable, glanceable form. They differ in governance and scale, not in essence. ## Feeds as disguised tables This becomes especially visible when looking at social media feeds. A feed feels like flow. Like presence. Like “now”. But structurally, it is a table. Each post is a row. Time is the primary ordering key. Metadata forms hidden columns. Filtering, ranking, and pagination are continuous table operations. The experience of “keeping up” does not come from reading everything. It comes from being able to glance at accumulated time and decide that nothing essential was missed. What looks like narrative immediacy is already ledger logic. 💡 Sidenote: A ****database server and a database file** invite very different ways of thinking. A server assumes continuity, coordination, and operational overhead; it needs to be running, managed, and integrated. A database file, as with SQLite, collapses all of that into something you can open, move, inspect, and reason about directly. I felt that difference again very concretely when rewriting an application with Claude Code: working against a file sharpened the feedback loop, reduced incidental complexity, and kept the data close to the code and the questions. I used [Datasette](https://datasette.io) to visualize the data to tables. ## Machines and linear time Large language models sharpen this contrast. They read linearly. When they encounter tables, they do not perceive a surface. They process sequences. JSON, YAML, CSV are not spatial to them. They are linear encodings of structure. What is glanceable to a human must be traversed again by a machine. This does not diminish either side. It clarifies roles. Tables are not primarily for machines. They are for humans who need to hold time still long enough to think. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/02/image-1.png) Table in linear format, JSON. ## Something that does not go away Software will keep changing. Interfaces will evolve. Data volumes will grow. But as long as events accumulate, there will be a need to turn time into space. And as long as humans remain bounded in attention, that space will need to be glanceable. That is why tables persist. Not as a legacy form, but as a recurring answer to a permanent problem. How to live with accumulation without being overwhelmed by it. For now, this is simply something worth noticing. Not a conclusion, but a grounding. A reminder that amid constant change, some forms endure because the pressures they respond to endure as well. ### From text to diagrams: working with Mermaid URL: https://rob-hoeijmakers.ghost.io/from-text-to-diagrams-working-with-mermaid/ Last updated: 2026-05-10T08:53:37.000Z There is something interesting about turning text into diagrams. Not illustrations, but diagrams that express structure: flows, dependencies, sequences. This becomes especially noticeable when the text itself is not prose, but something closer to a script. Ordered sentences that describe relations rather than tell a story. That is roughly the space where Mermaid sits. ## Text as structure, not as explanation Mermaid is usually introduced as a way to generate diagrams from text. That description is accurate, but incomplete. What matters more is that the text is the primary artefact. You are not drawing shapes. You are describing relationships. The diagram is a rendering of that description, not the other way around. Because of that, the process is reversible. You can always go back to the text, adjust it, and regenerate the diagram. Nothing is lost in translation. This is different from most visual diagramming tools, where structure slowly dissolves into manual layout decisions. Once you start dragging boxes around, the model lives in your head rather than in the artefact. Mermaid keeps the structure explicit. 💡 ****What is Mermaid?** Mermaid is an open-source, text-based diagramming language created around 2014 by ****Knut Sveidqvist**. It allows you to describe diagrams such as flows, sequences, and state machines using plain text, which can then be rendered into visual diagrams. Mermaid is not a formal standard governed by a standards body, but it has become a widely adopted de facto standard in documentation and developer tooling, much like Markdown. Its focus is on expressing structure rather than visual design, which makes diagrams easy to edit, version, and regenerate. This text-first approach is also why Mermaid works particularly well with modern language models, which can read, generate, and modify Mermaid syntax directly. ## Low resolution, by design I tend to think of Mermaid in the same category as [Markdown](https://rob-hoeijmakers.ghost.io/markdown-google-docs/). Both are intentionally low-resolution formats. They do not try to look good. They try to stay readable, portable, and precise. You can open them in any editor, version them, diff them, or move them between systems without friction. That constraint is not a drawback. It is what makes them durable. The result is usually good enough visually, but strong in terms of meaning. For thinking, that trade-off is often exactly right. ## What changes with language models What makes Mermaid more relevant now than it used to be is that large language models understand it natively. They can generate Mermaid diagrams from descriptions, modify existing ones, or explain what a diagram represents. That lowers the threshold for working with structured diagrams, especially when you are still figuring things out. You can start with a rough explanation, ask for a diagram, adjust the structure in natural language, and iterate. The diagram becomes part of the thinking process rather than a final documentation step. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/01/Mermaid-Chart-in-ChatGPT.jpg) Mermaid Chart GPT in ChatGPT ## A tighter feedback loop If you connect Mermaid generation into your workflow, for example via an MCP setup or similar tooling, that loop tightens further. You explain something. You get a diagram. You adjust the structure. You regenerate. No redrawing. No exporting. No loss of intent. This is particularly useful for architecture sketches, process descriptions, or system overviews that are still in flux. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/01/Mermaid-Chart-in-Claude.jpg) Mermaid integration into Claude ## Why it works for me Mermaid is not a general-purpose diagramming tool, and it does not try to be one. It assumes that structure matters more than appearance, and that text is a reasonable place to express that structure. If you are already comfortable working in Markdown, Mermaid tends to feel familiar rather than foreign. And if you combine it with a language model, it becomes a practical way to move back and forth between explanation, structure, and visualisation with very little overhead. Not a revolution. Just a tool that fits well with text-first ways of thinking. --- I personally have a paid plan for this service: [MermaidCreate diagrams and visualizations using text and code.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon-41.ico)Try Editor![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/mermaid-logo-horizontal-1.svg)](https://mermaid.js.org) Het is a flowchart example. Textual script on the left, graphical output on the right. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/01/Mermaid-Graph---FlowChart-example.jpg) ### Further reading - [From Ideas to Precision: Why I Use Mermaid with LLMs](https://rob-hoeijmakers.ghost.io/mermaid/) - [GPT Image 2 Changed My Mind on AI Visuals](https://rob-hoeijmakers.ghost.io/gpt-image-2-changed-my-mind-on-ai-visuals/) ### PersonaPlex marks a shift from dictation to conversation URL: https://rob-hoeijmakers.ghost.io/personaplex-marks-a-shift-from-dictation-to-conversation/ Last updated: 2026-05-10T08:53:37.000Z I have written quite a bit about speech recognition, dictation, and voice over the past years. Looking back, I now see that I was often talking about different things under the same label. [PersonaPlex](https://research.nvidia.com/labs/adlr/personaplex/) is a research model from NVIDIA that explores native, real-time speech-to-speech conversation, where listening and speaking happen at the same time. It helps make a distinction visible that had been forming for a while already. Not because earlier systems were wrong or outdated, but because *what we mean by voice* has started to change. 💡 ****PersonaPlex, in short:** PersonaPlex is a research model that treats speech as a continuous, real-time process rather than as input to be processed in turns. It listens and speaks at the same time, meaning interruptions, timing, and tone directly shape what the system does next. That makes it fundamentally about conversation, not dictation. ## Dictation solved one problem Models such as [Whisper](https://rob-hoeijmakers.ghost.io/openais-whisper-speech-to-text-technology/) are excellent at dictation. You speak. The system listens. You get clean, reliable text. That alone was a major step. It removed friction between thinking and writing. It made meetings, interviews, and spoken notes usable at scale. For this purpose, dictation models are still extremely strong. But dictation treats speech as something that already happened. Accuracy matters more than timing. [Speech recognition, Speech-to-Text, Dictation, and Transcription: What’s the Difference?Just like picking the right writing tool, choosing the right speech technology makes all the difference—here’s how to decide.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-506.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_4550-5.jpeg)](https://rob-hoeijmakers.ghost.io/asr-speech-to-text-dictation-and-transcription-whats-the-difference/) ## Conversation already feels solved, doesn’t it? If you use ChatGPT with voice, or Gemini Live, it is easy to think: *this already is full duplex*. You can interrupt them. They stop speaking immediately. The interaction feels fluid compared to older voice assistants. From a user’s perspective, that intuition makes sense. But under the hood, something else is going on. ## How today’s voice systems actually work Most production voice systems today rely on several fast components working together: - One part listens for speech and detects interruptions. - Another part reasons about what to say. - Another part turns that into audio. When you interrupt, a very fast detector notices this and simply cuts off the audio output. The system stops talking right away, even if another part is still finishing its thought elsewhere. To you, it feels like the system was listening while speaking. Technically, it mostly just stopped speaking very quickly. This is not a flaw. It is a sensible engineering choice. It is cheaper, more stable, and easier to control. But it is not yet what researchers mean by **full duplex**. ## What “full duplex” really means Full duplex simply means *listening and speaking at the same time*. Not taking turns. Not stopping first. Not restarting after an interruption. In a full-duplex speech system, incoming sound continues to shape what the system is doing while it is already talking. Interruptions are not just stop signals. They carry information: timing, tone, urgency. Speech is no longer just an interface around reasoning. It becomes part of the reasoning itself. That is the real shift. ## Why this matters Seen in that light, PersonaPlex is not just another voice demo. It is a concrete example of what changes when speech is no longer treated as input that must first be stabilised, but as a medium in which interaction itself takes place. Dictation systems listen, then act. Most current voice assistants react quickly. PersonaPlex listens *while* it speaks. That distinction may sound subtle, but it changes what kinds of conversations become possible. Especially outside the assistant setting: in phone calls, service conversations, and other situations where flow, interruption, and timing matter. PersonaPlex does not replace dictation models, nor does it invalidate today’s voice assistants. It shows that voice is becoming layered. And it gives a first, working glimpse of what that next layer looks like. ### Further reading - [NVIDIA PersonaPlex: Natural Conversational AI With Any Role and Voice](https://research.nvidia.com/labs/adlr/personaplex/) - [From Typing to Talking](https://rob-hoeijmakers.ghost.io/from-typing-to-talking/) ### One European Company (EU-INC) URL: https://rob-hoeijmakers.ghost.io/one-european-company-eu-inc/ Last updated: 2026-05-10T08:53:38.000Z In November, when I wrote about [Europe’s AI future](https://rob-hoeijmakers.ghost.io/one-european-company-eu-inc/), I described ideas like “[EU Inc.](https://www.eu-inc.org)” as early. Interesting, but distant. The kind of structural reform that makes sense on paper, yet tends to move slowly once it enters the European institutional machinery. Fast forward a few months, and the tone has shifted. At Davos, Ursula von der Leyen spoke openly about competitiveness, scale, and Europe’s position in a more confrontational geopolitical world. Around the same time, the European Parliament voted overwhelmingly in favour of a proposal that has been circulating for years: a so-called “28th regime”, an optional European company structure, often referred to as “EU Inc.” The headlines suggest momentum. Some people sense a turning point. Others respond with a shrug: *we’ve seen this before*. Both reactions are understandable. [EU–INC — One Europe. One Standard. — Pan-European legal entity.EU–INC is a proposal for a pan-European standardized legal entity to unlock pan-European startup scaling.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/hTcgaWRewUNZeB6wVrfBSXk9Mn0-1.svg)![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/DywwAp236AdnECpow4p9VcGluY-1.png)](https://www.eu-inc.org/) ## What actually happened? The recent vote in the European Parliament did not create a new European company overnight. It was not a law. It was a strong, cross-party signal: a request to the European Commission to come forward with a concrete proposal for a unified, optional company regime that would sit alongside the 27 national systems. In practical terms, Parliament is asking for a legal form that allows startups and scale-ups to operate across Europe without constantly translating themselves into different national legal languages. A company that is European by default, rather than Dutch first, German second, French third. The Commission is expected to present a proposal in the coming months. After that, the long part begins: negotiations with member states, legal design, compromises. This is why nothing has “changed” yet in the lived reality of entrepreneurs. But something has shifted in tone and alignment. ## “But I founded my company just fine” If you have incorporated a company in the Netherlands, Estonia, or a handful of other countries, you might reasonably ask what the problem is. Setting up a business can already be fast, digital, and inexpensive. I had the same experience. And that is precisely the point. The issue EU Inc. is trying to address is not the moment of incorporation. It is what happens *after*, when a company starts to grow across borders. Selling across Europe from one entity is often straightforward. Hiring people, opening offices, offering equity, raising capital, merging, or relocating headquarters is where friction accumulates. Each additional country brings its own company law, governance logic, equity rules, and legal expectations. A startup active in three or four European countries is already navigating multiple legal universes. Investors know this. Founders feel it. Over time, it shapes where companies choose to scale, or where they choose to exit. EU Inc. is an attempt to reduce that structural friction, not to erase national differences. [Signing the Future: How a Digital Incorporation Made eIDAS RealA first-hand look at how eIDAS and digital signatures replace paper rituals with cryptographic trust.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-505.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_0188-2.jpeg)](https://rob-hoeijmakers.ghost.io/digital-signature-eidas-experience/) ## What EU Inc. is, and what it is not The proposed “28th regime” would be optional. National company forms would remain. Taxes would remain national. Labour law would remain national. This is not a leap towards a European super-state. It is better understood as legal infrastructure: a shared base layer for companies that intend to operate across borders, with standardised rules for governance, shares, and cross-border activity. Think less “harmonising everything” and more “removing unnecessary translation costs”. That distinction matters, because it explains both the appeal and the resistance. ## Why scepticism persists If this idea sounds familiar, that is because it is. Variants of a European company regime have been discussed for decades. Previous attempts stalled in the face of national sensitivities, fears of regulatory arbitrage, and concerns about labour protections. Those concerns are still present. Trade unions worry about a race to the bottom. Some governments fear losing control over company law. Others question whether a new regime is necessary, or whether existing frameworks could be improved instead. Even the name has become political. “EU Inc.” sounds efficient and global, but also American. In Parliament, some prefer a neutral Latin label to avoid the symbolism altogether. These are not side issues. They are signals of how deeply company law touches questions of sovereignty, social models, and identity. ## So why does this feel different now? Not because Europe has suddenly become fast. But because pressures have aligned. Europe has struggled to produce globally scaled tech companies. Capital markets remain fragmented. Strategic dependence on the US and China feels riskier than it did a decade ago. And geopolitical tensions are no longer abstract background noise. In that context, ideas that once felt optional start to feel necessary. Not urgent in the sense of “done by next year”, but urgent in the sense of “no longer ignorable”. EU Inc. benefits from that shift. It is being reframed less as a market experiment and more as economic infrastructure. ## Holding two thoughts at once It is possible to say all of the following without contradiction: - This is a meaningful proposal that addresses a real structural problem. - It will not be implemented quickly. - It will involve compromises that disappoint almost everyone. - It still matters whether Europe even tries. If EU Inc. eventually materialises, it will not be because of a Davos speech. It will be because member states accept that a single market without scalable legal structures remains incomplete. In November, calling this “early” was accurate. Today, the world feels more compressed, more pressured. The timeline has not magically shortened, but the stakes have become clearer. That alone is already a change. ### Further reading - [Special Address by President von der Leyen at the World Economic Forum](https://ec.europa.eu/commission/presscorner/detail/en/speech%5F26%5F150) - [EU Inc.: between words and reality in Europe’s new company proposal](https://rob-hoeijmakers.ghost.io/eu-inc-europe-company/) ### What coding with AI feels like now URL: https://rob-hoeijmakers.ghost.io/what-coding-with-ai-feels-like-now/ Last updated: 2026-05-10T08:53:38.000Z ## Notes from the edge of building with LLMs I am all in on large language models as cognitive amplifiers. I use them to think, to structure, to explore, to write. That part feels natural to me. What I have never been particularly drawn to is using them as coding tools. Programming has always felt like a different craft. Adjacent, interesting, but not where I naturally place my attention. And yet, over the past months, that boundary has become harder to ignore. Not because I suddenly wanted to become a developer, but because clients started asking different questions. They see headlines about “AI eating code”. They see demos racing past on LinkedIn. And they wonder whether they are missing something important. So I decided to explore. Not to reskill, but to orient myself. To understand what is actually changing, where agency now sits, and what is realistically accessible to people who are curious, but not intent on becoming software engineers. What follows is not a survey. It is a short field note, guided by my own steps. ## Standing near the edge of coding I have never really been a coder. I know enough to follow along, to collaborate, to get something running from [GitHub](https://github.com/rhoeijmakers) if I have to, but it has always felt like a different craft. Interesting, sometimes impressive, but not where I naturally operate. You may recognise that position. Curious, not allergic to technology, but not inclined to disappear into documentation or frameworks either. What surprised me in these experiments was not how powerful the tools were, but how quietly the boundary shifted. Not because I learned a new skill, but because the effort moved. An idea could turn into something that actually works. Not a mock-up, not a diagram, but a usable thing. That was new to me. Not spectacular, but disorienting in a subtle way. ## First encounter: describing a thing into existence The first thing I tried was [Lovable](https://lovable.dev). Someone [shared a prompt](https://www.linkedin.com/posts/cassombroek%5Fbij-de-meeste-bedrijven-zijn-e-mailhandtekeningen-activity-7418204193548562432-Gv62?utm%5Fsource=share&utm%5Fmedium=member%5Fdesktop&rcm=ACoAAAAGPccBqY4Rvlcj2zZ2PdvOk7tl5dQrkUs) for building a simple email signature generator. I copied it, adjusted it slightly, and watched the result appear. A working interface. Inputs, preview, export. No setup, no explicit architectural decisions on my side. What struck me was how familiar this felt. This was not programming in the traditional sense. It felt much closer to product definition. Describing behaviour, constraints, and outcomes. The kind of thinking that normally stops at wireframes or tickets. Here, the description executed. That is the strength of tools like this. They let you stay at the level of intent and outcome. At the same time, the trade-off is clear. Many decisions are made for you. Hosting, structure, deployment, all of that is implicit. You gain speed and clarity, but give up control. For an experiment, that was exactly right. [Lovable AppLovable Generated Project![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon-40.ico)Lovable![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/id-preview-7f5b0dcc--bb3425a8-ad34-420a-81f6-0c4ac8723ba9.lovable.app-1768640349762.png)](https://brand-signature-scribe.lovable.app/) ## A step deeper: intent meets a project Curiosity pulled me one layer down. I wanted to see what would happen if I stayed close to intent, but entered a recognisable software project. I tried [Claude Code](https://www.anthropic.com/engineering/claude-code-best-practices) and connected it to GitHub. This time, the result was not an abstract app, but a repository. Files, versions, commits. I used it to build a bespoke email signature generator inside Google Workspace, backed by a Google Sheet and implemented with Google Apps Script. The experience changed in kind. Claude Code did not “build an app for me”. It helped me think through a project. What goes where. What changes when I adjust something. How to iterate, deploy, and improve. The LLM was no longer replacing structure. It was inhabiting it with me. I also felt a boundary. Systematic software development, long-term maintenance, deeper architectural decisions, that still felt like a different craft. Not inaccessible, but distinct. What mattered was that this boundary was now visible. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/01/image.png) The signature app in Google Workspace. ## What people mean when they say “IDE” This is where the term “IDE” often comes in. IDE stands for Integrated Development Environment, but that label tends to obscure more than it explains. In practice, an IDE is the place where software lives while it is being made and maintained. It is where files are organised, changes are tracked, versions are compared, and decisions accumulate over time. Not just a text editor, but a working environment with memory. A widely known example is [Visual Studio Code](https://code.visualstudio.com), used by many developers and familiar to a broader technical audience. What is changing now is not the existence of such environments, but their role. With LLMs embedded, IDEs start to absorb more of the thinking, explanation, and iteration that used to happen outside them. They begin to look less like specialist tools and more like shared workspaces between human and system. ## Three ways of building, loosely speaking Looking back, what helped me make sense of this was not taxonomy for its own sake, but orientation. Roughly speaking, I now see three modes people move between. First, **outcome-driven tools like Lovable**, where you stay at the level of intent and receive a working artefact. Second, **artifact-centric assistance**, where LLMs help you produce scripts, components, or configurations that you then integrate yourself. Third, **IDE-centred environments**, where the LLM is embedded into the development workflow itself and develops a persistent understanding of a system over time. Tools such as [Cursor](https://cursor.com) are often mentioned in this context, as are [Codex](https://openai.com/codex/)\-based systems from OpenAI. These are not strict categories. They overlap and evolve. But they differ in one crucial respect: where your agency sits. At intent. At artefacts. Or at systems. What has changed is that moving between these layers no longer requires a hard identity shift. ## A signal worth paying attention to This is also why it is telling to see Google presenting tools like [Antigravity](https://antigravity.google) explicitly as a “next-generation IDE”. Even if such tools are still emerging, the framing itself matters. It suggests that IDEs are no longer seen only as specialist developer tools, but as environments where complex work takes shape, especially when that work results in living systems rather than finished documents. That matters even if you never plan to work there yourself. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/01/image-1.png) ## Abundance, noise, and judgement One effect of this shift deserves a little more attention. As software becomes easier to create and easier to discard, the market around it changes shape. We move away from a relatively structured landscape, with high entry costs and clear thresholds, towards something more crowded and informal. A place with many small offerings, many claims, many tools competing for attention. At times, it starts to resemble a medina: lively, inventive, and full of possibility, but also noisy and disorienting. In such an environment, the hard part is no longer building something. The hard part is knowing what needs to be built, what can be thrown away, and what deserves care, continuity, and quality. Not every problem needs a robust system. Not every tool should live beyond its moment. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/01/30626148-E457-4C58-836E-EEAE655E76B5_1_105_c.jpeg) Finding your way in buying or using software is changing. ## What actually gets democratised Software development is becoming more democratic. Not because coding no longer matters, but because much of the handwork no longer determines who gets to participate. It is now relatively easy to spin something up. To test an idea. To build a situational tool and discard it again. Software starts to behave more like a medium than a monument. At the same time, this does not remove the hard parts. Vision still matters. Knowing what to build, for whom, and why does not get automated away. Nor does the work of building systems that are robust, scalable, secure, and efficient. If anything, that work becomes more valuable. Good developers do not become less important here. They become more important. Not because they type faster, but because they understand structure, trade-offs, and long-term consequences. The same goes for people who know how to deploy systems into production, test them properly, and keep them running once they stop being experiments. What changes is the division of labour. A large amount of manual work has quietly been absorbed by tools. That opens space for non-coders to explore, and for coders to focus on what actually requires judgement. So when people say “you no longer need coders”, they are pointing at the wrong thing. What is disappearing is not the craft, but the gatekeeping. And in a noisier, more abundant landscape, that kind of discernment becomes the real advantage. ### Further reading - [Excel and the future cockpit of business logic](https://rob-hoeijmakers.ghost.io/excel-and-the-future-cockpit-of-business-logic/) - [Smaller, Nimbler, Smarter: AI Taking Paths Of Least Resistance](https://www.forbes.com/sites/joemckendrick/2026/01/15/smaller-nimbler-smarter-ai-taking-paths-of-least-resistance/) - [Excel, the Hidden Operating System of Business Reasoning](https://rob-hoeijmakers.ghost.io/excel-hidden-operating-system-of-business-reasoning/) ### ChatGPT Memory Has Quietly Changed (Plus & Pro) URL: https://rob-hoeijmakers.ghost.io/chatgpt-memory-has-quietly-changed-plus-pro/ Last updated: 2026-05-10T08:53:38.000Z If you use ChatGPT **Plus or Pro**, something important has shifted in how memory works. This is not a cosmetic update. ChatGPT can now draw on *past conversations* in a way it simply could not before. Not just preferences you explicitly saved, but relevant material from earlier chats. This **does not apply to** [**ChatGPT Business**](https://rob-hoeijmakers.ghost.io/chatgpt-personal-vs-business/). There, memory remains deliberately limited, which makes sense organisationally, but is a real contrast in capability. ### The short version ChatGPT now appears to work with **three distinct memory layers**, on top of its fixed system behaviour: 1. **Saved memory** Explicit, persistent facts you asked ChatGPT to remember. Preferences, recurring projects, long-term details. These live at account level and are manageable in settings. 2. **Contextual (conversation) memory** The implicit memory inside an active chat. This is the familiar “it remembers what we were just talking about” behaviour, bounded by the conversation window. 3. **Reference chat history** A newer layer that allows ChatGPT to *retrieve relevant information from past conversations*, even if nothing was explicitly saved. The real change is the third layer. Reference chat history is **retrieval-based**, not ambient. Past conversations are consulted when relevant, rather than always being “in mind”. 0:00 /1:25 1× Short demo of retrieval of older conversation in ChatGPT Plus ### Why this matters This subtly changes how ChatGPT behaves over time. - You don’t have to restate everything. - Longer lines of thought can accumulate. - The tool starts to feel less like a reset-on-every-use assistant, and more like a continuing workspace. For Plus and Pro users, that is a meaningful shift. For Business users, the contrast is now sharper: less continuity, more containment. ### What to do now The best way to understand this update is to **experiment**. - Start a fresh chat. - Refer to something you discussed weeks ago. - Notice when it recalls it, and when it doesn’t. - Toggle memory and “reference chat history” settings and observe the difference. Once you see the pattern, you will almost certainly start using ChatGPT differently. That alone makes this update worth knowing about. ### Further reading - [ChatGPT Personal vs Business](https://rob-hoeijmakers.ghost.io/chatgpt-personal-vs-business/) ### Just eight European quantum computers? URL: https://rob-hoeijmakers.ghost.io/just-eight-european-quantum-computers/ Last updated: 2026-05-10T08:53:38.000Z I am new to quantum computing. That is worth stating upfront, because this piece grows out of uncertainty rather than expertise. The trigger was [a short message from the European Commission](https://x.com/DigitalEU/status/2011435884001079794?s=20) claiming that the EU is building eight quantum computers. It sounded clear and ambitious. But the more I thought about it, the less clear it became. What does it mean, exactly, for “the EU” to build a quantum computer? And how does that relate to what I thought I knew about places like Delft, which I associated with serious quantum research? This article is an attempt to sort out that confusion. Not to resolve it fully, but to put some structure around it. > ⚡️Quantum computers can perform calculations radically faster than classical computers. > > Do you know that the EU is building 8 of them? > > Practical applications include FinTech, cybersecurity, transport management, meteorology, earth observation & more! > > ➡️ [pic.twitter.com/I8yrq4TNcO](https://t.co/I8yrq4TNcO) > > — Digital EU 🇪🇺 (@DigitalEU) [January 14, 2026](https://twitter.com/DigitalEU/status/2011435884001079794?ref%5Fsrc=twsrc%5Etfw) ## What is actually being talked about? After some digging, it became clear that the message refers to quantum computers developed within the [EuroHPC Joint Undertaking](https://www.eurohpc-ju.europa.eu/index%5Fen). **EuroHPC Joint Undertaking** is a European programme focused on **high-performance computing (HPC)**, meaning large-scale, high-capacity computing infrastructure such as supercomputers that perform many calculations in parallel. Its core mission is to fund and coordinate these systems so they can be accessed by researchers across Europe. Quantum computing has been added to that portfolio in recent years. So when EuroHPC lists “its” quantum computers, it is not making a general claim about quantum computing in Europe. It is describing a specific set of machines that are: - co-funded at EU level - hosted by selected national centres - intended as shared research infrastructure That scope is narrower than the original message suggests, but also more concrete. 💡 ****Joint Undertaking (EU governance context)** ****EuroHPC Joint Undertaking** is structured as a Joint Undertaking. This is a formal EU governance model used when the European Union and its Member States jointly develop and operate large, strategic programmes. A Joint Undertaking has its own legal identity and shared governance, allowing long-term coordination and pooled investment in areas that are too capital-intensive or strategically sensitive for individual countries to manage alone. In the case of EuroHPC, this structure underlines that high-performance computing is treated as a shared European capability rather than a standalone research initiative. ## A first realisation: infrastructure is a category of its own This distinction led me to a more general realisation. Quantum computing seems to exist on multiple levels at once. There is the level of infrastructure, where the goal is access, coordination, and capacity building. And there is the level of research, where the goal is understanding, experimentation, and long-term progress. The machines funded through EuroHPC belong clearly to the first category. They are not meant to represent the cutting edge of quantum science. They are meant to make quantum computing available as a shared resource, much like supercomputers are today. That does not make them less important. It just places them in a specific role. [Our Quantum ComputersFacts and details about the EuroHPC Quantum Computers.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/hcp_favicon-1.ico)The European High Performance Computing Joint Undertaking (EuroHPC JU)![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/eu-socialmedia-fallback-1.png)](https://www.eurohpc-ju.europa.eu/eurohpc-quantum-computers/our-quantum-computers%5Fen#lucy) ## Looking at the Netherlands from that angle From a Dutch perspective, this helped explain something that initially puzzled me. The Netherlands does not immediately appear in some of these EuroHPC overviews, yet it is often mentioned in discussions about advanced quantum research. Especially Delft. Institutes such as [QuTech](https://qutech.nl), closely connected to TU Delft and TNO, work on fundamental aspects of quantum computing: qubit technologies, quantum networks, error correction. These are not infrastructure machines designed for broad access. They are research systems, often experimental, sometimes fragile, and deeply specialised. From what I can tell so far, they live largely outside the EuroHPC framing, even though they may be scientifically more advanced than some infrastructure systems. ## Two layers, loosely connected At this point, my working picture looks something like this: - European programmes such as EuroHPC focus on shared infrastructure and coordination. - Universities and research institutes focus on knowledge creation and experimentation. - The two are related, but not interchangeable. Confusion seems to arise when these layers are collapsed into a single story about “having” quantum computers. ## Where this leaves me, for now I am deliberately stopping short of conclusions. What I take away at this stage is mainly a habit of asking better questions. When I read that “Europe is building quantum computers”, I now pause and ask: in what sense? Infrastructure, research, capability, or aspiration? For now, that distinction is enough. This piece is not a statement of understanding, but a marker of where my understanding currently begins. ### Further reading - [QuTech - Research institute for quantum computing and quantum internet](https://qutech.nl) - [Homepage](https://www.eurohpc-ju.europa.eu/index%5Fen) ### Why WhatsApp Channels Trigger the Digital Services Act URL: https://rob-hoeijmakers.ghost.io/why-whatsapp-channels-trigger-the-digital-services-act/ Last updated: 2026-05-10T08:53:39.000Z For a long time, WhatsApp felt easy to place. It was a communication tool: private, conversational, and largely absent from debates about platforms and media. When doubts arose, they tended to focus on groups and communities. Is WhatsApp still “just messaging”, or has it become a form of social media? A poll I ran some time ago [reflected that ambiguity](https://rob-hoeijmakers.ghost.io/is-whatsapp-social-media/). Roughly half of respondents saw WhatsApp as a communication tool; the other half as a social platform. What almost no one mentioned was a third role that, in hindsight, turns out to matter most. WhatsApp has also become something people *read*. [Is WhatsApp social media?Is WhatsApp a connectivity tool or a social medium? A casual chat revealed differing views. I see it as a way to connect, while my friend considers it social media.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-499.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/Is-whatsapp-social-media-1.jpg)](https://rob-hoeijmakers.ghost.io/is-whatsapp-social-media/) ## From private exchange to public reach With the introduction of Channels, WhatsApp added a layer that sits alongside private chats and groups, but behaves very differently. Channels are one-to-many by design. They are asymmetric. No shared conversation is required. Updates flow outward and are often consumed passively. This matters not because Channels resemble broadcasting in a cultural sense, but because they introduce **public reach**. Information is no longer exchanged between known participants. It is distributed to an open or semi-open audience, defined by interest rather than relationship. That single shift is enough to move a service into a different regulatory category. ## How the Digital Services Act sees platforms The Digital Services Act does not regulate platforms based on how they are branded, nor on whether they feel like media. It regulates based on whether a service **structures public reach at societal scale**. This is why services as different as app stores, maps, marketplaces, and video platforms can all qualify as Very Large Online Platforms. What they share is not content type, but function: they intermediate access, visibility, or distribution for a broad public. Crucially, the Act applies at **service level**, not company level. Ownership alone is not decisive. Different services operated by the same company can fall under different obligations if their function and reach differ. Core WhatsApp messaging remains private, enclosed, and non-discoverable. It does not shape public attention, even at enormous scale. Channels do. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/01/WhatsApp-Channels-Updates.jpg) WhatsApp Channels can be found under Updates. ## Why scale alone is not enough This also explains why WhatsApp messaging itself is not treated as a Very Large Online Platform. Size is necessary, but not sufficient. Private communication, even when widely used, lacks public reach. There is no general audience, no system-level visibility, and no platform-level structuring of attention. The Digital Services Act is explicit about this boundary. It is not a law about communication. It is a law about **public intermediaries**. Channels cross that boundary by design. ## The quiet irony The irony is not that WhatsApp is now being examined under the Act. The irony is that this happened without a dramatic product shift. Channels felt incremental, almost modest. And yet they introduced exactly the condition the law is designed to notice: outward-facing distribution at population scale. At the time of writing, WhatsApp Channels have crossed the relevant user threshold, but no formal designation decision has yet been issued. Nothing has gone wrong. No judgement has been made. This is simply the law doing its work as intended: observing scale, then assessing responsibility. ## Why this example matters As an illustration of European digital legislation, this case is unusually clean. It shows that the Digital Services Act does not ask platforms what they *are*, but what they *enable*, once enough people rely on them. Platform identity, in this framework, is not a matter of self-description. It is inferred from public reach. And sometimes, that inference reveals that a messaging app has become something else. [\-> Whatsapp channel for this blog](https://whatsapp.com/channel/0029VaiZXbXLtOjBxoqBwN2o) --- ### Current designated Very Large Online Platforms (VLOPs) under the DSA *(service-level designation; illustrative, not exhaustive)* 1. **AliExpress** 2. **Amazon Store** 3. **Apple App Store** 4. **Booking.com** 5. **Facebook** 6. **Google Play** 7. **Google Maps** 8. **Google Shopping** 9. **Instagram** 10. **LinkedIn** 11. **Pinterest** 12. **Snapchat** 13. **TikTok** 14. **X** 15. **Wikipedia** 16. **YouTube** 17. **Zalando** ### Very Large Online Search Engines (VLOSEs) 1. **Google Search** 2\. **Bing** ### Further reading - [WhatsApp might soon be subject to stricter scrutiny under the EU’s Digital Services Act](https://www.engadget.com/social-media/whatsapp-might-soon-be-subject-to-stricter-scrutiny-under-the-eus-digital-services-act-191000354.html) - [WhatsApp faces new obligations under the EU’s toughest digital rules](https://www.euronews.com/next/2026/01/26/whatsapp-faces-new-user-protection-obligations-under-the-eus-toughest-digital-rules) - [Digital Europe in Two Acts: DSA and DMA](https://rob-hoeijmakers.ghost.io/dsa-dma-compared/) - [Europe Fines X: The Moment the DSA Became Real](https://rob-hoeijmakers.ghost.io/dsa-fines-x/) - [Hoeijmakers.net - Living and building the digital now](https://whatsapp.com/channel/0029VaiZXbXLtOjBxoqBwN2o) ### ChatGPT Personal vs Business URL: https://rob-hoeijmakers.ghost.io/chatgpt-personal-vs-business/ Last updated: 2026-05-10T08:53:39.000Z ### Why the difference is not features, but how thinking is shaped When people compare ChatGPT subscriptions, the discussion usually gravitates towards surface distinctions: access to models, message limits, collaboration features, privacy guarantees. Those things matter, but they are not where the most consequential difference sits. In practice, using ChatGPT in a personal context feels fundamentally different from using it in a business context. Not because one is more powerful than the other, but because each configuration embodies a different assumption about continuity, memory, and responsibility. What looks like a subscription choice is, in fact, a choice between two interaction regimes. And those regimes shape how you think, how you work, and what kinds of outcomes you can reasonably expect. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/04/ChatGPT-Personal-vs-Business--Team-.png) An overview of the differences between ChatGPT Personal and Business (Team) ## One system, two regimes At a technical level, both Personal (Plus) and Business access the same underlying models. There is no “smarter” or “dumber” intelligence at play. The difference lies in how conversations are treated over time. In a personal setup, conversations feel connected. You can have multiple parallel threads, return to earlier themes, and experience a sense that the system “knows what you have been working on”. Context appears to accumulate implicitly. Not as a carefully curated memory, but as an emergent state built from recent interaction. In a business setup, that continuity disappears. Each conversation is a bounded workspace. Context does not carry over unless you explicitly load it. Documents must be attached again. Background assumptions must be restated. Nothing is implicitly remembered across sessions. This is not a flaw. It is a deliberate design choice. ## Emergent continuity and personal synthesis The personal mode invites exploration. Because context lingers, you can think in fragments, circle around a topic, and allow ideas to mature over time. Separate conversations can subtly inform one another. The system becomes a kind of reflective surface that supports synthesis rather than execution. This works particularly well for: - exploratory research - writing and thinking in drafts - personal learning trajectories - loosely structured sense-making The cost of this mode is imprecision. Because continuity is implicit and uncontrolled, you cannot fully account for what influences a response. The “memory” is not inspectable, auditable, or shared. It is experiential rather than contractual. That is often acceptable, even desirable, when the work is personal and provisional. ## Explicit context and accountable production Business mode makes the opposite trade-off. Nothing is assumed. Everything must be stated. Context is not ambient but deliberate. This changes the character of the interaction. Conversations become more linear. Work becomes more document-centric. The emphasis shifts from exploration to production, from synthesis to traceability. This mode supports: - collaborative work - reproducibility and review - privacy and data boundaries - organisational accountability The system does not help you remember what you were thinking last week. Instead, it forces you to externalise that thinking into documents, prompts, and shared artefacts. Again, this is not a limitation. It is what makes professional use defensible. ## Linearity as a condition for cooperation What this structure really optimises for is cooperation. When work is linear, bounded, and explicit, it becomes shareable. Colleagues can follow the reasoning, assess decisions, and build on the same material without relying on invisible context. Implicit continuity supports individual cognition, but it undermines collective legibility. Explicit context does the reverse. The apparent rigidity of business mode is therefore not accidental. It is the condition that makes coordination, review, and shared responsibility possible. Seen this way, linearity is not a usability compromise. It is a design principle. ## Why this difference matters If you treat ChatGPT as a text generator, the distinction barely registers. If you treat it as a medium for thought, it becomes decisive. The personal regime supports cognitive continuity. The business regime supports organisational integrity. Problems arise when these regimes are mixed without awareness. When exploratory, half-formed thinking is treated as production-ready output. Or when accountable workflows are expected to emerge from implicit, private context. Much of the confusion around “why ChatGPT behaves differently” can be traced back to this mismatch in expectations. ## A question of posture, not preference This is not an argument for choosing one subscription over the other. Many professionals benefit from using both, consciously and separately. What matters is recognising that each setup invites a different posture: - one oriented towards personal synthesis and intellectual drift - the other towards shared context, responsibility, and repeatable work Once you see that, the subscriptions stop being a feature comparison and start being what they actually are: different ways of structuring human–AI collaboration. And that, ultimately, is the more interesting design question. ### Further reading - [Choosing ChatGPT Team: From AI-Confidant to Trusted Colleague](https://rob-hoeijmakers.ghost.io/choosing-chatgpt-team-from-ai-confidant-to-trusted-colleague/) - [The Workspace Advantage: Google’s Quiet Lead in the AI Shift](https://rob-hoeijmakers.ghost.io/google-gemini-workspace-future/) ### WhatsApp in 2026: From Phone Numbers to Names URL: https://rob-hoeijmakers.ghost.io/whatsapp-in-2026-from-phone-numbers-to-names/ Last updated: 2026-05-10T08:53:39.000Z For most of its existence, WhatsApp relied on a simple assumption: if someone knows your phone number, they can reach you. In 2026, that assumption changes. WhatsApp is introducing usernames, and with them a different way of thinking about identity, privacy, and something more subtle: **addressability**. This is not a story about features. It is about what happens when a platform stops using a piece of infrastructure as a stand-in for who you are. ## Borrowed identity A phone number is an infrastructural artefact. It was designed to route calls, not to function as a social identity. WhatsApp nevertheless turned it into exactly that. Your number became your address, your identifier, and your point of contact. The result was simplicity, but also friction. Starting a conversation meant sharing an identifier that worked far beyond the context in which it was given. Over time, that tension became harder to ignore. WhatsApp gradually detached itself from the phone: - accounts outliving devices - conversations living on the platform - multi-device use without an active handset Usernames are where that long shift becomes visible. ## Addressability instead of exposure Usernames do not make people anonymous. WhatsApp still knows who you are. What changes is *how others can reach you*. A username allows contact **within** WhatsApp without exposing an identifier that exists outside it. That difference matters most at the edges of social life: first contact, temporary groups, semi-public situations. Until now, sharing a phone number collapsed those contexts into one. A single identifier leaked from one situation into many others. Usernames introduce containment. Reachability becomes contextual rather than absolute. This is not about hiding. It is about not over-sharing by default. [Identity and Addressability in the Digital AgeUnderstanding digital identity is crucial in today’s connected world. Explore how addressability, through email, phone numbers, and online handles, shapes our online presence.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-491.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_4342-1-1-3.jpeg)](https://rob-hoeijmakers.ghost.io/identity-and-addressability/) ## Privacy as a structural outcome It is tempting to frame this as a privacy feature, but that misses the point. Nothing here depends on users making better choices or managing more settings. The change is architectural. It alters what happens *by default* when a conversation begins. Starting a chat no longer automatically means handing over something that persists beyond the platform. For most people, this will not feel like “more privacy”. It will feel like fewer awkward moments and less clean-up afterwards. That is often how privacy improvements show up in practice. ## Addressability works both ways Addressability does not only affect individuals. It also changes how organisations can be reached. Until now, finding a business on WhatsApp required a phone number. There was no real notion of search. You either had the number, or you did not. Usernames introduce a limited but meaningful form of discoverability. People can reach a business **by name**, without first handling a phone number. This does not turn WhatsApp into an open directory. Discovery appears controlled and exact-match rather than exploratory. Even so, it marks a shift. WhatsApp gains a way to reference organisations directly, as entities on the platform. And businesses gain a way to be reachable without turning a phone number into a public artefact. That difference matters. ## A platform with its own addresses Once a platform introduces its own addressing system, and allows people and organisations to be reached by name, it changes character. WhatsApp has long presented itself as a thin layer on top of phone numbers. Usernames undermine that framing. They give WhatsApp an internal way to point to participants without relying on external infrastructure. That is what platforms do. Everything else follows from that: easier sharing, clearer entry points, better placement on the web. But those are consequences, not the core shift. [WhatsApp is Quietly Becoming a Platform — And That Changes EverythingWhatsApp Is Becoming a Platform: What That Means for Identity, Contact Management and Strategy![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-492.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_3606-3.jpeg)](https://rob-hoeijmakers.ghost.io/whatsapp-platform-shift-contact-identity-strategy/) ## From numbers to addressable identities Seen in context, usernames are not a sudden reinvention. They are the next step in a long uncoupling. WhatsApp is moving: - from borrowed identifiers - to platform-specific addressability - from universal reach to contextual reach Or, more simply: > WhatsApp is moving from *a phone number with chat* > to *a chat platform with addresses*. It is a quiet change. But it reshapes who can reach you, and on what terms. ## A brief technical note Behind usernames sits a layer of internal identifiers that keeps conversations consistent even when phone numbers are hidden or usernames change. Most users will never see this. But it reflects the same underlying move: separating *identity* from *reachability*. That separation is where platforms grow up. ### Further reading - [Identity and Addressability in the Digital Age](https://rob-hoeijmakers.ghost.io/identity-and-addressability/) - [WhatsApp is Quietly Becoming a Platform — And That Changes Everything](https://rob-hoeijmakers.ghost.io/whatsapp-platform-shift-contact-identity-strategy/) ### Exploring ISO 42001 and AI governance URL: https://rob-hoeijmakers.ghost.io/iso-42001-ai-governance/ Last updated: 2026-05-10T08:53:39.000Z It is the start of a new year, and like many people working with AI, I have been spending time reading things that are not models, prompts, or product updates. Instead, I have been reading regulation. The EU AI Act. Privacy law. Compliance language. Procurement questionnaires. The kinds of documents that do not describe how AI works, but how organisations are expected to *live with it*. When you follow those threads for a while, you start to notice a different layer of the AI conversation. Less about innovation, more about responsibility. Less about what AI can do, more about who is accountable when it does it. That is how I recently ran into **ISO 42001**. I am not writing this to argue for or against it. At this stage, I am simply trying to understand what it is, why it exists, and why it is starting to appear in conversations around AI, compliance, and procurement. ## What ISO standards are, briefly ISO is the [International Organization for Standardization](https://www.iso.org/home.html). It does not regulate, and it does not certify organisations itself. What it does is define shared standards that describe how organisations can structure their work around specific themes. Many of the better-known ISO standards focus on *management systems*. **ISO 9001** for quality management. **ISO 27001** for information security. **ISO 27701** for privacy. These standards do not say “your outcomes must be good”. They say: if you care about this topic, here is how responsibility, risk, documentation, and continuous improvement are typically organised. Certification is handled by accredited third parties, not by ISO itself. Organisations are audited against the standard by certification bodies, which are in turn overseen by national accreditation authorities. It is a layered system. Procedural. Institutional. Familiar to anyone who has worked with large organisations or public-sector procurement. [ISO - International Organization for StandardizationWe’re ISO, the International Organization for Standardization. We develop and publish International Standards.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/apple-touch-icon-152x152-precomposed.png)ISO![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/iso-logo-og.png)](https://www.iso.org/home.html) ## What ISO 42001 adds to that landscape ISO 42001 is the first ISO standard specifically focused on AI. More precisely, it defines the requirements for an **Artificial Intelligence Management System**. That phrasing matters. This is not a technical AI standard. It does not evaluate models, measure bias, or certify that an AI system is “ethical”. Instead, it asks whether an organisation has a structured way to govern AI across its lifecycle. Things like: - defining which AI systems are in scope, - assigning responsibility and oversight, - assessing risks and potential impacts, - managing data, suppliers, and third-party components, - handling incidents and continuous improvement. In other words, it standardises *how questions about AI are handled*, not the answers themselves. If you are familiar with ISO 27001, the logic will feel recognisable. ISO 42001 follows the same management-system structure, just applied to AI. [ISO/IEC 42001:2023Information technology — Artificial intelligence — Management system![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/apple-touch-icon-152x152-precomposed-1.png)ISO![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/iso-logo-og-1.png)](https://www.iso.org/standard/42001) ## Why this shows up in procurement, not product design Formally, ISO 42001 applies to “organisations”. In practice, it attaches itself most strongly to **vendors**. That is not because vendors are morally more responsible, but because procurement needs certainty. Large organisations, public bodies, and regulated sectors often prefer not to deeply inspect every supplier’s internal practices. Certification becomes a shortcut. A way to say: an independent party has checked that there is a governance system in place. In that sense, ISO 42001 functions a bit like insurance. Not because it prevents problems, but because it redistributes risk and responsibility. It creates a shared baseline that allows organisations to move forward without fully understanding each other’s internal complexity. This also explains why such standards are often more relevant for AI vendors than for advisory or exploratory work. The closer you are to operating AI systems in production, the more likely these questions become unavoidable. [The Hot Potato of ComplianceFrom GDPR to the EU AI Act, a recurring pattern emerges: European regulation lands in procurement, spreading responsibility, caution, and friction.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-487.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/2FAE4B6B-ECF4-4870-AED8-93282D1EB0C4-2.jpeg)](https://rob-hoeijmakers.ghost.io/the-hot-potato-of-compliance/) ## How this relates to the EU AI Act ISO 42001 is not law. It does not replace the EU AI Act, and it does not guarantee compliance. But the two are clearly moving in the same direction. The AI Act introduces a risk-based regulatory framework. ISO 42001 introduces a risk-based governance framework. One is legal, the other procedural. What ISO 42001 offers is not legal compliance, but *organisational readiness*. A way for companies to show that they have thought about responsibility, oversight, and control in a structured way. For regulators, clients, and procurement teams, that distinction often matters less than one might expect. [Learning to Work with the EU AI ActI used to avoid EU regulation. Now I’m learning to work with it. The AI Act isn’t perfect, but it’s shaping how I think about risk, trust, and tech.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-488.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_4160-4.jpeg)](https://rob-hoeijmakers.ghost.io/learning-to-work-with-the-eu-ai-act/) ## Where this fits into my own exploration For me, this sits alongside a broader research question: how responsibility for AI is being redistributed. On the one hand, there is a strong emphasis on AI literacy. Organisations are encouraged to understand AI better, make informed decisions, and build internal competence. On the other hand, there is a parallel move to push complexity upstream. To require vendors to demonstrate that they have governance systems in place, so that not every organisation has to reinvent that wheel. ISO 42001 clearly belongs to the second category. It does not replace understanding. But it does change where questions are asked, and who is expected to answer them. In [my own work](https://schmuki.nl), this is not a primary goal right now. It is something I am keeping in my peripheral vision. Something to be aware of as I look at my own organisation, the vendors we work with, and the institutional environment forming around AI. Not as a decision to make, but as a signal. ## Why it is worth noticing, even if you do nothing with it You do not need to pursue ISO 42001 to find it interesting. Its value, at least for me, lies in what it reveals about the direction things are moving. AI is no longer just a technical or creative concern. It is becoming part of the same governance machinery that already exists for quality, security, and privacy. Whether that is reassuring, constraining, or simply inevitable is not something I am trying to resolve here. For now, it is enough to recognise that this layer exists, to understand roughly how it works, and to notice when it starts to appear in conversations. Sometimes orientation is more useful than opinion. ### Further reading - [ISO/IEC 42001: a new standard for AI governance](https://kpmg.com/ch/en/insights/artificial-intelligence/iso-iec-42001.html) - [ISO/IEC 42001:2023 Artificial Intelligence Management System Standards - Microsoft Compliance](https://learn.microsoft.com/en-us/compliance/regulatory/offering-iso-42001) - [ISO/IEC 42001 - Compliance](https://cloud.google.com/security/compliance/iso-42001) ### The limits of “unlimited” mobile data URL: https://rob-hoeijmakers.ghost.io/unlimited-mobile-data/ Last updated: 2026-05-10T08:53:40.000Z *A European perspective, based on lived experience* “Unlimited data” sounds simple. It suggests freedom, predictability, and one less thing to worry about. Yet if you have travelled within Europe, compared mobile plans, or suddenly hit a slowdown abroad, you will probably have felt that something about “unlimited” does not quite add up. This piece is not a technical deep dive, nor an exposé. It is an attempt to digest my own experiences with mobile data plans in a European context and to make the underlying patterns a bit more legible. The hope is that this helps others when they run into the same quiet confusion and have to make a choice themselves. The key idea is simple: *unlimited is not one thing*. ## Unlimited depends on context Within Europe, the word *unlimited* is used across very different situations. The confusion starts when we assume it always means the same thing. In practice, I keep running into three distinct models. ### 1\. Unlimited at home Within your home country, an unlimited mobile plan usually means: - No predefined monthly data cap - Normal 4G or 5G speeds - A fair use policy that exists, but is rarely felt in everyday use There is no daily allowance, no visible counter, and no routine throttling. Heavy use is acceptable as long as it looks like normal personal behaviour. This model works because domestic networks are predictable. Operators control the infrastructure end to end and rely on statistical averaging. Some people use a lot, most people do not, and it balances out. For many users, this is what “unlimited” intuitively feels like. ### 2\. Unlimited, but roaming within Europe The moment you cross a border, even within the EU, the meaning shifts. On an unlimited domestic plan, roaming typically comes with: - A clearly defined monthly roaming allowance - A hard ceiling, not a soft suggestion - A stop, surcharge, or required add-on once the allowance is used This is where many people first notice the limits of unlimited. It is not a trick or an exception. Roaming traffic runs on foreign networks and is settled at wholesale rates. EU regulation ensures access, but it does not make unlimited roaming economically free. As a result, limits become explicit and numerical. Unlimited at home quietly becomes limited abroad. ### 3\. “Unlimited” travel eSIMs Travel eSIMs add a third interpretation, and this is where the word starts to stretch. Many European travel eSIMs advertise “unlimited data”, but what they actually offer is: - A fixed amount of high-speed data per day, often around a few gigabytes - Throttling to much lower speeds after that - A reset after 24 hours You are not cut off. The connection remains active for the duration of the plan. But high-speed capacity is rationed daily. This is unlimited access, not unlimited performance. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/01/Unlimited-mobile-data-eSIM.jpg) I now see why 100GB is much more expensive than Unlimited. ## Two kinds of limits At this point, it helps to name the difference that is often left implicit. There are two fundamentally different ways a plan can fail. **Throttling** - You stay online - Speed drops, sometimes sharply - Messaging and light browsing still work - Video calls, uploads, or hotspot use become frustrating **Hard caps** - Full speed until the allowance is gone - Then data stops - Performance is predictable - The moment of failure is not Neither approach is objectively better. They optimise for different kinds of reassurance. ## The psychology behind the choice This is where the decision becomes less technical and more human. A throttled “unlimited” plan reassures you that you will never be completely disconnected. Something will always work, even if it becomes slow. A large capped bundle, say 100 GB for a trip, offers a different comfort. For many people, that amount is effectively unlimited. Until it suddenly is not. The anxiety shifts: - With throttling: *Will this still be usable today?* - With caps: *Will I hit zero at the wrong moment?* Interestingly, the throttled “unlimited” option is often cheaper. That only makes sense if you assume that many users will spend a significant part of their time at reduced speeds. ## What happens when a cap is reached? On most capped travel eSIMs, the behaviour is straightforward: - Once the allowance is used, data stops - You need to top up or buy a new plan - Any remaining validity period becomes irrelevant There is usually no slow fallback and no safety net. The line is clean. This makes capped plans less forgiving, but also more honest. You know exactly what you are buying, and exactly where it ends. ## So what does “unlimited” really mean? In a European context, it tends to mean different things in different economic settings: - At home: trust and statistical averaging - While roaming: regulated volume limits - On travel eSIMs: controlled speed over time The word survives because it is legally defensible and psychologically comforting. But technically, unlimited always has edges. The more useful questions are not whether something is *really* unlimited, but: - Where is the limit? - Is it a slowdown or a stop? - And which one fits the way I actually use my phone? Once those limits are visible, unlimited becomes less magical, but far more practical. ### Further reading - [Roaming: what you pay to use your smartphone in another EU country - Your Europe](https://europa.eu/youreurope/citizens/consumers/internet-telecoms/mobile-roaming-costs/index%5Fen.htm) - [Everyday life runs on daily data](https://rob-hoeijmakers.ghost.io/mobile-data-roaming-and-fibre/) ### Everyday life runs on daily data URL: https://rob-hoeijmakers.ghost.io/mobile-data-roaming-and-fibre/ Last updated: 2026-05-10T08:53:40.000Z I almost never exceed 10 or 12 gigabytes of mobile data per month. Not because I am particularly careful, or because I avoid my phone. It is simply how I use it. Messaging, email, maps, reading, some browsing. No music streaming, hardly any video. And whenever there is Wi-Fi, my phone switches over without me thinking about it. For a long time, I assumed this made me an exception. It turns out it does not. Recent [Dutch figures](https://www.acm.nl/nl/publicaties/acm-telecommonitor-ook-derde-kwartaal-2025-forse-stijging-mobiele-dataverbruik) suggest that mobile data use sits at roughly **10 to 11 gigabytes per SIM per month**. That is not a niche. It is the middle. Many people use less, a smaller group uses far more, and almost all of that difference comes down to one thing. **Video.** An hour of HD video can easily consume three gigabytes. Do that a few times a week and you move far beyond the average. Not gradually, but abruptly. What is interesting is that this does not happen automatically. For many people, mobile remains a lightweight layer. It is where you coordinate, check, skim, respond. The heavier consumption happens elsewhere, on fixed connections that are so present we barely notice them. ## Fixed internet does the heavy lifting At home, fixed internet usage in the Netherlands typically sits between **300 and 500 gigabytes per household per month**, and often more. Streaming, updates, cloud backups, video calls, gaming. Increasingly AI-driven services as well. Put next to each other, the contrast is stark. Fixed connections carry **roughly thirty times more data** than mobile connections do, per subscription. Not because mobile networks are underused, but because they are not built for sustained bulk traffic at that scale. This is structural, not temporary. Mobile operates over shared spectrum. Capacity is finite, local, and collective. Fibre scales differently. Once it is in the ground, adding traffic is mostly a matter of equipment, not physics. This is also why “unlimited” mobile plans always come with conditions, and why roaming limits persist, even within Europe. The language suggests abundance. The infrastructure quietly enforces boundaries. [Switching to Optical: My Journey to Fibre InternetI switched from coax to fibre — faster, cheaper, and finally competition. Fibre opens real choice, and even 5G (FWA) enters the game. Here’s how it went.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-479.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_3217.jpeg)](https://rob-hoeijmakers.ghost.io/switch-coax-fibre-internet/) ## Roaming as a reality check Those boundaries become visible when you travel. I often spend holidays in places that are genuinely off-grid. No cable. No fibre. Sometimes no Wi-Fi at all. And although it is a holiday, it can also be a working week. That is when mobile stops being a complement and becomes the backbone. On paper, everything looked fine. My roaming plan said “unlimited”. In practice, I ran into limits almost immediately. After about **one gigabyte of use**, speeds dropped to the point where video calls or sustained work became difficult. Nothing was broken. This was the plan working exactly as designed. “Unlimited” did not mean unconstrained. It meant continuous access, with an assumed usage pattern. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/01/IMG_4995.jpeg) Office. ## What “unlimited” actually means Across domestic mobile plans, EU roaming, and travel eSIMs, the logic is consistent. **Unlimited protects you from:** - hard cut-offs - surprise charges **It does not protect you from:** - capacity limits - congestion - fair use thresholds Especially for roaming and travel eSIMs, “unlimited Europe” almost always means unlimited *access*, not unlimited *throughput*. There is usually a daily or monthly full-speed allowance. After that, you are still connected, but on different terms. This works well as long as mobile remains secondary. The moment it has to replace fixed internet, the limits surface. [The limits of “unlimited” mobile dataUnlimited data rarely means the same thing twice. At home, abroad, or on eSIMs, the limits shift. This piece maps where they actually are.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-486.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_5709-1.jpeg)](https://rob-hoeijmakers.ghost.io/unlimited-mobile-data/) ## Holidays as stress tests Holidays make this visible. Fixed Wi-Fi disappears. Mobile usage spikes. Not because behaviour suddenly becomes extreme, but because context changes. Mobile absorbs traffic that would normally land on fibre. Limits appear. Expectations adjust. Video quality drops. Offline modes suddenly make sense again. What this shows is not excess, but elasticity. We adapt quickly when constraints are felt. 🇪🇺 EU roaming is not [unlimited](https://rob-hoeijmakers.ghost.io/unlimited-mobile-data/): even on domestic “unlimited” plans, roaming data is capped by EU fair-use rules that tie the allowance to wholesale data prices. From ****1 January 2026**, the EU wholesale cap drops to ****€1.30 per GB**, which increases roaming allowances modestly. In practice, this means that a ****€10 mobile plan** typically includes around ****18 GB of EU roaming data per month** (up from roughly ****15 GB** today), with higher-priced plans scaling accordingly. After that allowance is used, operators may throttle speeds or apply surcharges. These limits apply across the EU roaming zone, regardless of whether the domestic plan is marketed as unlimited. ## One system, different roles Mobile data, roaming, and fixed internet are often discussed separately. In practice, they form one system. Fixed connections carry the bulk. Mobile provides continuity. Roaming exposes the seams. When one layer weakens, the others compensate, up to a point. Total data usage keeps growing, but unevenly. It accumulates where the system can handle it most efficiently. Expectations follow, until they meet constraints and recalibrate. Video is where that recalibration becomes unavoidable. ## Knowing where you stand None of this requires optimisation. But having a rough sense of your own data usage is a form of digital literacy. Not to chase averages, but to understand what your habits rely on. How much of your digital life assumes fixed connectivity. And why “unlimited” so often turns out to be contextual rather than absolute. I still rarely exceed my mobile allowance. Not because I should, but because it fits how the system actually works. That, more than the headline numbers, is the pattern worth paying attention to. --- ### Who uses mobile data, and how To make this more concrete, the charts below show two complementary views: how mobile data is typically distributed across activities, and how mobile data use is distributed across users. They do not describe the Netherlands specifically, but reflect a pattern that shows up consistently across mature mobile markets. The key point is not the exact percentages, but the shape of the distribution: **video dominates total mobile data volume, while a relatively small group of heavy users accounts for a disproportionate share of usage**. **Mobile Data Usage by Category (typical distribution)** | Category | % of total mobile bytes | | ---------------- | ----------------------- | | Video streaming | 60–80% | | Social media | 8–15% | | Web browsing | 5–10% | | Audio streaming | 3–7% | | Other/background | 2–5% | **Mobile Users by Data Consumption Bucket** | Users (%) | Monthly Data Used | | --------- | ----------------- | | \~40–50% | < 5 GB | | \~25–30% | 5–20 GB | | \~15–20% | 20–50 GB | | \~5–10% | \> 50 GB | **Sources / methodology note** These distributions are based on recurring findings from international network measurement and traffic analysis reports, including **OpenSignal’s mobile experience reports** and **Sandvine’s Global Internet Phenomena reports**. Both analyse aggregated, anonymised mobile network traffic across multiple countries and operators. While national regulators such as the ACM do not publish application-level breakdowns for the Netherlands, the overall pattern shown here is stable across comparable European markets and helps explain why averages, fair-use caps, and roaming limits behave the way they do. ### Further reading - [Roaming: what you pay to use your smartphone in another EU country - Your Europe](https://europa.eu/youreurope/citizens/consumers/internet-telecoms/mobile-roaming-costs/index%5Fen.htm) - [Fibre Feels Faster — And It’s Not About Speed](https://rob-hoeijmakers.ghost.io/fibre-feels-faster-and-its-not-about-speed/) - [Taming Data Usage with a MiFi hotspot: My Experience and Solution](https://rob-hoeijmakers.ghost.io/taming-data-usage-with-a-mifi-hotspot-my-experience-and-solution/) - [Etymology and meaning of “Data”](https://rob-hoeijmakers.ghost.io/etymology-of-data/) ### Typography, in motion URL: https://rob-hoeijmakers.ghost.io/typography-in-motion/ Last updated: 2026-05-10T08:53:40.000Z I have always had a soft spot for [typography](https://rob-hoeijmakers.ghost.io/typography-insights/). Not as ornament or branding, but as infrastructure. Letters as dependable carriers of meaning, doing their work without drawing attention to themselves. For a long time, typography felt like a settled discipline. Important, certainly, but largely associated with print and slower forms of reading. Books, essays, editorial layouts. Contexts where time and attention are available. That assumption no longer holds. ## Where reading now happens Most of my reading today happens on screens. Short fragments on a phone. Longer stretches on a large desktop monitor. Sustained reading on an e-reader, where the typography is deliberately restrained and tuned for endurance rather than speed. Each of these environments has its own demands. Screen size, distance, lighting, posture. Typography adapts accordingly, often invisibly. The car introduces a very different condition. ## The car as a visual interface I would hesitate to call what happens in a car “reading”. It is closer to checking or registering. A glance rather than a gaze. In more traditional cars, [dashboards](https://rob-hoeijmakers.ghost.io/why-dashboards/) rely on physical indicators: dials, needles, icons. Information is conveyed through shape and position. Text plays a minor role. In newer vehicles, screens take over. Navigation, system messages, warnings, media, communication. Information appears on central displays, instrument clusters, windscreens, head-up displays. The car now communicates continuously, largely through visual means. Text becomes part of the interface. ## Reading in a fraction of a second In these environments, information must be processed under pressure. The driver’s attention is divided. Light conditions change constantly. The eye moves rapidly, guided by saccades: quick jumps from one fixation point to the next. Typography here is not about elegance or expression. It is about immediate recognition. Letters must be legible in isolation, unambiguous at a glance, readable before conscious interpretation has time to intervene. Text approaches the role of icons. This is where type design becomes a matter of ergonomics. ## A typeface designed for the dashboard [Volvo’s Centum typeface](https://www.volvocars.com/intl/news/articles/volvo-cars-introduce-typeface-designed-with-safety-in-mind/) is interesting precisely because it is framed in these terms. It is not presented primarily as a branding exercise, but as a functional one. Designed, tested, and refined for legibility in moving vehicles, across lighting conditions, and under cognitive load. The goal is reduction of friction. Clear differentiation between similar letterforms. Open shapes. Calm rhythm. Nothing that slows recognition or invites doubt. That it also reinforces Volvo’s brand values is almost incidental. The sense of calm, control, and restraint emerges naturally from the functional requirements. The brand is expressed through behaviour rather than decoration. ## A personal hesitation about screens At the same time, I remain ambivalent about screen-first car interiors. Screens are emissive. They glow, day and night. Considerable intelligence is required to keep them legible: adaptive brightness, contrast shifts, glare reduction, compensation for tunnels and oncoming headlights. It works, most of the time, but it is a complex solution to a problem that physical controls largely avoid. Knobs, switches, and dials communicate through touch and position. They can be operated without looking. Muscle memory plays a role. Information is conveyed through resistance and orientation, not light. From that perspective, the rise of typography in cars is slightly paradoxical. Text becomes more critical precisely because interfaces demand more visual attention. Good type design compensates for a medium that is, ergonomically, still open to debate. This is not nostalgia. It is a recognition of cognitive limits. ## Typography across contexts What makes this moment interesting is how clearly typography now responds to context. In the car, letters are pushed towards speed, clarity, and near-invisibility. On my own blog, I recently made the opposite move. I returned to a serif typeface for body text, favouring reading comfort over neutrality. A subtle shift, but a deliberate one. [Serifs](https://rob-hoeijmakers.ghost.io/typeface-to-serif-or-not-to-serif/) still support sustained reading. They guide the eye, soften rhythm, and encourage slower engagement. On large screens and e-readers, that matters. Placed side by side, these choices underline the same point: typography is not a fixed style or a nostalgic craft. It is a living discipline, shaped by use, medium, and human perception. 🔤 ****Typographic note** I recently updated the body text of this blog, moving from one serif to another. I replaced Merryweather with IBM Plex Serif, which I find subtler in rhythm and more comfortable for long-form reading. ## From clay tablets to dashboards From marks carved into clay and wood, to ink on paper, to pixels on glass, and now to luminous text embedded in moving machines, the role of typography has remained remarkably consistent. It mediates between humans and systems. What changes are the constraints. Speed. Light. Motion. Attention. In the dashboard of a modern car, those constraints are unusually tight. That makes typography less visible, but more consequential. Perhaps that is why it still fascinates me. Not because letters are beautiful, but because, millennia on, they are still doing the quiet work of making complex systems legible, even when time is scarce. ### Further reading - [Typography insights](https://rob-hoeijmakers.ghost.io/typography-insights/) - [Typeface: to serif or not to serif](https://rob-hoeijmakers.ghost.io/typeface-to-serif-or-not-to-serif/) - [From Typewriter to the Matrix](https://rob-hoeijmakers.ghost.io/single-spaced-fonts/) ### Colour, systems, and the moment it clicked URL: https://rob-hoeijmakers.ghost.io/colour-systems-and-the-moment-it-clicked/ Last updated: 2026-05-10T08:53:40.000Z For most of my working life, colour was a digital thing. I designed on screens, judged on screens, published on screens. Print existed somewhere in the background, but it always felt slightly opaque. Colours came back different. Deeper on screen, flatter on paper. More constrained. More brittle. I accepted that as one of those things you “just had to know”, without ever really understanding why. That changed recently, through a combination of a book, a branding question, and a slow realisation about how colour systems actually work. ## A book I still love The immediate trigger was *A Dictionary of Color Combinations*, based on the colour studies of Japanese artist **Sanzo** **Wada** and later compiled by Seigensha, which I was given as a present. It is a lovely object. Calm, precise, generous. It does not define colours in isolation, but shows how they behave together. Pairs, trios, quartets. Harmonies rather than absolutes. It is inspiring in a very practical way. For clothing. For interiors. For branding. I still use it, and I still love it. What struck me, though, was a review that mentioned that the colours “used to be richer and deeper” and that newer editions felt flatter, more CMYK-like. At the time, that comment lingered. Only later did it start to make sense. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/01/IMG_1944.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/01/IMG_1945.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/01/IMG_1946.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/01/IMG_1947.jpeg) A dictionary of color combinations ## Designing our own brand, digital first Around the same time, I was looking back at our own brand. We designed it ourselves, deliberately. Two colours. A deep blue and a pinkish glow. They are not arbitrary. They are derived from Earth’s shadow: the darkening blue of the shadow itself, and the pink glow caused by atmospheric scattering at its edge. Those colours came from observation, not from a colour book. From the start, the brand was digital-first. Screen-native. Light-based. So when I checked the original files, there were no Pantone values. No CMYK intent. Just RGB. That was not an oversight. It was simply where the design lived. Only later did the question arise: if we ever want to print this properly, what *are* these colours? ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/01/logo-schmuki-vof-klein.png) Logo of our company ## Two very different kinds of “print colour” This is where something important finally became clear to me. Pantone and CMYK are often mentioned together, but they are fundamentally different kinds of colour systems. Pantone colours are *defined*. They are closer to pigments. They exist as named, specified inks that are mixed deliberately and printed as such. If you choose a Pantone colour at design time, you are making a conscious commitment to a physical colour. CMYK works differently. It is not a set of colours, but a *reproduction system*. Colours are built up from tiny dots of cyan, magenta, yellow, and black, arranged in patterns that the eye blends together. It is about simulation rather than definition. That distinction matters more than I realised. Pantone answers the question: **Which colour do we want, materially?** CMYK answers a different one: **How can this colour be reproduced, reliably, at scale?** ## The Pantone myth For a long time, I implicitly assumed that Pantone values were somehow hidden inside files, waiting to be uncovered. That if you just looked hard enough, you would find “the real colour”. That turns out to be wrong. Pantone is not something you recover. It is something you choose. If Pantone was not defined at design time, it does not exist. Anything you assign later is a closest match, a judgement call, an approximation. However precise it looks. That was the first real eye-opener. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/01/pantone-color-chip-pq-295c.webp) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/01/pantone-color-chip-1785-c.webp) Quick search for our logo colours. ## CMYK as translation, not failure CMYK, by contrast, *is* a translation. Given an RGB or HEX value and a colour profile, the conversion to CMYK is deterministic and mathematical. The same input will produce the same output. But it is also constrained. CMYK lives in a smaller colour space. Ink on paper reflects light; it does not emit it. Some colours simply cannot survive the translation intact. So when people say CMYK “flattens” colours, that is not a technical failure. It is a physical reality. CMYK does not fail to express richness. It expresses the maximum richness that can be reproduced reliably. That reframing changed everything. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/01/IMG_1964-1.jpeg) CYMK ## What happened to the colours in that book With that understanding, the comment about *A Dictionary of Color Combinations* clicked into place. Earlier editions were closer to pigment-first thinking. Colour as material, ink, paper, light. Later editions prioritise reproducibility. Standardisation. Portability across printing processes. The colours did not become flatter by accident. They became legible to a system. That is not degradation. It is translation across worlds. And once you see that, the book does not lose its value. It gains context. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/01/IMG_1967.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/01/IMG_1969.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/01/IMG_1970.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/01/IMG_1972.jpeg) CMYK at work in the book. Magnification with iPhone. ## You can only define colour inside a system This is the core insight I wish I had earlier: You can only ever define colour *inside* a system. Moving between systems always requires judgement. RGB, CMYK, Pantone, pigment. None of them are wrong. They just answer different questions. If you start digital-first, you postpone certain decisions. When you later move to print, you are not uncovering hidden truths. You are making assumptions explicit. That is not a mistake. It is a normal sequence. 0:00 /0:08 1× RGB at work ## Where this leaves me This may sound like a narrow design concern, but for me it was a genuine moment of discovery. Not because I finally “understood print”, but because a whole area that used to feel fuzzy suddenly became legible. Colour stopped being something that either worked or didn’t, and became a set of choices, translations, and constraints that I can reason about. That shift is oddly energising. It raises better questions. It opens up new ways of thinking about branding, reproduction, and design decisions I had previously postponed or avoided. It also makes me look differently at objects like that colour dictionary, not as flawed or diminished, but as situated in a particular system with its own logic. I still love the book. I still use it. I just enjoy it more now that I understand what kind of space it occupies. And perhaps that is the real pleasure here: not having arrived at an answer, but having gained a new way of seeing, with many more directions to explore. ### Further reading - [The difference between slides and photos](https://rob-hoeijmakers.ghost.io/the-difference-between-photos-and-slides/) - [The miracle of your screen](https://rob-hoeijmakers.ghost.io/three-colors-of-light-and-a-drop-of-water/) - [Blue LED changed the world](https://rob-hoeijmakers.ghost.io/invention-of-blue-led/) ### AI literacy: from definition to practice URL: https://rob-hoeijmakers.ghost.io/ai-literacy-from-definition-to-practice/ Last updated: 2026-05-10T08:53:41.000Z For weeks, I found myself circling the same question: *what is AI literacy, really?* Not in the abstract sense. I could list skills, risks, competencies. Plenty of reports do that already. But in conversations with organisations, that approach consistently led to friction. People would nod, agree in principle, and then stall. The more I tried to pin AI literacy down as a clearly defined thing, the more it seemed to slip away. What eventually helped me get unstuck was a shift in perspective: stopping the attempt to define AI literacy as a stable concept, and instead asking how it actually shows up when organisations try to work with AI in practice. That changed everything. ## Where AI literacy actually appears AI literacy rarely enters organisations through enthusiasm. It appears at pressure points. [Procurement conversations.](https://rob-hoeijmakers.ghost.io/the-hot-potato-of-compliance/) Legal reviews. Governance discussions. Moments where someone asks: *are we comfortable with this?* or *can we explain this if needed?* It is not introduced as a learning objective. It emerges as a requirement for making decisions that feel defensible. That already tells us something important: AI literacy is not primarily about knowledge acquisition. It is about judgement under uncertainty. [The Hot Potato of ComplianceFrom GDPR to the EU AI Act, a recurring pattern emerges: European regulation lands in procurement, spreading responsibility, caution, and friction.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-455.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/2FAE4B6B-ECF4-4870-AED8-93282D1EB0C4-1.jpeg)](https://rob-hoeijmakers.ghost.io/the-hot-potato-of-compliance/) ## The unease behind the term Much of the discomfort around AI literacy comes from its vagueness. There is no licence to obtain, no test to pass, no moment where you are officially “done”. And yet, there is a growing sense that organisations will be held to account for how they use AI. That creates a particular kind of anxiety: fear without a clear trigger. People imagine regulators, enforcement actions, penalties. In reality, that picture is usually exaggerated. There is no AI police. There are no random inspections looking for insufficient literacy. What does exist is retrospective scrutiny. When something goes wrong, when a decision is challenged, when a system becomes visible to the outside world, organisations are expected to explain themselves. AI literacy is inferred afterwards, from how decisions were made, documented, and governed. That is uncomfortable, but it is also familiar. This is how responsibility works in many European regulatory contexts. ## Regulation without instruction The [EU AI Act](https://rob-hoeijmakers.ghost.io/learning-to-work-with-the-eu-ai-act/) reinforces this pattern rather than replacing it. It raises expectations, but it does not prescribe competence in procedural terms. It does not tell organisations exactly how to be “AI-literate”. Instead, it assumes that organisations deploying AI take responsibility for understanding what they are doing. That ambiguity is intentional. Fixed definitions would age badly. Context matters too much. The result is not clarity, but latitude. And latitude shifts responsibility inward. ## A practical reality check One of the most grounding ways to understand this dynamic is to look at how risk is treated outside regulation. There is no dedicated “AI literacy insurance”. AI-related risk is absorbed through existing instruments: professional liability, errors and omissions, cyber and legal expenses coverage. Insurers do not ask whether an organisation is literate in the abstract. They look for signs of seriousness: governance structures, training efforts, decision documentation, escalation paths. Not because they care about ethics, but because they need to price uncertainty. That tells us something important. AI literacy is already being interpreted operationally, even if it is not labelled as such. It affects insurability, risk exposure, and the ability to respond when challenged. This is not about avoiding risk. It is about keeping risk bounded. ## From fear to agency Once that becomes visible, something shifts. AI literacy stops feeling like an undefined threat and starts to look like a way of regaining agency. Not control, but orientation. It becomes possible to say: we understand the system’s role, we know its limits, we have thought about failure modes, and we remain accountable for outcomes. That does not make organisations safe. But it makes them *defensible*. And defensibility is often what determines whether experimentation feels possible at all. ## Literacy as organisational maturity Recent research reinforces this view. Studies such as the [*EY European AI Barometer 2025*](https://www.ey.com/nl%5Fnl/functional/forms/download/ey-european-ai-barometer-2025) show that while AI adoption is accelerating across Europe, AI-specific governance, risk management, and training remain uneven and underdeveloped. Only a minority of organisations conduct formal AI risk assessments, even as many expect substantial operational and financial impact. What is often framed as a skills gap is, in reality, a maturity gap. AI literacy, in this sense, is not about educating individuals in isolation. It is about aligning ambition, capability, governance, and responsibility across the organisation. It is also a signal. For vendors, partners, and regulators alike, it indicates how much foundational work has already been done, and how much remains. [EY European AI Barometer 2025Het EY European AI Barometer 2025 rapport onthult de dynamische impact van Artificial Intelligence op verschillende sectoren en de arbeidsmarkt, en belicht zowel de kansen als de uitdagende obstakels van AI-adoptie. Ontdek hoe organisaties zich kunnen bewegen in dit transformerende landschap en het potentieel van AI voor groei kunnen benutten.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/ey-favicon.jpg)EY![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/ey-ai-barometer-meta-20250702.jpg)](https://www.ey.com/nl%5Fnl/functional/forms/download/ey-european-ai-barometer-2025) ## Why I am comfortable with where this is heading After spending time with the darker scenarios first, I find myself broadly comfortable with the direction AI literacy is taking in Europe. Not because it is well defined, but because it reflects something reasonable: that powerful systems should not be deployed casually, and that responsibility should not dissolve into tooling. AI literacy is not a brake on innovation. It is the condition under which responsible experimentation becomes possible. By forcing organisations to think through their use of AI, it slows things down just enough to allow understanding to catch up with capability. That feels less like a burden, and more like a necessary recalibration. ## Where this leaves us AI literacy is not something we can neatly define and move on from. It is something that evolves as practice evolves. Trying to pin it down too early leads to paralysis. Watching how it manifests in real organisational behaviour brings clarity. For me, that shift from definition to practice was the key to getting unstuck. And it opens the space to engage with AI more openly, more confidently, and, paradoxically, more creatively than before. ### Further reading - [Reskilling the Mind: Europe’s Next Transition](https://rob-hoeijmakers.ghost.io/reskilling-the-mind-europes-next-transition/) ### Kobo and Kindle solve different problems URL: https://rob-hoeijmakers.ghost.io/kobo-and-kindle-solve-different-problems/ Last updated: 2026-05-10T08:53:41.000Z For a long time, I treated Kobo and Kindle as roughly the same thing: e-readers with different ecosystems. Only when I started using them side by side did it become clear that they are built around very different setups. Not different features. Different models. Once you see that, the choice becomes much simpler. Quick video on the topic. ## Two ways of getting content to a reader With Kobo, the device is the centre. You connect it to a computer, copy files onto it, and that’s where they live. The book is on the device, and the device is the destination. The cloud, if you use it at all, is secondary. With Kindle, the account is the centre. You don’t really put files on the device. You send them into Amazon’s system, and the device pulls them down. The book lives in the cloud first. The Kindle is one of several ways to read it. That single difference explains most of the experience. ## Kobo: device-first, edge-based Kobo works best when you treat it like a quiet, durable object. You put books on it deliberately. They stay there. Reading position and annotations live on the device. With tools like Calibre, you can run it almost entirely outside Kobo’s ecosystem. It has more friction. Often a cable. Often a computer. But that friction fits long-term reading. It assumes you care about the book, want to keep it, and might return to it. Kobo feels like a library. [Connecting My Kobo Directly to an iPhone 16 ProI connected my Kobo e‑reader to my iPhone 16 Pro with a USB‑C cable. It worked instantly. No adapter, no fuss but just a quiet moment of satisfaction.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-457.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_4390-4.jpeg)](https://rob-hoeijmakers.ghost.io/connecting-my-kobo-to-iphone/) ## Kindle: account-first, cloud-operated Kindle works best when you treat it like a service. You send documents to your account. They appear wirelessly. They sync across devices. You read them, and you can remove them again without much ceremony. That makes it ideal for: - PDFs - Reports - Business books - Things you read once or twice It also means you depend on Amazon being there. The system only really works because Amazon operates it end to end. Kindle feels like an inbox. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/01/IMG_1957.jpeg) Kindle notices reading progress on a different device. ## One extra thing that matters: screens Kindle is not just an [e-ink reader](https://rob-hoeijmakers.ghost.io/electronic-paper-and-digital-ink/). It’s a delivery system. The same document can be read: - On e-ink, quietly and with low energy use - On a phone or tablet, with colour and zoom - On a desktop screen, where layout really matters That is especially useful for professional PDFs. Many of them are designed visually, with columns, charts, and typography doing part of the work. With Kindle, you can switch between those forms without changing how the document is delivered. Kobo doesn’t really do that. There, the device *is* the destination. ## Where I landed Once I stopped trying to make one device do everything, it became obvious. - Books I want to keep live on the Kobo. - Documents and work-related reading go to the Kindle. Kobo is where books stay. Kindle is where documents pass through. The devices didn’t change. My expectations did. ### Further reading - [Electronic paper and digital ink explained](https://rob-hoeijmakers.ghost.io/electronic-paper-and-digital-ink/) - [Reading Webpages on Your Kindle: A Simplified Process](https://rob-hoeijmakers.ghost.io/reading-webpages-on-your-kindle/) ### Word of the year: model URL: https://rob-hoeijmakers.ghost.io/model-meaning-etymology/ Last updated: 2026-05-10T08:53:41.000Z When I looked back at the words that kept appearing in my work this year, one stood out more than I expected: *model*. Not because it felt central or dramatic, but because it was everywhere. In AI, obviously. In strategy documents. In conversations about organisations, processes, responsibility. Even in fairly ordinary moments. The word kept resurfacing, quietly doing work. At some point I realised I was using it constantly without really stopping to think what I meant by it. And the more I tried to pin it down, the more elusive it became. Not vague, but flexible. Almost suspiciously so. This piece is an attempt to understand that flexibility. Not to define *model*, but to ask: how did we get here, and why does this word fit so many domains so well? ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/12/blogging-year-in-review-2025-wordcloud-1.png) 2025 Wordcloud for my blog ## The everyday meanings we hardly notice If you search for *model*, you will likely land on fashion models first. Photo models. People. That might seem like a distraction, but it is actually a useful starting point. A model here is an example. Something you look at in order to orient yourself. This is what it looks like. This is what it could be. We use the word like this all the time. A role model. A model student. A model answer. In all these cases, the model is not a description of reality, but a reference point. It reduces complexity by embodiment. Instead of rules or explanations, you get an instance you can copy, approximate, or respond to. Alongside this, there is another everyday sense that feels more abstract. Scale models. Maps. Diagrams. Calendars. Dashboards. These are not things you imitate, but things you use to navigate. They deliberately leave things out so you can act. A map is not the territory, but it is still indispensable. Already, the word is doing two different jobs: showing what something looks like, and helping you move through complexity. That tension turns out to be important. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/12/image-1.png) ## A small detour into etymology The word *model* comes from the Latin *modulus*, a diminutive of *modus*. *Modus* means measure, manner, way, method. Not an object, but a way of doing something. A pattern that makes action possible. *Modulus* is a small measure. A manageable unit. This matters more than it might seem. From the start, a model was not meant to be the world in miniature, but a chosen scale. A way of handling something too large, too complex, or too messy to grasp directly. A model is already an admission: we cannot deal with everything at once. That quietly underpins almost every use of the word today. [Model - Etymology, Origin & Meaning“likeness made to scale; architect’s set of designs,” from French modelle (16c., Modern… See origin and meaning of model.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon-37.ico)etymonline![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/app-icon-medium.png)](https://www.etymonline.com/word/model) ## Two paths that never really split Historically, *model* developed along two closely related paths. On the one hand, the model **as exemplar**. A sculpture model. A pose. A prototype. Something you look at and emulate. On the other hand, the model **as representation**. A plan, a sketch, a proportional guide. Something that captures relationships rather than appearance. These were never cleanly separated. A sculptor’s model was both something you looked at and something you built from. It guided action without claiming to be the final thing. That dual role has always been there. The confusion around *model* today is not new. It is inherited. ## From craft to science When science and mathematics adopted the word, they did not change its meaning so much as tighten it. A mathematical model is a reduction of reality, expressed in symbols, designed to preserve certain relationships while ignoring others. An economic model does the same with incentives, behaviour, and constraints. These models are explicit about what they leave out. They are tools for thinking, not claims to completeness. This is why scientific models are always accompanied by assumptions, boundaries, and caveats. Not because they are weak, but because their strength lies precisely in being limited. They help you see a system. They do not pretend to be it. ## When models start to build things Engineering shifts the balance. Here, models are no longer only aids to understanding. They become instruments of construction. A [blueprint](https://rob-hoeijmakers.ghost.io/blueprints/) is a model. A [data schema](https://rob-hoeijmakers.ghost.io/etymology-of-data/) is a model. A software architecture is a model. **Change the model, and you change the system.** At this point, the model stops being merely epistemic and becomes operative. Errors are no longer just misleading. They propagate. This is where the stakes rise, and where the word *model* starts carrying real authority. Not because it is more accurate, but because it has consequences. ## Language models sit on the fault line This long history helps explain why *model* feels so overloaded in AI. A language model brings all these meanings together. - It is a statistical reduction of language, trained rather than reasoned into existence. - It produces exemplars: plausible sentences, answers, styles. - It is deployed as an operational system. - And it is used by people as a way to explore, understand, and make sense of domains. It is, at the same time: **something that generates behaviour, and something we use to think with.** This collapses an old distinction between models that help us see systems and models that are systems. No wonder the word feels unstable here. It is being asked to do everything at once. Much of the current confusion around AI is not technical, but semantic. We slide between treating the model as a tool for exploration and treating it as an authority. Between using it as a map and mistaking it for the territory. The word *model* quietly enables that slide. 💡 Next to **model*, I often use the word **system*. For the most part, I use **AI model* to refer to a trained computational artefact, and **AI system* to refer to what happens once that artefact is embedded in data flows, software, workflows, and decisions. A model can be seen as a system in miniature, but impact emerges at the level of the system. I will return to the notion of **system* later, as it is an interesting and consequential term in its own right. ## Why this word keeps appearing Looking back, I think this explains why *model* surfaced so often for me this year. It is a word that allows us to work with complexity without fully resolving it. It lets us act, decide, and build while acknowledging that what we are doing is partial and provisional. At the same time, it carries a risk. A model can easily stop being a choice and start feeling like reality. Especially once it is embedded in systems, dashboards, policies, or software. The problem is rarely the model itself. It is forgetting that it is a model. Seen this way, *model* is not just a technical term, but a cultural one. It sits at the boundary between understanding and authority, between representation and action. That is probably why it is almost everywhere now. And why it is worth pausing over, at least once, to ask what we are really doing when we invoke it. Not to pin the word down, but to keep it honest. ### Further reading - [Do dashboards protect against flying mud?](https://rob-hoeijmakers.ghost.io/why-dashboards/) - [From Blueprints to Bytes: How Original Designs Shaped Modern Technology](https://rob-hoeijmakers.ghost.io/blueprints/) - [Where does the word “benchmark” come from?](https://rob-hoeijmakers.ghost.io/where-does-the-word-benchmark-come-from/) ### Direct Debit: US versus EU URL: https://rob-hoeijmakers.ghost.io/direct-debit-us-europe/ Last updated: 2026-05-10T08:53:42.000Z A video has been circulating on social media in which an American banker gives very firm advice: never use a debit card if you can avoid it. Use credit cards for everything. Debit cards, he argues, expose your bank account. Credit cards protect you by putting distance between merchants and your money. The advice is confident, widely shared, and in a US context, largely reasonable. > This is why you should Never use a Debit card ‼️ [pic.twitter.com/PfwtEvoSAO](https://t.co/PfwtEvoSAO) > > — Manly Mentor (@manly\_mentor) [December 28, 2025](https://twitter.com/manly%5Fmentor/status/2005188279638151467?ref%5Fsrc=twsrc%5Etfw) From a European perspective, it also feels oddly off. Not wrong exactly, but mismatched. To understand why, you have to look past the cards themselves and focus on something else entirely: direct debit. ## Cards are not the problem Much of the discussion around the video focuses on debit cards versus credit cards. That is understandable, but it misses the real fault line. Card payments, whether debit or credit, work in broadly similar ways across regions. You initiate the payment. The merchant’s bank talks to your bank via a card network such as Mastercard, Visa, or Amex. Your bank checks whether the payment is allowed, based on balance or credit line, and settlement follows later. The card networks impose rules, including dispute and chargeback mechanisms. There are differences in detail, but structurally this is familiar territory on both sides of the Atlantic. The real difference sits elsewhere. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/12/massimiliano-graph1.jpg) ## Direct debit is a different rail entirely Direct debit works the other way around. Here, the merchant initiates the payment, based on permission you have previously given. This is not a card payment at all. It is designed for recurring charges: subscriptions, utilities, rent, memberships. In the Netherlands, this is incasso. Across Europe, it runs via SEPA Direct Debit. In the United States, it runs via ACH. On paper, these systems look similar. In practice, they behave very differently when something goes wrong. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/12/tla1437556783472.png) SEPA Direct Debit ## ACH versus SEPA: same idea, different assumptions ACH is an old, pull-based system. Once a merchant has your bank details, payments can be initiated with relatively weak ongoing checks. Disputes exist, but they are slow, procedural, and often stressful. Money can leave your account first and only later become a subject of investigation. This is the experience the American banker is speaking from. In that world, allowing direct access to your bank account is genuinely risky. Using credit cards instead is not a lifestyle choice, but a defensive strategy. SEPA Direct Debit, by contrast, is built on a different set of assumptions. In Europe: - Direct debit requires an explicit mandate - Mandates can be revoked easily - Unauthorised debits can be reversed predictably - Specific merchants can be blocked directly in a banking app The system assumes that mistakes, abuse, and misunderstandings will happen and treats reversibility as a feature, not an exception. This difference in default protections changes everything. 💵 ****What is ACH?** In the United States, most direct debit–style payments run through the Automated Clearing House (ACH). It is a batch-based bank transfer system that allows merchants to pull money directly from bank accounts once they have the relevant details. While reliable at scale, ACH was not designed for real-time control or easy reversal. Disputes are possible, but often slow and procedural, and money typically leaves the account before issues are resolved. This is the context in which American warnings about debit and bank access make sense. ## Why American advice sounds extreme in Europe When Americans say “never use debit”, they are often talking about direct debit exposure, even if they describe it in terms of cards. When Europeans hear the same advice, it sounds disproportionate, because direct debit here does not behave like that. The underlying system is simply more forgiving. This is why advice that feels like common sense in one place can sound like fear-mongering in another. The disagreement is not really about financial literacy. It is about infrastructure. ## The cultural layer on top There is also a cultural difference that reinforces the gap. In much of Europe, credit cards are still associated with borrowing money you do not yet have. Even people who use them tend to settle payments immediately or within days. Interest is avoided. Building a credit profile is not a goal. In the US, credit cards have become a core consumer protection mechanism layered on top of weaker direct debit infrastructure. Rewards, chargebacks, and credit history are part of the same system. The same tool exists in both places, but it plays a different role. ## Travel is the exception, not the rule There is one area where the American advice translates almost perfectly: travel, especially outside the EU. Car rentals, hotels, deposits, and unfamiliar legal environments introduce distance and risk. In those situations, limiting direct access to your funds makes sense. Credit cards provide a buffer that debit and direct debit cannot. Here, the logic is universal. ## What this is really about This is not a story about which card to use. It is about understanding how close a payment method is allowed to get to your bank account, and what happens when something goes wrong. Systems encode assumptions about trust, error, and recovery. Those assumptions differ by region, and advice only makes sense inside the system it was designed for. Once you see that, the viral video stops being provocative and starts being contextual. Not wrong. Just local. ### Further reading - [The Silent Upgrade: What Really Changed When Our Bank Replaced Our Cards](https://rob-hoeijmakers.ghost.io/the-silent-upgrade-mastercard-debit-transition/) - [Understanding the layers behind digital payments](https://rob-hoeijmakers.ghost.io/understanding-digital-payments-layers/) ### Blogging year in Review 2025 URL: https://rob-hoeijmakers.ghost.io/blogging-year-in-review-2025/ Last updated: 2026-05-26T19:38:47.000Z ## Writing, agency, and staying coherent in a changing landscape Looking back at 2025, what stands out most is not growth curves or individual articles, but the fact that writing itself remained enjoyable, even as almost everything around it shifted. That may sound trivial. It isn’t. For anyone writing about technology, AI, and digital life today, the conditions are volatile. Platforms change, discoverability shifts, and attention is increasingly fragmented, not just by other writers, but by AI systems themselves. Against that backdrop, continuing to write with curiosity and pleasure already matters. This year confirmed something important: **the system I work in still works**. ## Why this blog exists This blog serves three purposes. First, it is a place to **learn in public**. Writing is the thinking tool here. It allows exploration, testing understanding, and staying curious without pretending to have final answers. Second, it functions as a **relational surface**. People find an article, recognise something of their own questions, and reach out. That mechanism still works, even as the routes people take to get here continue to change. Third, it helps build a **position** in the space where I actually work: European digital and AI strategy. Not as branding, but as a long-form signal of how I reason, what I care about, and where I place responsibility. Seen from that perspective, 2025 was not a straightforward year. ## When infrastructure breaks Early in the year, a domain change [disrupted discoverability](https://rob-hoeijmakers.ghost.io/the-end-of-google-search-as-we-know-it/) almost completely. Traffic dropped sharply. Articles were written into what felt like a vacuum. Later, I reverted to an older, long-standing domain and had to rebuild trust and visibility from there. It was uncomfortable, and clarifying. Infrastructure matters, more than most writers like to admit. At the same time, the environment shifted more broadly. Google search sent less traffic. Social media felt saturated. In parallel, articles were increasingly surfaced and used inside large language models and AI tools, often without showing up as traditional referrers at all. Being read no longer always meant being visited. That changed how success looked. Visibility became less about spikes and more about durability. Articles that explain, frame, or structure ideas continued to find readers, sometimes months later, sometimes indirectly. [The End of Google Search (as we know it)Google didn’t warn me. It just erased my blog. What looked like a bug turned out to be a glimpse into the future of search—and it’s not built for us anymore.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-441.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_2069-6.jpeg)](https://rob-hoeijmakers.ghost.io/the-end-of-google-search-as-we-know-it/) ## Workflow, tools, and enjoyment Throughout all of this, the act of writing itself remained stable and enjoyable. I still work in [Markdown](https://rob-hoeijmakers.ghost.io/markdown-google-docs/). I still publish with Ghost. I still combine text with photography, not because photography is my profession, but because it keeps my thinking grounded and observational. The workflow held up even as external conditions changed. This was also the year where **power tools became genuinely useful**. Not in a way that pushed me to produce more, but in a way that increased control. Research, drafting, and editing became calmer and faster, freeing attention for judgement. ## From volume to agency Part of this year involved relatively high output. That was exploratory: mapping the terrain, stress-testing questions, seeing which themes persisted. That phase is now largely complete. What changed this year was not just efficiency, but **agency**. With better tools and more stable processes, it became easier to make deliberate choices: what to pursue, what to pause, and what not to do at all. Less volume and more depth is not a productivity goal. It is an agency decision. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/12/blogging-year-in-review-2025-wordcloud.png) Wordcloud on the year 2025\. ## Structure and subtraction Midway through the year, it became clear that the accumulated writing needed better internal structure. A new taxonomy emerged. Some themes were sharpened, others explicitly named: European digital sovereignty, the[ future of work](https://rob-hoeijmakers.ghost.io/tag/future-of-work/), finance and payment systems, and more foundational questions around infrastructure and agency. At the same time, some things were deliberately removed. A [Dutch-language sibling blog](https://robhoeij.nl) was archived, not out of disinterest, but because it could no longer be maintained at the level it deserved. Writing in Dutch continued instead through [guest contributions](https://www.enreach.com/nl/news-knowledge/blog?category=21). Photography was clarified as foundational but not professional. Social profiles were refocused rather than multiplied. These were not optimisations. They were acts of alignment. ## What made the year successful Despite the disruptions, readers kept finding their way here. Not in explosive numbers, but steadily, and often with thoughtful responses. New relationships formed. Existing ones deepened. Client work grew in ways that matched the direction of the writing. That alignment matters more than raw traffic ever could. If there is a single thread running through 2025, it is this: the blog continued to function as **a stable learning and connection surface in an unstable digital landscape**. ## Looking ahead That changes the posture for the year ahead. Less volume, more depth. Building on established themes rather than constantly extending the map. Allowing for occasional exploratory pieces, but focusing primarily on consolidation, synthesis, and clarity. The landscape will keep changing. Search will keep evolving. AI systems will mediate more reading and discovery. What can remain stable is the practice: writing to learn, sharing to connect, and exercising agency in where attention and effort go. **Thanks for reading, following along, and reaching out when something resonated. I’m looking forward to continuing the conversation in the year ahead.** --- A selection of well read articles from the year 2025: --- ### Further reading - [Running a Local LLM on Your iPhone](https://rob-hoeijmakers.ghost.io/running-a-local-llm-on-your-iphone/) - [When Research Turns Into Action: Understanding Deep Research vs. Agent Mode in ChatGPT](https://rob-hoeijmakers.ghost.io/deep-research-vs-agent-mode/) - [Why AI Image Editing Is Often Rebuilding](https://rob-hoeijmakers.ghost.io/why-ai-image-editing-is-often-rebuilding/) - [Google One vs Google Workspace: Where do you get real AI and Gemini Advanced?](https://rob-hoeijmakers.ghost.io/google-one-vs-google-workspace-for-gemini/) - [From Mini to Pro: My Experience Switching to the iPhone 16 Pro](https://rob-hoeijmakers.ghost.io/iphone-mini-to-pro/) ### The Hot Potato of Compliance URL: https://rob-hoeijmakers.ghost.io/the-hot-potato-of-compliance/ Last updated: 2026-05-10T08:53:42.000Z I have been noticing a pattern for some time now, without quite knowing how to name it. As an entrepreneur, and earlier inside larger organisations, I kept encountering the same kind of friction. Deals that looked straightforward would slow down late in the process. Not because of price or technology, but because of questions that appeared at the moment contracts were about to be signed. I first saw this clearly with GDPR. Now, with the [EU AI Act](https://rob-hoeijmakers.ghost.io/learning-to-work-with-the-eu-ai-act/) and the emerging idea of **AI literacy**, I recognise the same dynamic returning. ## Where regulation actually lands We tend to picture regulation as something that happens after the fact. Inspectors, fines, enforcement actions. In practice, much of European digital regulation lands much earlier. It lands at **procurement**. That is the point where responsibility becomes fixed. Once a contract is signed, an organisation cannot easily claim ignorance or shift liability elsewhere. So uncertainty flows downhill until it reaches the place where a signature is required. Procurement becomes the final gate, not by design of the organisation, but by design of the law. 💡 Procurement is the structured way organisations decide what to buy, from whom, and under which conditions. It covers the full journey of identifying a need, selecting suppliers, running tenders, negotiating contracts, managing risks, and checking that money is well spent. In many companies it acts as a gatekeeper between internal ambition and external markets, balancing cost, quality, legal obligations, and strategic fit. Done well, procurement protects organisations from financial, legal, and ethical exposure while helping them secure the best long-term partnerships. ## A familiar pattern from GDPR With [GDPR](https://rob-hoeijmakers.ghost.io/gdpr/), this became highly visible. The regulation was intentionally abstract. Concepts like “appropriate measures” and “demonstrable compliance” were left open. That openness created space for interpretation, but also for anxiety. Organisations responded by internalising that uncertainty. Procurement processes became more defensive. Vendor questionnaires expanded. Standard clauses multiplied. Sales cycles slowed, especially for smaller suppliers. This was often framed as bureaucracy. Seen differently, it was regulation working as intended. ## Spreading responsibility, spreading risk [European regulation](https://rob-hoeijmakers.ghost.io/playing-the-eu-game/) often spreads responsibility across the market rather than centralising enforcement. That is efficient. Supervisors do not need to inspect everything. Organisations are incentivised to police themselves, and each other, through contracts. But there is a side effect. Responsibility spreads, and so does risk aversion. Procurement teams often end up holding a kind of hot potato. They do not own the technology, nor the policy goals, yet they are asked to absorb uncertainty on behalf of the organisation. Saying “yes” can feel riskier than saying “no”. ## Why AI intensifies this effect With [AI](https://rob-hoeijmakers.ghost.io/tag/ai/), this dynamic becomes sharper. AI systems are complex, probabilistic, and often hard to explain fully, even by their creators. Asking procurement to confidently assess such systems is asking them to manage uncertainty they are not structurally equipped to resolve. **The result is not outright rejection, but friction. Slower decisions. More documentation. More caution.** From experience, I see the same lines forming again. Not yet fully drawn, but recognisable. ## A very European logic Regulation via procurement is not new. It has existed for decades in areas like environmental standards, safety, and finance. What feels distinctly European is how deeply this logic is now applied to digital infrastructure and AI. Responsibility is embedded ex ante, at the moment of decision, rather than enforced primarily after harm has occurred. Whether this ultimately hampers innovation or builds long-term trust is still an open question. Probably both. What seems clear to me is this: when procurement slows things down, it is not simply being difficult. It is where policy becomes real. And perhaps that is exactly the point. --- ### Further reading - [Learning to Work with the EU AI Act](https://rob-hoeijmakers.ghost.io/learning-to-work-with-the-eu-ai-act/) - [Introduction to GDPR](https://rob-hoeijmakers.ghost.io/gdpr/) - [Playing the EU Game: Learning the Digital Rules by Getting in the Game](https://rob-hoeijmakers.ghost.io/playing-the-eu-game/) ### Why AI Image Editing Is Often Rebuilding URL: https://rob-hoeijmakers.ghost.io/why-ai-image-editing-is-often-rebuilding/ Last updated: 2026-05-10T08:53:43.000Z ## How colour, text, and “fixes” revealed what generative images really are It started, as these things often do, with something that felt right at first glance. I shared a colourised version of a famous photograph of The Hague from around 1860\. A quiet street, soft tones, a sense of calm. It looked like a pleasant summer day. Only later did I notice something odd. The trees were bare. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/12/G7lMdh7XwAAmF9j-2.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/12/G7qKEiXWMAEesJP-3.jpeg) ## A summer that was not there The impression of summer came entirely from colour. Greenish hues suggested foliage, warmth, life. But when I looked more carefully, there were no leaves at all. Just branches. The colourisation had not lied outright, but it had *implied*. It filled in what usually makes sense, not what was actually present. At that point, this still felt like a limitation of colourisation. Interesting, but contained. ## Where text breaks the spell Then someone commented that they had tried to read the street sign. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/12/G7lMdh7XwAAmF9j-2-1.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/12/G7qKEiXWMAEesJP-3-1.jpeg) 2\. AFD, Hooge Nieuwstraat. In the colourised image, the sign had become unreadable. Random glyphs, shapes that resemble letters but form no words. In the original black-and-white photograph, the same sign is still legible to a human eye. Not perfectly crisp, but recognisable. This was the first moment where the problem no longer felt cosmetic. If the information was there in the original, why was it lost? ## “It rebuilds the image” Someone replied, almost casually: an AI does not really colour in an image. **It rebuilds the photo entirely.** What you see is a generated image that resembles the original as closely as possible. That sentence stuck. If that were true, a lot of things suddenly made sense. The invented sense of season. The loss of typographic fidelity. The overall plausibility paired with local incoherence. But I did not yet *know* it. So I started testing. > Den Haag, zomerdag 165 jaar geleden. > > Beetje kleur aangebracht. Je ziet beter dat de zon schijnt en het laat het Mauritshuis mooi uitkomen. [pic.twitter.com/K3yW1xgPH7](https://t.co/K3yW1xgPH7) > > — Rob Hoeijmakers (@robhoeij) [December 8, 2025](https://twitter.com/robhoeij/status/1998061200215822735?ref%5Fsrc=twsrc%5Etfw) ## Famous images behave differently I shared an image of a well-known book with ChatGPT and began experimenting. First, I asked for a small edit. Almost nothing seemed to change. Then I asked for a translation of the text. This time, the result was strikingly coherent. At one point, ChatGPT effectively recognised the book. It was not merely extracting text from the image, but identifying the object itself and drawing on prior knowledge to rebuild it accordingly. That moment was revealing. The system was not reading the text purely from pixels. Recognition anchored the reconstruction. Editing appeared to work here because the image belonged to a familiar visual and cultural pattern. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/12/IMG_1341.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/12/9D363EF9-F5AC-47B9-8487-F34DC4646D6F.png) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/12/2B048F3A-42B5-484F-9CFE-E17D9C7EFA85.png) The original photo, the true edit from yellow to green, and the regenerated translation. Check out the change of font type. ## When recognition disappears To remove that anchor, I changed the input. I took a photograph myself: a screen, shot at an angle, with reflections and distortion. No known layout. No shared cultural reference. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/12/69F0E2A4-DC83-4591-BEBF-7CA14A6B58DC.png) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/12/12934E04-FC9B-44F1-8C33-394869896EE8.png) Convincing at a glance, but wrong in the details: the image is regenerated, not edited, losing pixel noise, moiree patterns, precise perspective, and the fidelity of that specific screen and moment. A human reader can still make sense of such an image. We are remarkably tolerant of noise, perspective, and imperfection. When ChatGPT regenerated the image, the result was visibly weaker. Layout drifted. Text degraded. The seams became obvious. This contrast was decisive. The system was not preserving the original image and modifying it. It was rebuilding it, and it did so far more effectively when it already “knew” what it was looking at. ## Typography as a diagnostic tool Typography turned out to be the clearest signal throughout this process. Letters are not textures. They are symbols. Humans can tolerate blur and distortion and still recognise them. Generative image models, trained primarily on visual similarity, cannot reliably preserve symbolic correctness unless the text is already familiar or explicitly generated as text. That is why typography collapses first. It is where regeneration reveals itself most clearly. ## What “editing” usually means now By this point, the pattern had stabilised. Most consumer-facing AI image tools do not edit images in the traditional sense. They interpret an image, combine it with a prompt, and generate a new image that fits both. Sometimes that new image is almost identical. Sometimes it diverges in telling ways. True pixel-preserving edits exist, but they are the exception. Regeneration is the default because it is flexible, powerful, and forgiving. Once you see that, the confusion lifts. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/12/IMG_1581.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/12/2e7d7439-d4bc-4524-b3c9-ca6ce63fb747.png) Here is colorisation where the image seems to be edited rather than regenerated. ## Learning to see the machine What changed for me was not my opinion of these tools, but my way of looking at their output. I stopped asking why something was wrong, and started noticing *where the system was guessing*. I could see when it was anchored in recognition, and when it was improvising. That shift, from frustration to understanding, is quiet but profound. It is not about demystifying AI. It is about aligning our expectations with the kind of machine we are actually dealing with. And once that alignment happens, the images stop being misleading. They become instructive. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/12/IMG_5064.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/12/d664a851-ed47-4b62-adc2-4fc16e52e82f.png) Here I tested ChatGPT and asked it to change from wood to metal. As you can see it carried on the handwritten note (but lost the second bit of Scotch). ### Further reading - [The AI Image Technology Landscape for images: Generation, Alteration, and Recognition](https://rob-hoeijmakers.ghost.io/ai-image-technology-landscape/) ### Understanding the layers behind digital payments URL: https://rob-hoeijmakers.ghost.io/understanding-digital-payments-layers/ Last updated: 2026-04-25T13:58:59.000Z *A companion piece to “*[*How Money Moves*](https://rob-hoeijmakers.ghost.io/how-money-moves/)*”* Over the past months I’ve been exploring the mechanics behind digital payments. It began with a simple curiosity about how Apple Pay and Google Wallet actually work, what happens when you tap a terminal in a shop, and why online payments behave differently from card-present transactions. That curiosity eventually led to *How Money Moves*, an attempt to sketch the broader landscape and offer an intuitive sense of its underlying structure. This new piece continues that exploration. After writing the first article, news broke that Mollie intends to acquire GoCardless, a company specialising in bank-to-bank payments. It made me realise that the earlier map could be coloured in more sharply. The physical and online worlds are converging; companies that once occupied narrow roles now stretch across multiple layers. The Mollie–GoCardless story is a concrete moment within that larger shift. I’m not approaching this as a payments expert. I’m simply trying to understand how things fit together: which actors sit where, why so many layers exist, and why the distinction between customer-initiated and business-initiated payments turns out to be foundational. If *How Money Moves* traced the outline of the terrain, this piece adds a few contour lines. [Dutch fintech Mollie buys GoCardless for €1.1bnSifted previously reported Mollie was the frontrunner to acquire payments startup in August![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon-36.ico)SiftedTom Matsuda![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/GoCardless-CEO-Hiroki-Takeuchi-copy.jpg)](https://sifted.eu/articles/mollie-buys-gocardless) ## A simple model of the four layers Much of the confusion around payments comes from mixing up categories. It helps to strip the system back to four layers. They are not industry terms, but they reflect how the system behaves. ### **Layer 1\. The infrastructure** The deep foundation where value actually moves: card networks, bank-to-bank systems, and the rulebooks that govern them. Examples include **Visa** and **Mastercard** for card rails, **SEPA Credit Transfer** and **SEPA Direct Debit** for euro bank payments, and systems such as **ACH** (US) or **UPI** (India). Most businesses never interact with this layer directly. ### **Layer 2\. The processors** Institutions allowed to operate on the infrastructure. They submit transactions, perform authorisations, handle settlement, and take on risk. Examples include **Worldpay**, **FIS**, and **TSYS** for card processing, and **GoCardless** for Direct Debit and other bank-to-bank pull payments. Some processors specialise in cards, others in Direct Debit, others in specific regions or schemes. ### **Layer 3\. The service layer for businesses** The space most digital companies touch. Here sit the platforms that make it easier to accept payments without dealing with banks or networks directly. They provide a unified interface over underlying complexity. Examples include **Stripe**, **Adyen**, **Mollie**, **PayPal**, and **Checkout.com**, often bundling gateway, processing, reporting, and compliance into a single integration. ### **Layer 4\. The consumer-facing experience** Wallets, checkouts, and authentication flows. This is the layer users see. It shapes convenience and trust but does not move money. Examples include [**Apple Pay**](https://rob-hoeijmakers.ghost.io/apple-pay-apple-wallet-apple-card/), **Google Pay**, **PayPal Checkout**, and local methods such as **iDEAL** or **Bancontact**, which sit on top of lower layers rather than replacing them. Seen this way, the landscape becomes less opaque. You can begin to place companies by the part of the system they attach to, and understand their announcements as shifts within that structure rather than isolated moves. [Apple Pay, Apple Wallet, Apple Card: How Apple Reshaped How Money FlowsWhile writing about how money flows, a simple confusion between Apple Pay and Apple Wallet exposes the hidden layers of modern payments and power.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-430.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_1383-1.jpeg)](https://rob-hoeijmakers.ghost.io/apple-pay-apple-wallet-apple-card/) ## Two different payment logics: push and pull Another distinction helps the picture settle: the difference between *push* and *pull* payments. **Push payments** are initiated by the customer. They include card payments, iDEAL, bank transfers, and digital wallets. They work well for occasional purchases and checkout flows. **Pull payments** are initiated by the business, with prior consent. **Direct Debit is the canonical example.** These payments underpin subscriptions, utilities, insurance premiums, and much of B2B invoicing. They reduce friction and cost but require more trust and careful handling. For years, these two logics developed in parallel. Providers typically focused on one or the other. When companies began blending subscription models with one-off purchases, the separation became visible: two systems, two logics, sometimes two providers. Understanding this divide makes the Mollie–GoCardless news easier to interpret. One platform blending both logics is not a feature expansion but a structural change. ## How companies occupy different parts of the system With this model in mind, the landscape becomes easier to navigate. Companies can be understood in terms of the **positions** they take up rather than the labels they use. Some sit close to the financial machinery; others build tools on top of it. Some shape the checkout experience; others handle risk, routing, or settlement. The exact boundaries are less important than the orientation. When a provider expands to another region, introduces a new payment method, or launches risk and analytics tools, it is often shifting its position within the system rather than redefining itself. This narrative view is more forgiving than strict classifications. It offers enough structure to orient yourself without requiring encyclopaedic precision. You begin to see a pattern: the pieces of the payments ecosystem are aligning, slowly but noticeably, as digital commerce pulls formerly separate functions closer together. ## Why the Mollie–GoCardless story fits the broader movement Seen through this lens, the acquisition points to something larger than a corporate deal. Online payments were long built around push transactions. Recurring revenue, however, relies on pull transactions. For a long time these worlds remained separate, even as many businesses now blend both modes of interaction. Bringing them together within a single platform reduces friction for the companies using it. But more importantly, it reflects a maturing of the European payments landscape. The system is not becoming simpler, yet the way the pieces relate to each other is becoming easier to recognise. That, to me, is the significance of the Mollie–GoCardless story. It is not about scale or product consolidation but about coherence. A reminder that understanding digital payments is less about memorising terms than about seeing where things belong, how they evolve, and why some combinations begin to make sense only when the underlying structure comes into view. --- If [*How Money Moves*](https://rob-hoeijmakers.ghost.io/how-money-moves/) offered a map, this piece adds a few landmarks. Not a full guide, but enough to help you move through the terrain with a clearer sense of what sits where, and why. ### Apple Pay, Apple Wallet, Apple Card: How Apple Reshaped How Money Flows URL: https://rob-hoeijmakers.ghost.io/apple-pay-apple-wallet-apple-card/ Last updated: 2026-05-10T08:53:43.000Z I was working on an article called [*How Money Moves*](https://rob-hoeijmakers.ghost.io/how-money-moves/). Collecting scenes, interfaces, moments where money quietly moves from one place to another. Somewhere between screenshots and diagrams, something small but unsettling happened. I realised that I had never really separated Apple Pay from Apple Wallet. In my head, they had always been the same thing. I pay with my phone. A card appears. Money moves. End of story. That confusion is not careless. It is designed. And once you notice it, an entire layer of modern payment infrastructure becomes visible. ## Apple Wallet and Apple Pay are not the same thing Apple Wallet is a **container**. It holds credentials: payment cards, tickets, boarding passes, transit cards, keys, identity documents. Apple Pay is a **payment service**. It is the system that turns a stored payment credential into an authorised transaction. As a user, you rarely experience them separately: - You double-click. - Wallet appears. - Face ID confirms. - The terminal beeps. Visually, you are in Wallet. Functionally and legally, you are using Apple Pay. Wallet answers *what you have*. Apple Pay answers *how you pay*. The elegance of the system is that you are not asked to care. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/12/Apple-Pay-Apple-Wallet.jpg) ## The first layer: paying in a shop The most familiar case is the physical point of sale. Here Apple Pay behaves like an unusually capable contactless card: - NFC handles proximity - Face ID or Touch ID confirms intent - a one-time token is generated - the terminal never sees your real card number From the outside, this looks like a card replacement. Underneath, it is a carefully choreographed interaction between device hardware, operating system, payment service, and card networks. Visa or Mastercard rails still carry the transaction. Banks still settle the money. Apple controls the interface and the sequence. ## The second layer: in-app payments The picture changes when there is no terminal. Inside apps, Apple Pay becomes more than contactless. It becomes a **secure authorisation layer**: - no NFC - no form filling - no card details passed to the app The device identifies the user. Apple Pay confirms intent. The app receives confirmation, not credentials. Here Apple Pay starts to resemble identity infrastructure as much as a payment method. The same gesture now works without physical presence. ## The third layer: paying on the web Online, Apple Pay quietly competes with iDEAL, PayPal, and card entry forms. In Safari: - a button appears - Face ID confirms - the transaction completes The similarity with iDEAL is striking from the user’s point of view. One action, strong trust, fast completion. The architecture, however, differs completely. iDEAL starts at the bank and pushes money. Apple Pay starts at the device and pulls via card rails. Same outcome. Different centre of gravity. ## A geographical extension: Apple Card In the United States, Apple goes one step further. Apple Card is a credit card designed by Apple, issued by Goldman Sachs, and running on Mastercard rails. It lives almost entirely inside Apple Wallet. Apple does not become a bank here. It becomes the **primary interface to the bank**. - transactions are visualised in Wallet - rewards are calculated there - support starts there - the physical card is secondary Apple Card only exists in the US. That constraint is revealing. Consumer credit law, risk models, and regulation differ sharply across regions. Apple advances where the stack allows it, and stops where it does not. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/12/Apple-Card.jpg) ## What this reveals about how money flows None of this changes how money ultimately moves. Banks still hold accounts. Card networks still run rails. Settlement still follows familiar paths. What changes is **where trust is anchored**. Not in the plastic card. Not in the checkout. But in the user–device pair. Apple Pay works in shops, apps, and browsers because it is not tied to a place. It is tied to identity, hardware, and control over the payment moment. The reason Apple Pay and Apple Wallet blur together is not marketing. It is strategy. The container is what you see. The service is what acts. The infrastructure stays politely out of frame. Until you stop and look. --- ### Further reading - [How Money Moves](https://rob-hoeijmakers.ghost.io/how-money-moves/) ### How Money Moves URL: https://rob-hoeijmakers.ghost.io/how-money-moves/ Last updated: 2026-05-10T08:53:43.000Z *A reconnaissance of wallets, cards, networks, and digital payment rails* I did not set out to write about money. I noticed a feature. Klarna appeared in the news with something new: tap-to-pay in physical shops, including on iPhone. That small announcement was enough to unsettle a comfortable assumption. I use my phone to pay. I thought I knew how that worked. I did not. This piece is an attempt to look around rather than draw conclusions. A reconnaissance of how money now moves through digital systems, and why a change that looks minor at the surface may signal a deeper transition underneath. ## The disappearing card For most people, the payment card has already vanished as an object. We no longer reach for plastic. We tap a phone or a watch. Yet the system beneath remains card-based. What changed is the interface. A physical card is replaced by a token stored on a device, authenticated by biometrics, transmitted via NFC. The card fades from view, but the card network persists. This distinction matters because when payments appear to change, they often only shift shape rather than substance. [Apple Pay, Apple Wallet, Apple Card: How Apple Reshaped How Money FlowsWhile writing about how money flows, a simple confusion between Apple Pay and Apple Wallet exposes the hidden layers of modern payments and power.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-429.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_1383.jpeg)](https://rob-hoeijmakers.ghost.io/apple-pay-apple-wallet-apple-card/) ## Wallets as interfaces, not money A digital wallet is not money and not a bank. It is an interface. [Apple Pay](https://rob-hoeijmakers.ghost.io/apple-pay-apple-wallet-apple-card/), Google Pay, Klarna, and various bank wallets are containers for credentials. They determine how payment is initiated, how it feels, and which rails are used underneath. Those rails may be global card networks, bank-to-bank transfers, or lines of credit. Crucially, some wallets are not just software interfaces but **hardware-level integrations**. Apple Pay is tied to a physical gesture, biometric sensors, and lock-screen access. That coupling collapses payment into a reflex. Other wallets, even when technically capable, must operate at the application layer. That difference alone shapes habit, adoption, and power in the payment landscape. Seen this way, [Klarna’s move into physical shops ](https://www.klarna.com/international/press/klarna-launches-tap-to-pay-for-in-store-purchases-across-14-markets/)becomes clearer. It did not invent a new payment method. It extended its interface into a space that had long been dominated by device-level wallets. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/12/Tap_to_Pay_3x2.jpeg) ## The networks beneath the surface Although users rarely notice them, global card networks sit at the centre of most retail payments. Visa and Mastercard handle routing, acceptance rules, tokenisation, fraud management, and dispute resolution at planetary scale. Banks issue cards and manage customer relationships, but they depend on Visa and Mastercard for interoperability and reach. Without these networks, most everyday payments would stop working outside narrow domestic contexts. Their influence is powerful precisely because it is not experienced directly. ## Europe’s counter-current Initiatives such as [Wero](https://rob-hoeijmakers.ghost.io/wero-europes-new-payment-brand/) arise from a different ambition. They do not primarily compete by refining interfaces. They attempt to rebuild the rails. Wero is designed around bank-to-bank transfers using SEPA Instant. Its aim is to reduce reliance on card networks such as Visa and Mastercard altogether. The trade-off is explicit. Infrastructure control is prioritised over frictionless experience. This difference shows in use. Account-to-account payments tend to require visible action. The user initiates, confirms, and authorises. Paying does not disappear into the device. [Wero: Europe’s new payment brandWero is Europe’s bid for payment sovereignty: pay across borders with just an email or phone number, without relying on US networks.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-423.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_7472-5.jpeg)](https://rob-hoeijmakers.ghost.io/wero-europes-new-payment-brand/) ## Regulation as a point of revelation The [Digital Markets Act](https://rob-hoeijmakers.ghost.io/eu-digital-markets-act/) forced Apple to open NFC access within the European Union. No new technology was introduced. An exclusive constraint was removed. Once that constraint disappeared, wallets that already existed could surface at the point of sale. Klarna’s tap-to-pay feature is one of the first visible outcomes. Others will follow. Here, regulation does not dictate outcomes. It reveals structure. It brings into view the layers that had shaped behaviour quietly for years. ## A transition, not a replacement What is unfolding is not a clean handover from one system to another. Cards persist beneath wallets. Wallets multiply above Visa and Mastercard. Banks, platforms, and regulators apply pressure from different directions. Power shifts slowly and sideways. Some models optimise speed and invisibility. Others prioritise control and resilience. None fully replaces the rest. ## Why this is worth mapping For those who are digitally fluent, payments feel solved. For others, they feel opaque. In both cases, the underlying structure remains largely unseen. A simple tap connects hardware, wallet software, banks, global networks, and public policy. Mapping how money moves is not about prediction or prescription. It is about noticing how everyday gestures travel through layered systems shaped by technology, corporate incentives, and regulation. This is reconnaissance. The surface keeps shifting. The rails underneath are where the real negotiations continue. --- ### Further reading - [The Digital Markets Act (DMA) is Here and It’s Changing How Apps Work](https://rob-hoeijmakers.ghost.io/eu-digital-markets-act/) - [The Silent Upgrade: What Really Changed When Our Bank Replaced Our Cards](https://rob-hoeijmakers.ghost.io/the-silent-upgrade-mastercard-debit-transition/) - [Understanding Digital Identity in the EU: eID, eIDAS, and the EUDI Wallet](https://rob-hoeijmakers.ghost.io/eu-eid-eidas-eudi-wallet/) - [Wero: Europe’s new payment brand](https://rob-hoeijmakers.ghost.io/wero-europes-new-payment-brand/) - [Direct Debit: US versus EU](https://rob-hoeijmakers.ghost.io/direct-debit-us-europe/) ### Europe Fines X: The Moment the DSA Became Real URL: https://rob-hoeijmakers.ghost.io/dsa-fines-x/ Last updated: 2026-05-10T08:53:43.000Z ## Quick takeaways - The EU has imposed its first fine under the Digital Services Act: €120 million against X. - This marks a shift from expectation to enforcement. - The DSA is not competition law, but ongoing governance of platform design and power. - Earlier platform changes, such as Meta stepping back from political content, were anticipatory, not imposed. - For Europeans, this now carries financial, political, and media consequences. ## Watching a law take shape When I first wrote about the [Digital Services Act](https://rob-hoeijmakers.ghost.io/europes-dsa-and-the-new-contours-of-online-freedom/) last summer, I did so consciously as a learner. I am still one. No exams passed, no formal endpoint reached. What I had was curiosity, and a decision to follow this law closely as it moved from text to practice. At the time, the DSA felt like a chessboard. The pieces were laid out, the rules visible, but the game had barely started. Platforms adjusted cautiously. Regulators prepared. The effects remained indirect. A few months later, the first serious move has been made. With the European Commission fining X €120 million under the Digital Services Act, the law has crossed a clear threshold. From design into action. [Digital Europe in Two Acts: DSA and DMAEurope’s DSA and DMA work in tandem. One governs digital content, the other digital competition. Together they form the EU’s Digital Services Package.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-418.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/D4A6BF61-3D84-472E-90FC-DE844F696546-1.jpeg)](https://rob-hoeijmakers.ghost.io/dsa-dma-compared/) ## Why this fine is different It is tempting to read this as just another EU tech fine. That would miss what is new here. This is not competition law. It is not about market dominance or pricing power. The DSA governs something more structural: how very large platforms are designed, governed, and held accountable over time. The violations identified in the X case are revealing. They concern misleading verification signals, limited advertising transparency, and restricted access for independent researchers. These are not marginal issues. They shape trust, visibility, and oversight at a systemic level. This is governance, not punishment. And crucially, it is ongoing. Under the DSA, a fine is not the end of a case, but the point at which expectations become enforceable. [Commission fines X €120 million under the Digital Services ActToday, the Commission has issued a fine of €120 million to X for breaching its transparency obligations under the Digital Services Act (DSA).![](https://static.ghost.org/v5.0.0/images/link-icon.svg)European Commission - European Commission![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/sm_ec_logo_big.jpg)](https://ec.europa.eu/commission/presscorner/detail/en/ip%5F25%5F2934) ## Anticipation versus enforcement Earlier, I used Meta’s decision to reduce the visibility of political content as an example of regulatory impact. The comparison still holds, but with an important clarification. Meta was not fined. There was no enforcement action. What we saw was anticipatory behaviour: a platform adjusting strategy to reduce risk under a new regulatory environment. The X case is different. Here, the regulator completed a procedure, established violations, and imposed consequences. Together, these moments show how the DSA progresses. First, platforms adapt in anticipation. When they do not, enforcement follows. The law does not rush, but it does arrive. ## Why this is no longer abstract European law often feels distant. Procedural. Written in a language that rarely invites curiosity. I shared that instinct myself. What changed for me while following the DSA is that it is no longer theoretical. With its first enforcement action, it has entered a tangible world. Design choices now have financial consequences. Governance failures carry political weight. And the interfaces we use every day operate under a different set of expectations. This is not legal detail for its own sake. It touches media ecosystems, product decisions, trust signals, and the tone of public discourse online. Once you see that, the law becomes legible. ## A small personal stake I should also be transparent about why this matters to me beyond professional interest. I have been active on these platforms for years. Not at massive scale, but enough to notice shifts. A few thousand followers. Enough to sense when a space invites exchange, and when it rewards antagonism. I enjoyed [Twitter](https://x.com/robhoeij) for a long time. I even had the blue checkmark at one point, which felt oddly meaningful until it didn’t, and I let it go again. Over time, my relationship with the platform changed. Not because of “Big Tech” as an abstract enemy, but because the atmosphere hardened. More violence. More ideological signalling by ownership. Less sense of an open public space. I did not abandon commercial platforms. Newspapers and broadcasters are businesses too. Products can make money without being corrosive. I simply moved some of my attention elsewhere, including to [Meta’s Threads](https://www.threads.com/@robhoeij), which has its own dynamics but currently feels governed differently. That experience sharpened my attention to governance. Because platform power does not announce itself only through policy, but through product design. And that is what the DSA ultimately addresses. ## The broader picture There is also a geopolitical dimension to this moment. A European regulator enforcing structural obligations on a US-owned platform is not an act of hostility, but of jurisdiction. The message is simple: access to the European digital public space comes with conditions. This is not about banning platforms, nor about moral posturing. It is about making power commensurate with responsibility, and doing so through law rather than improvisation. You do not need to agree with every aspect of this approach to recognise its significance. ## Coda: learning in public I am still learning this law. But there is a difference between not knowing and not seeing. Over the past months, I’ve gone from reading legal texts to recognising moves as they happen. From abstract concerns to concrete consequences. That has given me a particular kind of confidence. Not in having the right opinion, but in understanding where things come from, and why certain moves are possible at all. If EU law feels complex or distant, that is understandable. But moments like this matter precisely because they make the system visible. The Digital Services Act is no longer potential. It is operative. And that makes it worth watching, move by move, because the game is now being played in the open. --- ### Further reading - [Europe’s DSA and the New Contours of Online Freedom](https://rob-hoeijmakers.ghost.io/europes-dsa-and-the-new-contours-of-online-freedom/) - [Playing the EU Game: Learning the Digital Rules by Getting in the Game](https://rob-hoeijmakers.ghost.io/playing-the-eu-game/) - [European Commission hits Elon Musk’s X with €120 million fine](https://www.euronews.com/my-europe/2025/12/05/european-commission-hits-elon-musks-social-network-x-with-120-million-fine) ### Structuring Knowledge with Obsidian URL: https://rob-hoeijmakers.ghost.io/structuring-knowledge-with-obsidian/ Last updated: 2026-05-10T15:55:45.000Z Obsidian is a simple text notebook that lets you connect individual notes and then see those connections as a map. That small shift changed how I understood the knowledge I had gathered over time, and it also changed how I could use it. Much of my work involves building conversational and agentic AI systems, and these systems depend on well-structured sources. Once my own material became a network rather than a sequence, it became far more suitable for that purpose. If you work with notes, research or any growing body of text, this approach may help you see relationships that usually remain hidden and prepare your material for more advanced uses. This piece focuses on the core idea: how non-linear knowledge can be structured, visualised and reused for symbolic or hybrid AI. 💡 I use Ghost for this blog. It is a lightweight publishing platform with a deliberately flat tag system. The tags help readers find their way, but they offer no hierarchy or semantics, so they cannot function as a real taxonomy. Seen in that light, they become simple way-markers rather than structural tools. The actual conceptual structure lives elsewhere, in Obsidian, where relationships and meaning can take shape. ## Knowledge Is Rarely Linear We often store information as if it were linear: in folders, lists, or chronological archives. But ideas rarely develop that way. They branch, overlap and return. Understanding a domain requires seeing those relationships rather than the order in which material happened to be written. The first step is to treat text not as a sequence but as a field of connections. ## A Text-Based Foundation Working entirely with plain text ([markdown](https://rob-hoeijmakers.ghost.io/using-markdown-for-text/) to be precise) keeps the material flexible. It can be reshaped, linked and repurposed without being locked into a particular system. When moved into Obsidian, each piece of text becomes a node that can be connected to others. The structure emerges from the links rather than from an imposed hierarchy. This simplicity is essential for later use in symbolic or hybrid AI systems. ## Visualising Structure Obsidian’s graph view does not create meaning, but it reveals how meaning is distributed. Dense clusters show where ideas reinforce each other. Sparse areas show where thinking is thin or disconnected. Connections that were intuitive become visible. Gaps that were unnoticed become clear. Visualisation here serves a practical purpose: it prepares the ground for knowledge retrieval. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/12/image.png) ## From Structure to Reuse in AI Symbolic and hybrid AI systems operate best when they have access to [structured knowledge](https://rob-hoeijmakers.ghost.io/tag/knowledge-engineering/). A linked set of texts, even at the page level, is already more usable than a flat archive. An AI agent can navigate the map, follow connections and retrieve material with more precision than free text alone allows. This is the real benefit: once knowledge is structured and visualised, it becomes reusable. ## A Case Study: My Own Archive I discovered this by using my blog (in Ghost) as a test case. Hundreds of posts spread across years became a network once they were in Obsidian. Themes surfaced. Outliers appeared. The structure became workable rather than opaque. But the blog is only an example. The method applies equally to research notes, project documentation or any long-running collection of text. ## Looking Ahead Looking ahead, the next step is finer granularity: moving from pages to concepts, where a true knowledge graph begins to form and where symbolic AI can work with greater precision. But even at the current level, Obsidian has proven to be a remarkably effective instrument. It keeps knowledge close to plain text, makes structure visible and leaves room for growth. For many kinds of work, it may be one of the most practical ways to shape and reuse knowledge, both for human thinking and for AI systems. [Obsidian - Sharpen your thinkingThe free and flexible app for your private thoughts.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon-13.svg)Obsidian![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/banner.png)](https://obsidian.md) ### Further reading - [Making Content Ready for Intelligence](https://rob-hoeijmakers.ghost.io/platecms-content-ready-for-intelligence/) - [Rules and Guesses: The Twin Engines Behind Smarter AI](https://rob-hoeijmakers.ghost.io/rules-and-guesses/) ### Qivalis, a New Euro Stablecoin URL: https://rob-hoeijmakers.ghost.io/qivalis-a-new-euro-stablecoin/ Last updated: 2026-05-10T08:53:44.000Z News of a new euro stablecoin caught my eye this week. At first it felt familiar. I had used a euro-pegged token before, so this did not look like a breakthrough. Stablecoins have been around for years. Then I saw who stood behind it: a new company called **Qivalis**, jointly owned by some of Europe’s largest banks. That small detail changes the entire meaning. It turns a niche instrument into part of Europe’s financial architecture. [#stablecoins #qivalis | Qivalis | 35 commentsExcited to share that Qivalis is here! Built by a consortium of European banks - BNP Paribas, CaixaBank, Danske Bank, DekaBank, ING, KBC Bank & Verzekering, Raiffeisen Bank International AG , SEB, Banca Sella, UniCredit -, Qivalis represents a pivotal moment for European digital commerce. Designed to be backed 1:1 by the euro and, to operate 24/7, Qivalis will empower businesses and consumers with seamless, trusted digital payments. Jan-Oliver Sell, CEO of Qivalis: “European businesses and consumers deserve a digital economy built for their success, and we will enable new opportunities for European companies and consumers to benefit from on-chain payments and digital asset markets in their own currency.” The Amsterdam-domiciled company is working towards Dutch Central Bank (De Nederlandsche Bank) authorization and supervision as an Electronic Money Institution (EMI). Its objective: to launch the first on-chain, fully regulated euro-denominated stablecoin in H2 2026.This is what European strategic autonomy in the digital age looks like. #stablecoins #qivalis | 35 comments on LinkedIn![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/al2o9zrvru7aqj8e1x2rzsrca-7)LinkedInQivalis![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/1764666657929)](https://www.linkedin.com/posts/qivalis%5Fstablecoins-qivalis-activity-7401568316172955649-jKBE?utm%5Fsource=share&utm%5Fmedium=member%5Fdesktop&rcm=ACoAAAAGPccBqY4Rvlcj2zZ2PdvOk7tl5dQrkUs) ## Quick takeaways - Euro stablecoins existed before, but none had institutional weight. - Qivalis is a company owned by major European banks and is built as infrastructure. - Stablecoins use different financial rails from traditional payments. - The euro lacked strong digital rails and often ceded ground to the dollar. - The project fits into a wider European push for financial sovereignty. ## What a stablecoin is Stablecoins are simple at their core. They are digital tokens that follow the value of a currency and are backed one-to-one by reserves. In this case, each token represents one euro, not in a bank account but in a programmable digital form. They are not new. Fintechs have issued euro tokens for years. You could already use them in certain wallets or on some exchanges. But they never became part of Europe’s mainstream financial systems. ## What “rails” actually mean Rails are the **infrastructure** along which payments travel. - When you make a bank transfer, the payment runs on **SEPA**. - When you use a card, the transaction runs on **Visa** or **Mastercard**. - When you tap your phone, it runs on **Apple Pay** as an interface on top of bank rails. These are all rails: the underlying systems that carry money from one place to another. Stablecoins run on a different type of rail. Instead of moving through banks or card networks, they settle value directly across distributed, programmable systems. There is no need for correspondent banks or multi-day clearance. The rails are written in software. This difference is why stablecoins matter more than the tokens themselves. ## Why earlier euro stablecoins stayed marginal Earlier euro-stablecoins were useful for developers and specific trading platforms, but they remained peripheral. They lacked: - large-scale liquidity - recognition by banks - integration with payment systems - institutional trust - regulatory backing under the new MiCA framework They existed, but they did not shape the financial plumbing. As a result, most digital settlement defaulted to the dollar. Not because the dollar is inherently better, but because the **dollar had digital rails and the euro largely did not**. 💶 [MiCA](https://www.esma.europa.eu/esmas-activities/digital-finance-and-innovation/markets-crypto-assets-regulation-mica) is the EU’s new rulebook for crypto and digital assets. It introduces clear categories, licensing, and conduct requirements, with [ESMA](https://www.esma.europa.eu) providing the technical standards and cross-border supervision. The aim is to replace today’s fragmented landscape with a predictable, union-wide framework that allows fintech firms to scale while keeping consumer protection and market integrity at the centre. ## Enter Qivalis Qivalis is different because of who created it. It is a new company, jointly owned by major European banks such as ING, UniCredit, BNP Paribas, KBC and others. This changes the purpose entirely. Qivalis is designed as: - **infrastructure**, not a product - part of the European settlement layer - fully regulated under MiCA as an e-money token issuer - interoperable with banking and payment systems - suitable for institutional and cross-border use This gives the euro a credible presence inside the programmable layer of finance. It is not a niche tool anymore; it becomes part of the mainstream rails. [Qivalis - Secure. Trusted. Future-ready.A euro stablecoin intended to be fully regulated, powered by Europe’s leading banks, designed to make digital finance secure, simple, and ready for tomorrow![](https://static.ghost.org/v5.0.0/images/link-icon.svg)Secure. Trusted. Future-ready.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/caixa-bank.png)](https://qivalis.eu) ## The sovereignty angle Sovereignty here is not a grand statement. It is practical. Europe has seen how digital payments and tokenised markets drift toward American infrastructure. US dollar stablecoins dominate because they are easy to use and deeply integrated. European firms using them implicitly rely on US technology, US liquidity pools and US regulatory regimes. A bank-issued euro stablecoin changes the balance. It allows European institutions to transact in euros without passing through foreign rails. It aligns with EU rules on privacy, identity and compliance. It strengthens the euro by strengthening its digital foundations. This is sovereignty expressed in infrastructure rather than rhetoric. ## Closing I find it helpful to view developments like this not as financial novelties but as adjustments to the foundation. Money is only part of the picture. The rails beneath it determine who sets the rules, who has leverage, and how value moves in a digital economy. With [Qivalis](https://qivalis.eu), Europe adds an essential piece that was missing: a euro that can travel at the speed of software, on its own terms. [Qivalis to launch euro stablecoin by 2026 backed by major European banksQivalis, backed by top European banks, plans a euro stablecoin launch by 2026 to challenge US-led payments and boost EU digital sovereignty.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/ios-144.png)Crypto BriefingEstefano Gomez![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/7bd8dfa5-0a3b-4705-aacc-064774b5dee1-800x420.jpg)](https://cryptobriefing.com/qivalis-euro-stablecoin-european-banks-2026/) ### Further reading - [Wero: Europe’s new payment brand](https://rob-hoeijmakers.ghost.io/wero-europes-new-payment-brand/) - [The Digital Euro: My First Exploration](https://rob-hoeijmakers.ghost.io/digital-euro-cbdc/) ### When the Future Starts Knocking Quietly URL: https://rob-hoeijmakers.ghost.io/when-the-future-starts-knocking-quietly/ Last updated: 2026-05-10T08:53:44.000Z I recently had a long and stimulating conversation with someone I had not met before. His name is [James Myers, publisher of The Quantum Record](https://thequantumrecord.com/about/) and a Canadian accountant with a deep interest in how our systems of value and knowledge have developed. What began as an exchange on accounting and trust turned into a wake-up call about something I had been treating as distant: [quantum computing](https://rob-hoeijmakers.ghost.io/tag/quantum-computing/). ## From Numbers to Trust Our first topic concerned how societies record value. Accounting seems technical on the surface, but beneath every entry lie [choices about what matters](https://rob-hoeijmakers.ghost.io/the-new-reckoning/), what lasts and what deserves protection. We discussed trust, the vulnerability of democracy and the way financial systems can drift away from the tangible world they are meant to represent. [What’s the Value of Time in the Digital Era? In the Long Run, Slowing Down and Being Bored Can Produce a Wealth of Benefits - The Quantum RecordAI promises to create value by speeding work, but maybe far greater future value can be generated by taking a break for creativity.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/cropped-577326-200-1-270x270-1.png)The Quantum RecordJames Myers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/Value-of-time.jpg)](https://thequantumrecord.com/blog/value-of-time-being-bored-can-produce-wealth-of-benefits/) ## A Shift in Perspective Late in the conversation, James introduced his current focus. He writes about quantum computing. I had assumed this was a distant horizon, still confined to laboratories. He challenged that view. The landscape is changing sooner than many realise. ## Identity and the Quantum Threat One concrete example changed my sense of timing. Germany has begun issuing identity documents designed to be resistant to attacks from quantum computers. Cyber security experts have been preparing for this shift for years, yet most people in civic life have never heard of it. The foundations of digital trust can move long before the public sees them move. [Post-Quantum Cryptography: Germany Prepares Next-Generation ID Cards for the Quantum Eraermany takes a global lead in digital security: Bundesdruckerei, Giesecke+Devrient, BSI, and Infineon develop the world’s first post-quantum–secure ID card technology—protecting citizens’ digital identities against future quantum-computer attacks.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/cropped-PRCOM_favicon-270x270.png)PR-COM![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/prcom_newlogo-1.svg)](https://pr-com.de/company%5Fnews/gd-germany-sets-new-standards-for-secure-id-documents-in-the-quantum-computing-era/) ## Why This Belongs in Public Conversation This brought us back to our starting point. Value is not only what appears in books. It includes the continuity of trust across generations, and the systems that hold identity, privacy and public life together. Cryptography is part of that foundation. If quantum capability arrives in practice before it arrives in public attention, we will adapt only after the fact. ## A Beginning This article marks the start of a new enquiry for me. I want to understand what quantum computing means for societies that depend on trust. I plan to look at what is becoming possible, who is preparing and where we still have choices. I am grateful that James encouraged me to take this seriously. We agreed to continue working together. With his support I will write foundational pieces that help me get a firm grip on the technology, while also beginning to reflect on the implications for the systems that keep our societies stable and free. [Home - The Quantum RecordA journal of science, technology, philosophy, and time featuring the ideas of the good people who add to knowledge for the benefit of all![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/cropped-577326-200-1-270x270.png)The Quantum Record![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/nf_car_t_patient-430x280.jpg)](https://thequantumrecord.com) ### Further reading - [Quantum computing - Rob Hoeijmakers](https://rob-hoeijmakers.ghost.io/tag/quantum-computing/) - [Review of The Reckoning, in the light of the AI boom](https://rob-hoeijmakers.ghost.io/the-new-reckoning/) ### Making Content Ready for Intelligence URL: https://rob-hoeijmakers.ghost.io/platecms-content-ready-for-intelligence/ Last updated: 2026-05-26T19:34:20.000Z A short exchange on LinkedIn about [llms.txt](https://rob-hoeijmakers.ghost.io/llms-txt/) led me to meet [Pieter Versloot from Plate](https://www.linkedin.com/in/pieterversloot/). I expected a technical discussion about AI crawling behaviour. Instead, we ended up talking about the foundations of organisational memory. How symbolic structure can emerge from the unstructured content organisations already produce every day. That is the quiet breakthrough. Content doesn’t need to be forced into formal structure when it is created. The structure can be discovered. And once it is discovered, it can be maintained. Meanwhile, publishing still serves humans. We write pages. We measure reach. That hasn’t changed. What has changed is who reads first. Language models digest our content before most people do. They internalise what we publish, form assumptions, and reuse that knowledge elsewhere. Content behaves like knowledge. We still treat it like decoration. [When Bots Become Readers: Publishing in the Age of AI CrawlersListening to Matthew Prince on Azim Azhar’s podcast made me reflect on who actually reads my blog. People (like you), machines, or both.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-395.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_9168-3.jpeg)](https://rob-hoeijmakers.ghost.io/when-bots-become-readers/) ## Fragmentation as the default As organisations scale, content spreads and slowly drifts apart. Small copies accumulate. Facts diverge. Metadata becomes optional. Editors continue. Machines continue. Trust becomes fragile. This is not a CMS problem. It is a knowledge integrity problem. ## A conversation that connected the dots Pieter approached this from the inside: a decade of building websites and CMS for large organisations. He saw publishing succeed at the surface while coherence failed underneath. Plate’s response is not to replace systems. It is to reveal, repair and sustain what organisations already know, but cannot easily see. ## The layer that sees At the centre of Plate Delta is CAS, the **Content Analysis Service**. It reads content across systems, notices patterns and detects gaps. Repeated product names. Inconsistent facts. Duplicated assets. Tone shift across teams. Metadata that is present in one place and missing in another. Machines do the auditing humans cannot scale. > “With Plate Delta, structure doesn’t come first, it follows the work. CAS reads how content is actually used and guides from real patterns, not preset rules. Structure stops feeling forced and starts doing its job.” > — Pieter Versloot Structure emerges from reality rather than being mandated. ## The layer that repairs and maintains The **Content Engineering Layer (CEL)** then reinforces what CAS discovers. It creates a continuous loop: • detect divergence • align to a shared model • prevent new drift Editors remain expressive. Systems remain flexible. Coherence becomes infrastructural rather than personal. > “The bigger an organization gets, the more content rules drift into people’s heads or scattered tools. That doesn’t scale. You end up with systems that lock everyone down or with shadow IT that fuels chaos. Only a central, intelligent content base breaks that cycle.” > — Pieter Versloot CEL does not centralise content. It centralises alignment. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/12/Plate---CEL-1.jpg) CEL acts as a smart layer on top of your existing systems, restoring control and clarity to your content management. ## Where symbolic structure takes shape Unstructured writing is how humans think. From it, CAS and CEL surface: • entities • attributes • relationships • contexts and constraints What begins as text becomes symbolic knowledge. A knowledge graph forms organically out of use. Meaning becomes computable. AI can draw from a reliable foundation. [Rules and Guesses: The Twin Engines Behind Smarter AIAI isn’t just smart, it’s a duet. One part rigid and logical, the other fluid and generative. Together, they form the neuro-symbolic systems behind real insight.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-394.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_4826-3.jpeg)](https://rob-hoeijmakers.ghost.io/rules-and-guesses/) ## What this enables A continuously maintained knowledge base improves daily work: • faster editing, less hunting for inconsistencies • reliable product information everywhere • multilingual content aligned over time • AI assistants grounded in verified truth Structure becomes governance. Consistency becomes trust. ## Plate’s product set Three components, one intelligence loop: 1. **Content Health Audit** See what is broken 2. **CEL**: Content Engineering Layer Repair and maintain alignment 3. **Plate CMS (Delta)** Publish without reintroducing chaos Machines reveal. Machines align. Humans stay in control. ## Looking ahead Content becomes knowledge when it stays aligned over time. And that memory becomes the internal logic on which every service relies. If you recognise that your content is working against you as much as for you, there might be a better foundation available. I’m not affiliated. Just impressed. And curious where this thinking leads. [Multi Site CMS & Digital Experience Platform | PlateAt Plate, we’ve been pushing the boundaries of content management for 10 years. Our mission is to give content and website managers full autonomy over their websites and content. Over 1,400 customers rely on Plate’s CMS products to navigate the fast-paced online world. Plate Multisite is the all-in-one solution for managing multiple websites, while Plate Delta is the headless CMS designed to structure and manage content seamlessly across various channels, all with minimal development needs and full content reusability.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Plate-20Multisite-20Wordpress-20plugin-20128-1.png)Logo\_leliezorggroep (1)![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/ByManonvB_PlateBBQ-34-1.jpg)](https://platecms.com/) ### Further reading - [What llms.txt can do for your website](https://rob-hoeijmakers.ghost.io/llms-txt/) - [The right CMS for your online content](https://rob-hoeijmakers.ghost.io/content-management-system-cms/) ### Voice Mode Without the Screen URL: https://rob-hoeijmakers.ghost.io/voice-without-the-screen/ Last updated: 2026-05-10T08:53:44.000Z ## When the phone locks but the conversation with Gemini / ChatGPT continues Voice mode is a way of using the phone when the screen gets in the way. Walking outside. Carrying bags. Looking around instead of down. In those moments, a conversational assistant can help you think aloud, check something quickly, or have the phone read what is on the display while your hands and eyes are elsewhere. Modern iPhones add a new capability to that scenario. The lock screen is no longer just a barrier. It has become a minimal interface. It can keep part of a session visible and under control even when the device locks itself. A stop button. A subtle indicator that the microphone is active. Presence without demanding attention. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/11/Live-with-Gemini---Lock-Screen.jpg) Lock Screen with Live Activity ## A real-world test Recent improvements in the voice modes of ChatGPT and Google’s Gemini made me curious to try them again in motion. So I took a long walk and compared both, with the phone slipping into the background. Both were fluent and responsive. But something unexpected happened. ChatGPT’s Live Activity never appeared. The screen locked, the conversation went on, and all signs of it disappeared. To pause or stop, I needed to unlock the phone and find my way back into the right view. It disrupted the flow and chipped away at the sense of ease. The failure was small. The impact was not. It showed how much trust depends on what happens after the screen goes dark. 🔒 The Lock Screen is the minimal interface that appears when the iPhone wakes. It keeps the device secure while still showing essential information and allowing a few quick actions, such as pausing media or ending a call, without unlocking. ## The invisible support structure Live Activities and the Dynamic Island are designed to prevent such confusion. They bridge the gap between app and operating system. A thin, ambient surface that keeps the session anchored. They tell you that the assistant is still there. They give you a way to take control back instantly. We only notice this layer when it disappears. There is no dashboard where you can see what is active. No way to manage it directly. We rely entirely on iOS to make the right call about visibility. That reliance is already shaping the user experience of voice. Not the intelligence. Not the personality. But the subtle infrastructure that keeps interaction grounded when attention moves elsewhere. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/11/iOS-Dynamic-Island---Live-Activity.jpg) Live Activity in and out the Dynamic Island, the black pill around the camera. ## The condition for confidence Voice mode is a useful modality alongside reading and [dictation](https://rob-hoeijmakers.ghost.io/asr-speech-to-text-dictation-and-transcription-whats-the-difference/). Not faster. Not always better. But perfect for the moments when screens are unwelcome. For that to feel natural, continuity must be visible and control must be immediate. A locked phone still listening must give clear reassurance. Otherwise, the interface feels like it has vanished while the microphone remains open. A dependable voice assistant requires more than a good voice. It needs the operating system to support the interaction quietly and consistently. ## Learning to see the hidden layer I am oddly grateful the feature failed. It forced me to dig into this system layer and learn the vocabulary for what I was experiencing. It revealed where hands-free interaction actually lives: not in the model itself, but in the orchestration around it. Voice without the screen can feel liberating. It can also feel unsettling. The difference lies in the smallest handles that keep us aware of what is happening when we stop looking. P.s. If anyone knows how to fix the Live Activities on ChatGPT's Voice Mode I [would be really happy](https://rob-hoeijmakers.ghost.io/about/). ### Further reading - [Exploring ChatGPT’s New Voice Features: More Hype than Substance?](https://rob-hoeijmakers.ghost.io/chatgpt-new-voice-features/) - [Speech recognition, Speech-to-Text, Dictation, and Transcription: What’s the Difference?](https://rob-hoeijmakers.ghost.io/asr-speech-to-text-dictation-and-transcription-whats-the-difference/) - [From Typing to Talking](https://rob-hoeijmakers.ghost.io/from-typing-to-talking/) ### Cybernetics: The Idea Behind Cyber URL: https://rob-hoeijmakers.ghost.io/cybernetics/ Last updated: 2026-05-10T08:53:45.000Z When I first encountered the word [*cyber*](https://rob-hoeijmakers.ghost.io/web-online-digital-and-cyber/) in the early 1990s, it meant something entirely different from what it means today. The web was just beginning to take shape, and *cyber* felt like an invitation: *cyberculture, cyberart, cyberspace.* It carried imagination. It suggested exploration. Over the years, that openness hardened. “Cyber” became the vocabulary of defence: cybersecurity, cybercrime, cyberwarfare. A word that once felt expansive narrowed into a border. Recently, reading Dan Wang’s [*Breakneck*](https://rob-hoeijmakers.ghost.io/dan-wang-breakneck/), I realised that to understand this shift, you need to trace the word back to its origins, long before it had anything to do with computers. ## The Original Meaning of “Cyber” The Greek word **κυβερνήτης (kybernetes)** means *steersman*, *helmsman*, one who guides a ship. From this root we get *governor* and *government*. The earliest meaning of *cyber* is **to steer**, to guide a system through uncertainty. In the nineteenth century, André-Marie Ampère revived the term as **la cybernétique** or the *science of governing*. In 1948, Norbert Wiener deliberately borrowed the word for a new science of feedback, control, and communication. Cybernetics described how systems steer themselves, whether animals, machines, organisations, or societies. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/11/1200px-Exekias_Dionysos_Staatliche_Antikensammlungen_2044_n2.jpg) Dionysus Cup ## A Book That Re-Opened the Question Reading *Breakneck* brought this tension into sharp relief. Wang describes how cybernetics entered China not as an abstract import but as a full intellectual project: taught in universities, applied in engineering, and used to frame problems of technology and society. One of the policymakers behind the one-child policy appears as a follower of this thinking, convinced that complex systems, *including populations*, could be steered through measurement and correction. Seen through that lens, the policy becomes a large-scale cybernetic loop: a system adjusting itself through quotas and feedback, guided by the belief that stability could be engineered. [Reading Breakneck by Dan WangFew books capture our accelerating era like Dan Wang’s Breakneck: global, personal, and clear-eyed about what drives modern change.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-384.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_1229.jpeg)](https://rob-hoeijmakers.ghost.io/dan-wang-breakneck/) ## A Parallel Insight: Varoufakis and the Corporate Cybernetic Plan Around the same time, I revisited[ Yanis Varoufakis’s work](https://rob-hoeijmakers.ghost.io/review-technofeudalism-by-yanis-varoufakis/). In one talk, he argues that **Amazon is not a capitalist market** but a **centrally planned, cybernetic command economy**. It looks like a market but functions as a feedback machine: - total information collection - demand prediction - supplier ranking - consumer steering through recommendation loops Placed alongside Wang’s observations, a symmetry appears. A Chinese ministry and an American tech giant both rely on **cybernetic governance**: prediction, measurement, correction, optimisation. ## How “Cyber” Became Digital The journey from cybernetics to the digital “cyber” we use today unfolded gradually, through a series of shifts in meaning. **1\. Cybernetics → Computing** Early AI and information theory borrowed cybernetic ideas. **2\. Cybernetics → Cyberspace** In the 1980s, William Gibson imagined a virtual world and named it *cyberspace*. **3\. Cyberspace → Cybersecurity** By the 1990s, governments used “cyber” for digital threats and defence. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/11/51FYSPKAN4L._UF1000-1000_QL80_.jpg) The German language edition of William Gibson's book. ## The Two Faces of Cybernetics Cybernetics contains two impulses: \- **Regenerative**, when feedback enables learning \- **Regulative**, when feedback enforces stability ## At the Helm of Our Own Thinking Feedback shapes many of the systems we use today, from platforms to public policy. It’s interesting to think how an idea from mid-century mathematics ended up guiding so much of daily digital life. Wiener saw feedback as a way to notice consequences, not to settle arguments. That still feels like a useful way to approach the world: understanding the patterns, and leaving space for new interpretations. Reading and writing about all this encourages me to keep being the helmsman in a rapidly changing intellectual landscape and especially now, when talking to AI can make it easy to give away a bit of our own agency without realising it. 💡 In a recent thread, [Sakeeb Rahman revisits a historical tension](https://www.threads.com/@sakeeb.rahman/post/DVZedJEESzn?xmt=AQF0cqTkNVY3PVoEYYrhCrfVhAi47zyIuFfN8oXEVRc73GTOplvOZLsT20ByAr6waCLxlhUx&slof=1): in 1956 John McCarthy coined “artificial intelligence” partly to distance the field from Norbert Wiener’s cybernetics. Yet today, as the MIT AI Agent Index traces agent definitions back to Wiener and Ross Ashby, cybernetics reappears as a missing design blueprint. The argument is that while AI advanced model capability, it neglected feedback, control architecture and second-order interaction. As agentic systems gain autonomy, cybernetic thinking about regulation, steering and co-evolution between human and system becomes newly relevant. ### Further reading - [Understanding the Terms: Web, Online, Digital, and Cyber](https://rob-hoeijmakers.ghost.io/web-online-digital-and-cyber/) - [Review: Technofeudalism by Yanis Varoufakis](https://rob-hoeijmakers.ghost.io/review-technofeudalism-by-yanis-varoufakis/) ### Group Chats in ChatGPT: Solving a Problem That Isn’t There URL: https://rob-hoeijmakers.ghost.io/chatgpt-group-chat/ Last updated: 2026-05-10T08:53:45.000Z New features often promise a glimpse of the future. Sometimes they simply reveal what we value in the present. ChatGPT now [offers group chats](https://openai.com/index/group-chats-in-chatgpt/). On paper it brings collaboration and intelligence into our everyday messaging. In practice it highlights something more basic: not every conversation wants or needs a machine in the room. ## Quick takeaways - Messaging is already a saturated and socially delicate domain - Adding an AI participant risks disrupting the human rhythm of interaction - The current implementation introduces friction rather than value - A healthier approach would separate models from the social interface - This is a personal take, based on direct experience ## The missing use case Group messaging apps have settled into clear roles. WhatsApp and iMessage serve intimacy and immediacy. Slack and Teams coordinate work. Discord hosts communities. These spaces feel human because they follow the grain of our conversation: mutual attention, timing, pauses, private humour. Into this arrives ChatGPT group chat. It is billed as a collaborative experience, but the design assumes a desire that does not feel widespread: moving personal or social conversations into an AI-first environment. When the assistant interjects uninvited, it breaks the tacit norms that keep these conversations comfortable. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/11/ChatGPT-Group-chat.jpg) From chat to chit chat. ## Friction at the foundation The issue is not only conceptual. The underlying mechanics struggle to support the new ambition. Today the feature reveals: - region restrictions without context - unclear boundaries between personal and business accounts - link-sharing that fails across user types These problems would be minor in a tool that solved a pressing need. Here they become reasons to step away entirely. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/11/Group-chat---ChatGPT---Not-in-this-region.jpg) Group chats not available in your region it says. ## Strategy without sociology There is a broader question: if some collaborators use other models such as Gemini or Claude, why would the shared space be a proprietary chatbot app? Human relationships do not align neatly with vendor ecosystems. Forcing them into one feels artificial. A more grounded architecture might look like: - models delivered as utilities through standard APIs - a memory layer owned and controlled by the user - interfaces chosen to suit context, not platform strategy The AI becomes an invisible capability rather than a visible participant trying to be social. ## Where this could work There are places where collaborative intelligence fits naturally. Study groups experimenting together. Teams preparing documents. Shared planning sessions. These are tasks, not chit-chat. In those environments, a bot can enter as a tool requested by the group, not an entity sitting at the table by default. ## Closing This is not a rejection of AI in communication. It is a reminder that our most important conversations rely on trust, timing, and the comfort of speaking freely. Technology can support that, but only if it understands when to stay quiet. ### Further reading - [ChatGPT wants to enter your group chats. I tried it — and it’s a little weird.](https://www.businessinsider.com/chatgpt-group-chats-why-needed-i-tried-2025-11) - [OpenAI is launching group chats in ChatGPT](https://www.theverge.com/news/825544/openai-chatgpt-group-chats-launch) ### The Workspace Advantage: Google’s Quiet Lead in the AI Shift URL: https://rob-hoeijmakers.ghost.io/google-gemini-workspace-future/ Last updated: 2026-05-10T08:53:45.000Z ## The centre of gravity is starting to move For most of my personal thinking, I still reach for ChatGPT. It feels quick, intuitive and multilingual in a way that supports my habits. Yet, as more of my work sits inside Google Workspace, there is a quiet shift. Gemini 3 makes it difficult to ignore Google’s position. Not because their app is delightful. Because their AI is embedding itself into the substrate of work. ## Two different kinds of intelligence I am not a financial analyst. My interest is how the product fits into everyday life, and what incentives shape its evolution. Still, some basic facts matter. Google owns the environments where billions collaborate and store knowledge: mail, documents, identity and devices. They have their own chips, distribution and deep access to organisational context. OpenAI, despite exceptional models, does not have that infrastructure. It buys compute and earns integration deal by deal. This leads to two forms of intelligence. • One stays close to the human mind. • The other sits inside the systems that run organisations. Gemini 3 highlights the distinction. [Andrew Lokenauth | TheFinanceNewsletter.com (@fluent.in.finance) on ThreadsSam Altman admits OpenAI is in trouble after Gemini 3 release in a leaked memo, per The Information. Google has the World’s data, its own chips, and unlimited cash. Google has YouTube, Search, Gmail, Maps, and Android. Billions of users. No other company has that combo. OpenAI is projected to lose $7B by 2028\. They’re burning $8.5B a year trying to compete. OpenAI’s valuation is $500B but its revenue is $13B. That’s a 38x revenue multiple. Google trades at 7x revenue with actual profits.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Pcnemah90K8-5.png)Threads![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/331268954_506850924730132_1701068779772126227_n.jpg)](https://www.threads.com/@fluent.in.finance/post/DRX1P2OEYCl?xmt=AQF07EextlCVs6oVfHKpg3wPCGSwyUY6r6gd7BpfaHNF5A) ## How I work today: power with friction I already use ChatGPT Team with connectors. It can read documents in my Drive, search my Gmail, and extract the context I need. In principle, I have access to the information layer. In practice, there is still a gap between understanding and action. I can draft in ChatGPT. I can check facts inside my files. But replying to email still means copying and pasting. Creating or updating documents requires manual steps and permission juggling. The intelligence knows the context, yet remains slightly outside the workflow. It feels as if my thinking lives in one place, and my work lives in another. [ChatGPT Connectors: The Emergence of a New Interface LayerChatGPT just got a new superpower: connectors. Real-time access to your tools like Drive, HubSpot and Notion—without uploading a thing.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-369.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_3877-1-1.jpeg)](https://rob-hoeijmakers.ghost.io/chatgpt-connectors/) ## From helpful advice to actual execution When AI sits directly inside the tools we use, it does more than understand. It can act. It can send the message, schedule the meeting, update the document, move the work forward. That makes it more influential than a good conversation on the side. It may still feel awkward at the surface. The interface is not yet elegant. But the real power lies underneath, in the systems of permissions, records and coordination. Once intelligence controls those layers, it stops being optional. ## Google’s own dilemma Search still prints money. The model depends on sending users out into the web, and being paid for those journeys. A conversational AI gives the answer directly. No journey, no revenue. Google must decide whether to protect the past or build the future fully. It cannot do both without cost. [The End of Google Search (as we know it)Google didn’t warn me. It just erased my blog. What looked like a bug turned out to be a glimpse into the future of search—and it’s not built for us anymore.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-370.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_2069-5.jpeg)](https://rob-hoeijmakers.ghost.io/the-end-of-google-search-as-we-know-it/) ## Our dilemma too: the shape of dependency With ChatGPT, my dependency feels like a dependency on an engine. Painful to swap out, but not impossible. With Google, the dependency goes deeper. Knowledge, communication and archives accumulate inside one ecosystem. Most users will never extract them again. That is not an accident. It is part of the value proposition. Google can bundle intelligence into what organisations already pay for. Gemini appeared in my Workspace account without debate. ChatGPT, by contrast, is an additional subscription. Cost alone shapes choices. ## Why Europe should pay attention When intelligence becomes a utility, the owners of distribution gain power over identity, pace and rules. Convenience today can be structural dependence tomorrow. Sovereignty is no longer only about where data lives. It is about who mediates judgment. ## Two trusted relationships I trust **ChatGPT** more as a companion. I trust **Google** more to handle the messy details of work. Which trust will matter more. The one guiding our thinking, or the one controlling our tools. ## A beginning, not an end Gemini 3 marks a deeper entanglement between AI and the environments where decisions are made. The important question is no longer which model is best, but who we allow to sit at the centre of our working lives, and how easily we can move on if we decide to take that centre back. --- P.s. Where will the price for all this go? I find this an interesting and thought provoking view: > Everyone thinks Google will drop the price of Gemini 3.0 to zero because their chips (TPUs) are so cheap. > > I ran the game theory on this, and I think we have it backwards. > > Google actually has the strongest incentive to keep AI expensive. Here’s the trap. 🧵 > > 1/ The "Obvious"… > > — Carlos E. Perez (@IntuitMachine) [November 22, 2025](https://twitter.com/IntuitMachine/status/1992201738409951550?ref%5Fsrc=twsrc%5Etfw) ### Further reading - [🚀 Can AI escape Google’s gravity well?](https://www.exponentialview.co/p/can-ai-escape-googles-gravity-well) - [OpenAI vs Google Gemini vs Meta AI: Comparing AI for Businesses](https://www.humai.blog/openai-vs-google-gemini-vs-meta-ai-comparing-ai-for-businesses/) ### Introducing the EU’s Digital Omnibus Proposal URL: https://rob-hoeijmakers.ghost.io/eu-digital-omnibus-proposal/ Last updated: 2026-05-15T12:24:41.000Z ## Where I come from Over the past months I have been looking more closely at how Europe is shaping the digital environment. It is a mix of personal curiosity and practical reality. The work I do with organisations depends on how we use data, automate processes and introduce AI into services that matter. Regulation is part of that reality, not a footnote. ## What has happened The European Commission has introduced the Digital Omnibus Proposal. On paper it sounds technical. In practice it adjusts a wide group of regulations that define how digital products work in Europe: GDPR, the AI Act, the Data Act and rules around cookies and consent. The ambition is to simplify. Regulations that were written in parallel now need more alignment. The Commission wants to reduce overlap, remove friction and give Europe’s builders more confidence to innovate. [Digital Omnibus Regulation ProposalThe Digital Omnibus proposal includes a set of technical amendments to a large corpus of digital legislation, selected to bring immediate relief to businesses, public administrations, and citizens alike, to stimulate competitiveness.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon-12.svg)Shaping Europe’s digital future![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/Omnibus_package_2YbNhISfcKk71dywR9oiLQYC6g_214993.jpg)](https://digital-strategy.ec.europa.eu/en/library/digital-omnibus-regulation-proposal) ## What it could mean in practice The effects will arrive gradually and many will only notice when deadlines become visible. But it is already a signal that Europe is recalibrating how AI and data governance should work in everyday practice. For people creating digital services, this matters. Architecture, governance and product design move slowly. Early awareness helps. ## The human side A recurring pattern in my work is how strongly people react to consent screens, forms and processing statements. Many do not understand what is happening to their data and instinctively assume the worst. That mistrust slows adoption, even of services that are safe and useful. The current system often asks users to trust blindly. When trust must be given without clarity, suspicion becomes rational. If the Digital Omnibus can reduce opacity rather than only reduce burden, that would be a meaningful gain. ## What happens next This is a proposal from the European Commission. For it to become law it must now be negotiated and amended by the European Parliament and the Council of the EU. The final outcome could shift during that process. Timelines are still emerging, so it is worth keeping an eye on the next steps. ## Why I am paying attention I am not trying to become a legal interpreter. My interest is what this means for those designing and deploying real systems. Better aligned rules can remove hesitation. Clearer expectations can help smaller players act with confidence. A Europe that protects rights and still enables progress is a Europe with momentum. This update is simply a marker. Europe’s pursuit of digital and AI capacity is entering a more practical phase. I plan to follow how this unfolds from a builder’s perspective. 💡 In plain terms, the EU’s Digital Omnibus is a regulation for regulating regulations. It is not a new digital law in itself, but a package of amendments to existing rules such as the AI Act, GDPR, ePrivacy, the Data Act, NIS2 and DORA. Its promise is simplification: fewer overlaps, clearer obligations and less administrative burden. Its risk is that “simplification” becomes a polite way to reopen protections that were politically hard to win. ### Further reading - [Finding My Place in Europe’s AI Future](https://rob-hoeijmakers.ghost.io/finding-my-place-in-europes-ai-future/) - [Europe is scaling back its landmark privacy and AI laws](https://www.theverge.com/news/823750/european-union-ai-act-gdpr-changes) - [Playing the EU Game: Learning the Digital Rules by Getting in the Game](https://rob-hoeijmakers.ghost.io/playing-the-eu-game/) ### Excel and the future cockpit of business logic URL: https://rob-hoeijmakers.ghost.io/excel-and-the-future-cockpit-of-business-logic/ Last updated: 2026-05-10T08:51:17.000Z Excel has always been more than a spreadsheet. For decades it has been the place [where business logic quietly lives](https://rob-hoeijmakers.ghost.io/excel-hidden-operating-system-of-business-reasoning/). Not in software systems designed for control, but in the free space where analysts, planners and managers actually think. What interests me today is how this space is changing as AI becomes a co-worker rather than a tool. And why Microsoft’s CEO Satya Nadella keeps returning to Excel when he talks about the future of knowledge work. It suggests that spreadsheets are not a relic of a pre-AI world. They may become one of the primary surfaces where humans steer autonomous agents. ### What spreadsheets already are A spreadsheet is a modelling environment. It shows reality and scenarios at once, across time and uncertainty. It is where risks, margins, and alternatives can co-exist and be compared. It is also a subtle form of programming without the ceremony: a model emerges step by step, and the “code” is simply what the sheet becomes. Nadella makes this explicit when he calls spreadsheets **Turing complete**. In principle they can express any logic a general computer can. Which means they are not just containers for numbers. They are a programming language that millions know by instinct, without naming it as such. > Excel Agent is not a UI-level wrapper. It’s actually a model that is in the middle tier. … I have a full understanding of all the native artifacts of Excel. Because if you think about it, if I’m going to give it some reasoning task, I need to even fix the reasoning mistakes I make. That means I need to not just see the pixels, I need to be able to see, ‘Oh, I got that formula wrong,’ and I need to understand that. - Satya Nadella. ### Rows, columns and the human mind There is a reason this format never felt like learning a new skill. Tables appear across all recorded history. They give us a humane geometry for complexity. Rows group cases, columns define attributes, and the grid allows order without forcing abstraction too early. You can just see what is happening. That deep familiarity matters. New interfaces rarely displace structures that are shaped by human cognition. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/11/1-SCjXvxDE8zyDXs2FdxHyfg.png) ### Excel as the IDE we already use Nadella pushes the idea further. He calls Excel an IDE (Integrated Development Environment) for people who are not thought of as programmers. In software development the IDE is the cockpit where you run, test, and inspect the behaviour of code. Something similar happens in Excel: forecasting, checking dependencies, tracing errors, comparing scenarios. And that cockpit becomes even more important when thousands of AI agents are working in parallel. He describes a shift from **micro execution** to **macro delegation, micro steering**. We will give missions to agents, then watch for anomalies, approve changes, redirect effort. The work will be choosing what matters, not executing each step. For that you need a surface where the structure is visible and steerable. Something like a spreadsheet. ### The sheet joins the workflow This is not about adding AI features to an old tool. It is about moving business logic into a live environment where AI participates: • agents draft projections and proposals directly in the grid • humans intervene at the right moments • the surface shows risk, uncertainty, and alternatives • acceptance and rejection become learning signals for the model In this reading, Excel is not being replaced. It is completing a long arc. From modelling… to programming… to a place where humans and agents collaborate in real time. ### The question that follows If the spreadsheet becomes a cockpit for autonomous agents, what does that mean for governance, for transparency, for the authority of human judgement? We are not only updating a tool. We are updating the boundary between what people decide and what systems propose. The table might remain familiar, but the activity inside it will feel new. It is worth paying attention. The future of AI in organisations may not arrive as a new interface. It may arrive in the grid we already trust. > never forget Satya Nadella in 1993 as a Microsoft Technical Marketing Manager showing how Excel works [pic.twitter.com/q7evvELa0r](https://t.co/q7evvELa0r) > > — Bearly AI (@bearlyai) [November 19, 2025](https://twitter.com/bearlyai/status/1991195520489189497?ref%5Fsrc=twsrc%5Etfw) ### Further reading - [Excel, the Hidden Operating System of Business Reasoning](https://rob-hoeijmakers.ghost.io/excel-hidden-operating-system-of-business-reasoning/) - [Satya Nadella — How Microsoft is preparing for AGI](https://www.dwarkesh.com/p/satya-nadella-2) ### Finding My Place in Europe’s AI Future URL: https://rob-hoeijmakers.ghost.io/finding-my-place-in-europes-ai-future/ Last updated: 2025-11-20T10:59:08.000Z Over the past year I have been circling the same question from different angles: **what is the actual playing field in which I work when I build AI systems in Europe.** Not just in a narrow, technical sense, but politically, economically and geopolitically. Who sets the rules. Where does the money come from. What is Europe trying to become. When the [Draghi report on European competitiveness](https://rob-hoeijmakers.ghost.io/europe-between-innovation-and-regulation/) appeared, I read it and felt both alarmed and energised. It was a signal that the ground under our work is shifting. More recently I spent time with two substantial pieces: [*Europe in the Age of AI*](https://institute.global/insights/tech-and-digitalisation/europe-in-the-age-of-ai-how-technology-leadership-can-boost-competitiveness-and-security) and [*The Constitution of Innovation*](https://constitutionofinnovation.eu). Together they forced me to ask a simple question: **where do my own views sit in this landscape, and what does that mean for my company and the organisations I advise.** This is a developing arena for me. My perspective will evolve. But the topic became too important to treat as background noise. The work we do is shaped by forces that reach far beyond project scopes and product roadmaps. When I write Europe here, I mean the Union and the wider ecosystem of companies, institutions and regional networks that give it real capability. ### Quick takeaways - Europe’s AI debate is now about power, security and the economy. - Fragmentation, hesitation and underinvestment slow us down. - My focus is shifting toward local pioneers and practical capability building. ## Why this became urgent for me I run [an agency](https://schmuki.nl) that designs and implements AI-driven tools for healthcare, public services and mid-sized companies. These organisations operate under European rules and budgets, and their ability to modernise is shaped by Europe’s strategic choices. Another element matters too. Public attitudes toward technology in Europe are oddly ambivalent. It is not fear or moral outrage, but a quiet scepticism. As if digitisation is something that happens elsewhere and mostly brings inconvenience. The Blair report, Europe in the Age of AI, confirms this intuition: **digitalisation is not a priority for European voters**, and innovation and AI consistently **rank near the bottom of topics citizens want the EU to focus on**. This creates a political environment where urgency is low and hesitation feels safe. That makes it even more important to understand the terrain: what is pulling us forward and what is holding us back. ## Two lenses on Europe’s AI future One text,[ *Europe in the Age of AI*](https://institute.global/insights/tech-and-digitalisation/europe-in-the-age-of-ai-how-technology-leadership-can-boost-competitiveness-and-security), frames AI as a geopolitical turning point. Technological capability is presented as a requirement for security, prosperity and democratic confidence. It calls for political leadership, serious investment in compute and energy, and a more coherent narrative around technology in daily life. The other, *The Constitution of Innovation*, focuses on economic mechanics. It argues that the Union’s prosperity engine has stalled: the internal market remains incomplete, regulation piles up, incumbents are protected and risk-taking suffers. The answer is to restore economic dynamism, with fewer barriers and faster scaling across borders. [The Constitution of Innovation: A New European RenaissanceAn academic paper examining Europe’s innovation challenges and proposing a constitutional framework for economic renaissance.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/apple-icon.png)The Constitution of InnovationLuis Garicano![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/og-image.png)](https://constitutionofinnovation.eu) One starts from politics and security. The other from markets and law. Yet they strike the same nerves. ## Where they agree - **Fragmentation** is the central obstacle. - **Underinvestment** leaves Europe consuming more than it creates. - **Hesitation** preserves the status quo at the expense of future capacity. - **Sovereignty** is about leverage and capability, not isolation. Reports have been diagnosing these weaknesses for years. Execution remains thin. ## The field where I actually play My work sits far from Brussels. It lives in: - a health organisation where front-line staff need instant access to procedural knowledge - a social-care team trying to manage complex decisions with limited time - a business struggling to align its website, data and customer workflows with AI-driven search and support Recently, I worked with a social-care organisation that introduced a lightweight conversational system to support case workers. It did not transform everything at once, but it freed time, improved clarity, and gave the team confidence to take the next step. These are not prototypes. They are real performance improvements, realised cautiously and under constraints. They also reveal something hopeful: - there is strong local ingenuity - European values like privacy and dignity can be translated into practical systems - when the basics are in place, AI can improve everyday institutions without breaking what people trust This is why I look beyond continental blueprints and toward **local pioneers**: places where capability already exists and could scale if given room. One idea in the *The Constitution of Innovation* caught my attention: a proposed “28th regime” for companies that want to grow across the EU. Instead of navigating 27 different national corporate rules, firms could opt into a single European legal framework, fully digital and directly connected to the internal market. It is presented as a way to turn Europe’s scale into actual innovation power. Still early, still debated, but it feels like a concrete step worth examining if we want Europe to remain competitive in AI and beyond. ### The 28th regime as a structural escape hatch Europe’s single market is still mediated by 27 corporate systems. A company that wants to operate across borders must translate itself into each national regime. It scatters energy. It favours remaining small. The “28th regime” pushes in the opposite direction. Instead of harmonising everything through compromise, it offers an optional EU-level framework that founders can adopt from the start. One legal identity capable of expanding across the Union without continual reinvention. This shapes how capital flows, how talent moves, how IP is protected, and how quickly organisations can reach European scale. It is part of the foundation on which a confident digital economy must rest. 💡 [****EU-INC**](https://www.eu-inc.org) A coalition of European founders and investors is already campaigning for such a structure. One central EU registry. One set of investment documents. Stock options that work everywhere in the Union. A practical attempt to turn the 28th regime from concept into operating reality. A structure like this would not solve everything. But it would remove one of the quiet constraints that keep European innovation fragmented, localised, hesitant to grow. ## A working direction Trying to locate myself in this landscape, a direction emerges that is both realistic and ambitious. These are the places where capability can grow: grounded commitments that help institutions move from caution to confidence. 1. **Start from existing strengths** Focus on sectors with data, expertise and strategic value: healthcare, public administration, industrial automation. 2. **Let pioneers lead and remove their barriers quickly** Scale what works across borders. Use practice to guide policy. 3. **Think sovereignty as interdependence with leverage** Control what matters. Stay open where it benefits citizens and industry. 4. **Tie regulation to delivery** Judge rules by whether they enable responsible deployment. 5. **Accept creative destruction** Make space for new entrants and new approaches. ## What this means for my work This direction reshapes the questions I ask and the advice I give. - Track not only technology, but energy policy, capital flows and regulatory change. - Design for cross-border reuse, not single-site dependency. - Be explicit that AI adoption is part of a wider European shift toward resilience and competitiveness. The micro and macro are not separate worlds. They meet in the systems we build for real use. 💡 Europe’s digital rule-set is quietly entering a moment of potential change. Alongside the big visions and political arguments in the Blair and Manifesto documents, the European Commission is preparing a [****Digital Omnibus Package**](https://ec.europa.eu/commission/presscorner/detail/en/ip%5F25%5F2718) that would consolidate and simplify the dense web of data, AI and platform regulation. If it lands as intended, the regular landscape that has shaped innovation in Europe for the past decade could look very different. Some see a chance for more clarity and room to build. Others worry that in the name of simplification essential protections and sovereignty could weaken. ## Closing Europe is not one actor. It is a layered field where policy, national governments, companies and research networks shape the possible together. The future will not emerge from directives alone, but from places where people build new capabilities and learn to trust their own momentum. This map helps me decide where I can contribute to that fabric. Local strength that can scale when it works. Public actors that know when to support and when to step aside. A Europe that builds from what is already strong. A Europe that trusts its pioneers. A Europe that enables Europeans to move. --- [EU–INC — One Europe. One Standard. — Pan-European legal entity.EU–INC is a proposal for a pan-European standardized legal entity to unlock pan-European startup scaling.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/hTcgaWRewUNZeB6wVrfBSXk9Mn0.svg)![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/DywwAp236AdnECpow4p9VcGluY.png)](https://www.eu-inc.org) [Signing the Future: How a Digital Incorporation Made eIDAS RealA first-hand look at how eIDAS and digital signatures replace paper rituals with cryptographic trust.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-365.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_0188-1.jpeg)](https://rob-hoeijmakers.ghost.io/digital-signature-eidas-experience/) [🔮 How Europe outsourced its future to fearWhen precaution becomes stagnation — a guest essay![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/https-3A-2F-2Fbucketeer-e05bbc84-baa3-437e-9518-adb32be77984.s3.amazonaws.com-2Fpublic-2Fimages-2Faa7453ed-eb02-46d7-b523-74d86b1c1e24-2Fapple-touch-icon-180x180-10.png)Exponential ViewExponential View![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/https-3A-2F-2Fsubstack-post-media.s3.amazonaws.com-2Fpublic-2Fimages-2F54eb0ad1-5328-4bca-9772-f7b9cb93e1a4_730x411.png)](https://www.exponentialview.co/p/how-europe-outsourced-its-future ) [Introducing the EU’s Digital Omnibus ProposalThe Digital Omnibus Proposal signals a shift in Europe’s approach to data and AI governance. Early awareness matters for those building digital tools.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-368.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_7134-1.jpeg)](https://rob-hoeijmakers.ghost.io/eu-digital-omnibus-proposal/) ### Notes - [The 28th legal regime: enhancing European competitiveness and innovation](https://www.loyensloeff.com/insights/news--events/news/the-28th-legal-regime-enhancing-european-competitiveness-and-innovation/) - [Establishing the 28th regime in Europe: A unified legal framework to support growth and business](https://www.ecmi.eu/publications/depth-analysis/establishing-28th-regime-europe-unified-legal-framework-support-growth) - [The state of AI in 2025: Agents, innovation, and transformation](https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai) - [AI-gedreven groei: Nederland blijft achter bij de VS](https://www.ing.nl/zakelijk/economie/nederland/ai-gedreven-groei-nederland-blijft-achter-bij-de-vs) ### Review of The Reckoning, in the light of the AI boom URL: https://rob-hoeijmakers.ghost.io/the-new-reckoning/ Last updated: 2026-05-10T08:53:46.000Z The Reckoning by [Jacob Soll](https://www.linkedin.com/in/jacob-soll-7b4817106/) traces how accounting shaped states, markets and the very idea of value. It shows how ledgers do not simply record reality. They help create it. I read the book some time ago, but came back to it now because we are in the middle of a new accounting challenge. The AI boom depends on investment and on the stories that justify that investment. This piece is a review of Soll’s book, with examples from today’s digital economy. ## Key points - Accounting choices decide what counts as value - Capital investment in AI relies on optimistic assumptions about time - Depreciation schedules for GPUs shape expectations of growth and profit - Numbers often smooth over uncertainty and risk - The past suggests that reality eventually forces a reckoning ## From ledgers to data centres Soll follows a long tradition of people trying to master the future by writing it down. Double entry bookkeeping allowed merchants and states to stretch time, to commit to future income before it arrived, and to hide fragility behind precision. It created accountability, but also imagination. Growth often needed belief first and proof later. Modern cloud infrastructure feels similar. Hyperscalers and AI companies invest billions into data centres, chips and power. These assets are justified on the basis of a horizon: three years of useful life for a GPU, or perhaps six. The difference between those two assumptions can turn a narrow margin into a headline success story. When model updates accelerate and hardware cycles shorten, the ledger is forced to catch up with physics. > “I didn’t want to get stuck for four years, five years of depreciation on one generation…” — Satya Nadella, Microsoft Accountants decide how quickly servers and GPUs lose value on the books. Investors then take those numbers as signals. But the reality beneath the figures may move differently. A chip that is technically fresh can already be strategically old if the next generation trains models at half the cost. In the AI economy, value decays along performance curves, not in tidy annual increments. Soll’s history is full of moments where growth masked fragility. Narratives filled the gap between ambition and evidence. When confidence broke, books were corrected, sometimes brutally. The question today is not only whether companies mislead. It is how long optimism can overshadow the material constraints of energy, location, supply chains and latency. ## What we choose to see A ledger is a lens. It highlights some aspects and fades others. If cost is pushed into the future, present gains look larger than they are. If complexity is simplified, risk becomes hidden. In AI, where the cost of progress is both upfront and ongoing, it is tempting to treat almost everything as long lived capital and relatively little as immediate expense. We are still learning how to record assets that learn and decay at the same time. Training improves value. Rapid technical change, however, accelerates obsolescence. Traditional bookkeeping is not fully equipped for this tension. The choice of a three year or a six year horizon for hardware life is not just a technical footnote. It encodes a view of how fast the world will move. Soll never presents accounting as pure deception. He shows that accounting is a judgement. Someone decides what future to believe in, and how much uncertainty to expose on the page. ## The book itself As a book, The Reckoning is broad in scope. Soll moves from Renaissance merchants to emerging nation states and modern corporations, showing how shifts in bookkeeping enabled expansion but also created new fragilities. For readers in digital infrastructure, finance or technology, it sharpens the sense that numbers are always choices and that those choices carry political intent. ## Closing reflection This book helped me see the present moment more clearly. For anyone working with AI strategy, energy heavy compute or digital infrastructure, The Reckoning offers a historical grounding that feels necessary. It reminds us that every boom is fuelled by expectations and that eventually the books must match the world. I find it highly relevant and very much worth reading. [Jacob Soll - USC DornsifeUSC Dornsife - Jacob Soll![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/cropped-Favicon_Monogram_Gold-Black-270x270.png)USC Dornsife![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/photo_1042779.jpg)](https://dornsife.usc.edu/profile/jacob-soll/) [The Reckoning: Financial Accountability and the Rise an…A “brilliant” (Los Angeles Review of Books) history of …![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon-35.ico)GoodreadsJacob Soll![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/18210762-1.jpg)](https://www.goodreads.com/book/show/18210762-the-reckoning) [Nvidia’s ‘I’m Not Enron’ memo has people asking a lot of questions already answered by that memoThe Streisand effect strikes again![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/apple-touch-icon-22.png)The VergeElizabeth Lopatto![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/STKS521_AI_BUBBLE_C.jpg)](https://www.theverge.com/business/828047/nvidia-enron-conspiracy-accounting) ### Further reading - [Compute: A New Measure of the World](https://rob-hoeijmakers.ghost.io/compute-a-new-measure-of-the-world/) - [Tracing and Telling: From GPX Tracks to KML Stories](https://rob-hoeijmakers.ghost.io/gpx-kml-ai-maps/) ### What llms.txt can do for your website URL: https://rob-hoeijmakers.ghost.io/llms-txt/ Last updated: 2026-05-10T08:53:46.000Z Language models increasingly act as intermediaries between organisations and their audiences. If a client asks ChatGPT about your services, the response depends on what the model believes to be true about you. [Llms.txt](https://llmstxt.org) offers a simple way to nudge that understanding in the right direction. ## When AI becomes a source of truth More people ask ChatGPT about your organisation than you might expect. The answer they receive is based on whatever the model has learned about you so far. Sometimes that is recent and accurate. Often it is not. If language models are becoming a practical interface to your services, it helps to offer them a reliable starting point. ## Introducing llms.txt **llms.txt** is a simple idea. Just as robots.txt guides search engines, llms.txt offers guidance to language models. It is a plain text file that can describe: - what your organisation actually does - which pages or sources are authoritative - which terms are correct or preferred It does not give you strict control. It offers context that models can use to avoid guesswork. This idea is gaining traction. I recently exchanged thoughts with Pieter Versloot from [PlateCMS](https://www.platecms.com) on LinkedIn, who added useful pointers and examples from the CMS perspective. I will include that reference below. [I was getting seriously annoyed by how often ChatGPT and Gemini got Plate's product details and even our company facts wrong. So I dug into the best practices distilled from 1,500 llms.txt files and… | Pieter Versloot | 23 commentsI was getting seriously annoyed by how often ChatGPT and Gemini got Plate’s product details and even our company facts wrong. So I dug into the best practices distilled from 1,500 llms.txt files and applied them to our own setup. These are the key insights. Google currently indexes somewhere between 30,000 and 60,000 llms.txt files worldwide. The average llms.txt weighs just 9.8kb. That’s around 275 times smaller than the average modern webpage. In other words: extremely crawl-efficient and a real visibility lever if you care about how LLMs interpret your content and brand. At Plate, we saw a clear pattern: AI systems invented details about our product specs, team roles, and even our company history. We added the correct information to our llms.txt and started tracking how quickly models pick it up. If you want a reference example: https://lnkd.in/epDhS-KZ If your platform supports agent-to-agent interactions or exposes an MCP (Model Context Protocol), add those details too. We have an MCP for both our products, and until it becomes public, our llms.txt tells people where to request access. You can let AI draft your llms.txt as long as you give it clear instructions, examples, and a concrete goal. Ours was straightforward: stop hallucinations and get the facts straight. Creating the file, writing this post, and setting up a quick baseline of the most frustrating hallucinations took me about 30 minutes in total. For Plate CMS users it’s a few minutes of work, but the real value of our CMS products goes much deeper: keeping your content correct, consistent, and genuinely AI-ready instead of hoping models guess the right version. Plate Delta is built exactly for that. Want to see more? Just hit me up. Link to the original source: https://lnkd.in/evPGEebe | 23 comments on LinkedIn![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/al2o9zrvru7aqj8e1x2rzsrca-6)LinkedInPieter Versloot![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/1763027405135)](https://www.linkedin.com/feed/update/urn:li:activity:7394672905554440192/?commentUrn=urn%3Ali%3Acomment%3A%28activity%3A7394672905554440192%2C7394682471956586496%29&dashCommentUrn=urn%3Ali%3Afsd%5Fcomment%3A%287394682471956586496%2Curn%3Ali%3Aactivity%3A7394672905554440192%29&dashReplyUrn=urn%3Ali%3Afsd%5Fcomment%3A%287396195495435542528%2Curn%3Ali%3Aactivity%3A7394672905554440192%29&replyUrn=urn%3Ali%3Acomment%3A%28activity%3A7394672905554440192%2C7396195495435542528%29) ## Implementing it in practice I added [llms.txt to our own website](https://rob-hoeijmakers.ghost.io/llms.txt) to see what happens in the real world. Creating the file was the easy part. Making it reachable turned out to be the real challenge. My hosting provider Hostinger blocked requests from LLM crawlers by default. The file lived on the server but returned a 403 to any AI agent. It was effectively invisible. The fix was to move DNS to [Cloudflare](https://cloudflare.com) and let their CDN handle the traffic. Suddenly the file was accessible and visible in analytics, including visits from AI-related bots. 💡 ****Updated April 2026**: llms.txt is now served as clean plain text at [hoeijmakers.net/llms.txt](https://rob-hoeijmakers.ghost.io/llms.txt) via a Cloudflare Worker, replacing the previous Ghost page. The same Worker also serves a custom robots.txt — something Ghost Pro doesn't allow natively. Both files are managed in one place and trivial to update. Testing is straightforward. You can simply ask ChatGPT whether it can access your llms.txt. If something blocks access, it will say so and often suggest where the issue might sit. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/11/image-1.png) After implementation it started working for ChatGPT straight away ## Why this matters for digital professionals Your organisation already works hard to keep its messaging consistent. Llms.txt extends that effort to automated systems that summarise you for others. It will not guarantee perfect accuracy. It provides a baseline, a reference that can reduce outdated assumptions over time. ## What I am tracking next I plan to observe whether the presence of llms.txt influences how AI agents describe our business. If the effect is positive, this practice will likely spread further through the web development and [CMS](https://rob-hoeijmakers.ghost.io/content-management-system-cms/) ecosystem. The main lesson so far is straightforward. The barrier was not writing the file but making sure it is visible to the systems that need it. A small adjustment can prepare your website for a future where AI is often the first audience. ### Further reading - [The /llms.txt file – llms-txt](https://llmstxt.org) - [Guests That Should Behave](https://rob-hoeijmakers.ghost.io/guests-that-should-behave/) - [Flying Blind: Measuring Traffic When Your Readers Are Machines](https://rob-hoeijmakers.ghost.io/measuring-traffic-machines-bots/) - [When Bots Become Readers: Publishing in the Age of AI Crawlers](https://rob-hoeijmakers.ghost.io/when-bots-become-readers/) ### Where deeptech really happens URL: https://rob-hoeijmakers.ghost.io/where-deeptech-really-happens/ Last updated: 2026-05-10T08:53:46.000Z A few months ago, I [interviewed Alain le Loux of Cottonwood Technology Fund](https://rob-hoeijmakers.ghost.io/alain-le-loux-on-deeptech/). It helped me see deeptech as something rooted in physics, materials and advanced engineering. Not the quick-turn world of consumer software. But I still pictured technologies that fit in labs or data centres. Then I came across [Allseas](https://www.allseas.com/en). An offshore construction company with Dutch roots working on nuclear reactors for ships. That made the definition stretch again. This piece is about that shift in perspective. And why I think it matters for Europe. ## Quick takeaways - Deep tech is physical as much as digital - Financing models shape which ideas become real - Industrial companies sometimes drive breakthroughs that startups cannot - Europe’s strengths may lie in places we rarely celebrate - R&D investment is strategic infrastructure, not a luxury ## What Alain le Loux taught me From Alain, I learned how difficult it is to finance deep tech at the early stage. Cottonwood invests long before the market is clear. Their risk is scientific uncertainty. Their reward, when it works, is transformative. This model fits well when technologies can scale quickly and generate global demand. But it becomes harder the moment the innovation is big, slow and tightly regulated. I thought I had absorbed that lesson. Then something unexpected tested it. [Investing in Atoms: Alain le Loux on Building the Deeptech FutureDeeptech demands patience, guts, and hands-on conviction. This article profiles Alain le Loux and the rare venture model behind Europe’s hardware hopes.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-348.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/C8095225-ED9A-4896-9EC3-F900E3354459-1-2.jpeg)](https://rob-hoeijmakers.ghost.io/alain-le-loux-on-deeptech/) ## A surprising example Allseas is a private, family-owned group with its engineering heart in the Netherlands. They design ships that can lift entire platforms out of the sea. Now they are investing [hundreds of millions of euros in a small modular reactor](https://nltimes.nl/2025/11/17/dutch-offshore-giant-allseas-invests-eu300-million-nuclear-reactor-power-fleet). The headlines treat the reactor as a climate move. That is certainly part of it. But I suspect an equally important motive lies in industrial sovereignty: securing energy independence for vessels that perform demanding work far from shore. This is deep tech of a very physical kind. It cannot be rushed. It requires trust, long planning horizons and a tolerance for complexity. Startups rarely get the space to operate like that. [Allseas delivers ground-breaking engineering and construction solutions for the offshore energy industry.We are a major contractor in the global offshore energy industry, leading the way in subsea pipeline construction and transport, installation and removal of offshore facilities. Pioneers in heart and soul, we integrate cutting-edge innovation with engineering excellence to meet the market’s ever-changing needs. Through our commitment to facilitate the energy transition through sustainable, technology-driven solutions, we are securing the future.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon-32.ico)Allseas![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/image.png.webp)](https://www.allseas.com/en) ## How money shapes deep tech My interest in R&D has grown sharply in recent years. I see it disappearing as a share of the Dutch economy, even as we talk loudly about innovation. That contradiction bothers me. Allseas reminded me that there are still companies willing to put serious capital into the hardest questions. Their investment looks like corporate R&D funded from their own balance sheet. The Dutch state appears mainly in a regulatory role, providing safety oversight and public legitimacy. This is not a typical venture story. There is no pitch deck. No exit. The value is realised in future operational capability and reduced dependence on scarce fuels. It makes me think we underestimate how often Europe’s major innovations are financed by industry itself. [Researchers warn: the Netherlands needs more ASML-style innovators to stay competitiveThe Netherlands needs more companies like ASML to drive innovation. For years, spending on research and development (R&D) has lagged behind the European target. ASML is investing a lot of money in innovation, but the chip machine manufacturing company from Veldhoven spends more than the next 10 companies in the R&D top 30\. This makes the Netherlands vulnerable, TNO researchers warn.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/icon60x60@3x.png)NL TimesNL Times![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/ASML_Containers_RZY_VRR_AKP_8_May2021.jpg)](https://nltimes.nl/2025/07/18/researchers-warn-netherlands-needs-asml-style-innovators-stay-competitive) ## Where Europe stands I want to be careful here. It is too early to say that Europe as a whole is absent. Funding and cooperation may appear later. Still, the early signal is instructive. When technology becomes big and slow, Europe’s support systems can feel hard to access. Innovators sometimes move faster by acting nationally first. That tension sits at the heart of the European project: the desire for sovereignty, and the difficulty of aligning effort at scale. We need both coordination and decisiveness. Right now, the balance is not easy. ## Bits and atoms together Even in the world of steel and reactors, digital capabilities are pivotal. Simulations, control systems and verification are where most of the risk is reduced. The entire safety case lives first in software, not in metal. There is a quiet convergence underway. The infrastructure that builds the physical world increasingly depends on the same intelligence that powers digital services. If we imagine the future of technology as purely virtual, we miss where the largest investments will flow: into systems that merge computation with energy, materials and logistics at scale. [Compute: A New Measure of the WorldCompute once meant calculation. Now it shapes work, art, and power. An invisible current running through the modern world’s every action.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-351.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_0897-2.jpeg)](https://rob-hoeijmakers.ghost.io/compute-a-new-measure-of-the-world/) ## Industrial strengths we overlook We often describe innovation in terms of what we lack: missing platforms, fewer unicorns. But there is another picture. Europe holds deep competence in nuclear engineering, maritime systems and regulatory governance. These are technologies that operate in the background until they fail, and failure is unacceptable. We rarely celebrate this kind of capability because it is not fast or visible. Yet it underpins climate action, industrial competitiveness and security. If we do not recognise these strengths, we risk underfunding the very sectors that carry the energy transition. Sovereignty is not only built in software. It is also welded, certified and maintained over decades. ## Conclusion This small discovery changed how I think. Deep tech is not a niche. It is a spectrum. From algorithms to reactors. And if I care about innovation, I have to care about how it is financed. R&D is not a nice-to-have. It is the part of the economy that builds the conditions for the rest. Europe talks confidently about sovereignty in the age of AI and energy transition. Delivering it will depend on supporting the kind of innovation that is heavy, slow and ambitious. The kind that does not show up on our screens, but moves quietly across the water. ### Further reading - [My Venture into Venture Capital: What I Learned About DeepTech Investing](https://rob-hoeijmakers.ghost.io/venture-capital-deeptech-investing/) - [An interview on the structural gaps holding Europe back in deeptech](https://rob-hoeijmakers.ghost.io/the-deeptech-dilemma-europe/) ### From Silicon to Intelligence: Understanding the Hardware Behind AI URL: https://rob-hoeijmakers.ghost.io/cpu-gpu-tpu-npu-explained/ Last updated: 2026-05-10T08:53:46.000Z It began with a short video. [FinninTech](https://www.threads.com/@tiffintech) explained the difference between TPUs and NPUs. A brief clip that suddenly made the invisible world of AI hardware tangible. [Tiffany Janzen (@tiffintech) on ThreadsWhat is the difference between NPUs and TPUs?! Here is a simple explanation! You’re going to start hearing about NPUs everywhere so it is good to understand why tech companies have become so obsessed with them 💡 #tech #stem #techexplained![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Pcnemah90K8-4.png)Threads![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/581125064_1364211398429336_6709421694528799212_n.jpg)](https://www.threads.com/@tiffintech/post/DQ9ZqbCkb2i?xmt=AQF0D5RHCALynlVo-UH8yolQTbjYw3kWnH4c4koanjtyNw) That curiosity sent me down a path connecting the chip in my phone to the massive processors that train models like ChatGPT. ### Quick takeaways - CPUs, GPUs, TPUs and NPUs form a spectrum of *specialisation*: from flexible generalists to highly efficient AI specialists. - The iPhone’s **Neural Engine** is Apple’s name for its NPU, a miniature AI processor for local tasks. - **FLOPS** and **TOPS** measure different kinds of computing power: precision versus speed. - Export limits on chips such as NVIDIA’s H100 show how computing power has become a geopolitical factor. ## The spectrum of specialisation Artificial intelligence may feel abstract, but it’s built on physical hardware — billions of transistors arranged for specific kinds of work. At one end stands the **CPU**, a flexible all-rounder that handles logic and control. Then come **GPUs**, vast grids of simple cores designed for parallel maths. Beyond those lie **TPUs** and **NPUs**, processors made specifically for neural networks. You can picture it as a line: > **CPU → GPU → TPU / NPU** > As you move right, flexibility decreases, but efficiency for AI tasks rises sharply. Where a CPU handles general tasks, a GPU multiplies matrices, a TPU accelerates training in Google’s data centres, and an NPU performs small-scale AI tasks efficiently on your device. ## The chip in your pocket Apple’s **A17 Pro** chip, used in the iPhone 16 Pro and newer models, combines three types of processors: a CPU for everyday applications, a GPU for graphics, and a **Neural Engine** for machine learning. This Neural Engine performs around **35 trillion operations per second**, powering on-device features such as transcription, photo recognition, and real-time translation. It consumes only a few watts, roughly a hundred thousand times less power than a data-centre GPU, yet fast enough for personal AI. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/11/apple-iphone-15-pro-a17-pro.jpg.webp) A17 Pro has the Neural Engine on it, Apple's NPU. ## FLOPS and TOPS: the language of compute **FLOPS** (*floating-point operations per second*) measure the ability to handle precise arithmetic -> needed for ***training*** large models. **TOPS** (*tera-operations per second*) describe simpler, lower-precision calculations -> ideal for ***running*** those models efficiently. Training requires floating-point accuracy and immense power; inference, which happens on your phone, can use integer maths to save energy. In short: GPUs and TPUs are measured in FLOPS, NPUs in TOPS. ## TPU vs GPU: same idea, different philosophy A **GPU** is a programmable engine for parallel work and it is built for graphics, later adopted for AI. A **TPU** is Google’s own design: a *tensor processor* built from the ground up for machine learning. It’s not a GPU, but it draws on the same principle and that is performing many operations in parallel. While GPUs remain flexible, TPUs are hard-wired for the algebra behind neural networks, making them faster and more efficient for that single purpose. 💡 Coming soon: Microsoft’s Maia. Maia is Microsoft’s own AI accelerator, optimised for transformer workloads. Functionally it resembles Google’s TPU family, but it has its own architecture, software stack, and integration into Azure. ## The far end of the spectrum In data centres, processors such as NVIDIA’s **H100** or **B100** dominate. Each consumes hundreds of watts and delivers several **petaflops** of performance. These chips now sit at the centre of export restrictions, because such computing capacity determines who can train the next generation of large models. To comply with U.S. limits, NVIDIA built slower versions (A800, H800) for the Chinese market. It is the same hardware, with reduced interconnect speed. The boundaries of computing power have become geopolitical borders. 💡 The NVIDIA H100 and Google’s TPU both power today’s AI revolution, but they aren’t one-to-one rivals. The H100 is a flexible, general-purpose GPU evolved for deep learning; the TPU is a purpose-built tensor processor optimised for Google’s own ecosystem. They meet at the same goal, **accelerating neural computation*, from opposite ends of the design spectrum. ## From abstraction to atoms Once you see AI through its hardware, it feels less ethereal. Every neural network, from the model in your phone to the ones shaping global research, depends on physical constraints: heat, energy, and silicon. Understanding this spectrum, *from the Neural Engine in your pocket to the Tensor Processor in Google’s data halls*, brings AI down to earth. It reminds us that intelligence, however artificial, still runs on very real machinery. > Wow. Classic Jensen style, he ended the Nvidia vs. custom ASIC competition for good. 🫡 > > The level of confidence with which he explains. 🎯 > > He was answering to UBS research analyst question on how custom ASICs will affect NVIDIA or how they are going to compete with custom ASIC.… [pic.twitter.com/dc5Y3tQSDh](https://t.co/dc5Y3tQSDh) > > — Rohan Paul (@rohanpaul\_ai) [November 16, 2025](https://twitter.com/rohanpaul%5Fai/status/1990154716475802103?ref%5Fsrc=twsrc%5Etfw) ### Further reading - [Infinite scale: The architecture behind the Azure AI superfactory - The Official Microsoft Blog](https://blogs.microsoft.com/blog/2025/11/12/infinite-scale-the-architecture-behind-the-azure-ai-superfactory/) - [From Sand to Software: A Whistle-Stop Tour of the AI Value Chain](https://rob-hoeijmakers.ghost.io/ai-value-chain-from-quartz-to-cloud/) - [Compute: A New Measure of the World](https://rob-hoeijmakers.ghost.io/compute-a-new-measure-of-the-world/) - [The Neural Engine Does Not Run Your LLM](https://rob-hoeijmakers.ghost.io/the-neural-engine-does-not-run-your-llm/) - [Running Gemma 4 on Your iPhone](https://rob-hoeijmakers.ghost.io/running-gemma-4-on-your-iphone/) - [Running a Local LLM on Your iPhone](https://rob-hoeijmakers.ghost.io/running-a-local-llm-on-your-iphone/) ### Flying Blind: Measuring Traffic When Your Readers Are Machines URL: https://rob-hoeijmakers.ghost.io/measuring-traffic-machines-bots/ Last updated: 2026-05-10T08:53:47.000Z [Youp van der Graaf, a freelance data and CX consultant ](https://data.youpvandergraaf.com)with long experience in analytics and experimentation, asked how we can still measure website traffic when so much reading now happens inside language models. [Chris Schneemann](https://www.linkedin.com/in/chrisschneemann/) had raised similar concerns from his CX practice. Their questions landed because my own dashboards were quiet, yet my writing clearly circulated somewhere beyond them. For months, I felt like a pilot flying at night with the sky full of signals, but no instruments that showed where they were going. Then I opened Cloudflare and saw something new: lines of traffic labelled *ChatGPT-User*, *GPTBot*, *PerplexityBot*. Suddenly the invisible became faintly visible. ## **Quick takeaways** - A growing share of online reading now happens *within* AI systems, unseen by traditional analytics. - There are three forms of AI interaction: training crawls, conversational mentions, and human click-throughs. - The new attention economy is dominated by ChatGPT, Perplexity, and Google’s AI Overviews. - Cloudflare’s bot analytics offer the first faint radar of this new airspace. - We need measurement tools that can operate without cookies, scripts, or clicks. ## **The discovery** In my earlier pieces, [*The End of Google Search as We Know It*](https://rob-hoeijmakers.ghost.io/the-end-of-google-search-as-we-know-it/) and [*When Bots Become Readers*](https://rob-hoeijmakers.ghost.io/when-bots-become-readers/), I wrote about the shift from discoverability to readability. Google’s role as the great referrer was already eroding, and AI crawlers had begun to treat blogs as training material rather than destinations. This time the story became tangible. In Cloudflare’s new **AI Crawl Control** dashboard, I watched my site being accessed hundreds of times a day by *ChatGPT-User*. Not the training bot (*GPTBot*), but the live agent, a sign that real people, through ChatGPT, were drawing on my writing for answers. Plausible, my privacy-first analytics tool, showed nothing. Because there was no browser, no JavaScript, no session. Just raw retrieval: a machine fetching text to interpret for someone else. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/11/Metrics-LLM-s-Cloudflare.jpg) Cloudflare report on requests by bots over time. ## **Three kinds of interaction and only one we can see** When we talk about “AI traffic”, we’re actually talking about three different layers of activity: 1. **Training exposure** – when large language models crawl or scrape your site to *learn from it*. Done by bots like `GPTBot`, `ClaudeBot`, or `CCBot`. It leaves traces in Cloudflare, but it’s not tied to any user or conversation. Your content becomes part of a statistical memory, a silent ingestion. 2. **Mention in conversation** – when your work is *quoted, paraphrased, or referenced* inside an AI chat. The model may recall it from training or fetch it briefly in real time. You won’t see it. There’s no referrer, no metric, only the indirect sign when an AI agent reaches out to your page. 3. **Click-throughs from AI interfaces** – when a human follows a citation from ChatGPT, Perplexity, or Bing Copilot and actually visits your site. These are the only visible events in Plausible or analytics, with referrers such as `chat.openai.com` or `perplexity.ai`. And then there’s **Google’s AI Overviews**, the most polished black box of them all. We may appear there as sources, but we can’t know how often or in what form. Not even Google Search Console reveals it. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/11/Google-AI-overview.jpg) Google AI overview (AIO). ## **Three new gravity fields** Three new centres of gravity are reshaping how information moves across the web, and each bending attention in its own way, away from the open web and toward enclosed, conversational systems. ### **1\. ChatGPT — the conversation interface** It has quietly absorbed the space once occupied by a thousand blogs and forums. Users no longer click; they ask. The model responds, often by drawing from sources like ours, but the visit never reaches us. This is traffic without a trace, attention without visitation. ### **2\. Perplexity — the search-dialogue hybrid** A middle ground between ChatGPT and Google. Perplexity still crawls and cites, and some of that activity shows up in my Cloudflare logs. The volume is smaller, but more transparent: you can see when and what they retrieved. It’s a semi-permeable layer where references still pass through. [Perplexity, the AI-powered answer engineFrom accurate answers to cultural insights, Perplexity.ai does it all. Explore the future of search with this AI-powered answer engine.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-333.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_6848.jpeg)](https://rob-hoeijmakers.ghost.io/perplexity-the-ai-powered-answer-engine/) ### **3\. Google’s AI Overviews — the incumbent transformed** Google remains the web’s largest gateway, yet its *AI Overviews* are turning search results into compact answers. Our content still fuels them, but from the outside we see only silence, no signal, no metric. It’s the most refined black box: full visibility for users, none for creators. > In my work with organisations that deploy chatbots, the same broader issue returns: the effect on traffic, Google visibility, and a brand’s presence inside AI platforms like ChatGPT. > > [Chris Schneemann, Head of Sales, Conversed.ai](https://www.linkedin.com/in/chrisschneemann/) ## **Flying with new instruments** For the first time, [Cloudflare](https://cloudflare.com) offers a partial radar. It tells us when AI crawlers train on our content, and when assistants fetch it on behalf of users. It’s imperfect, only 24 hours of history, no context, but it’s something. From these traces, we can infer that writing has become a dual act: communicating with humans *and* maintaining semantic clarity for machines. Writers are used to measuring attention through pageviews and shares. Now, we may need to measure resonance through *retrievals*. The audience still exists, it’s just reading through an interpreter. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/11/Cloudflare-LLM-paths.jpg) Cloudflare report on paths taken into my blog by ChatGPT. ## **What’s next** I’d like to pick up this conversation with the **Plausibles of this world**, privacy-minded analytics teams who already think beyond cookies. Could they help us measure AI-mediated attention without reintroducing surveillance. Could we imagine a new kind of open signal, one that lets site owners know when their work is being read *in* the model rather than *on* the web? Should we look towards tools like WordPress and Ghost? These are still early days, but I sense a new metric emerging: not sessions, not clicks but *citations in the synthetic layer*. > In our LinkedIn exchange, [Youp van der Graaf, freelance data & CX consultant](https://data.youpvandergraaf.com), noted that optimising a site now means working with two distinct streams of behaviour: human traffic and AI-driven traffic. He also stressed that organisations need a way to bring the most relevant insights from both streams into their decision-making. ## **Closing reflection** The web hasn’t disappeared; it’s been internalised. Our words travel further than we can see, carried by systems that read on our behalf. For now, Cloudflare’s dashboard is just a glimmer on the radar, it is faint, but reassuring. At least we know we’re still in the sky. If you’re exploring the same terrain, or if you work on analytics tools like Plausible, I’d love [to compare notes](https://rob-hoeijmakers.ghost.io/about/). We may still be flying through cloud, but the outlines of a new map are beginning to appear. ### Further reading - [When Bots Become Readers: Publishing in the Age of AI Crawlers](https://rob-hoeijmakers.ghost.io/when-bots-become-readers/) - [The End of Google Search (as we know it)](https://rob-hoeijmakers.ghost.io/the-end-of-google-search-as-we-know-it/) - [Web analytics made simple by using Plausible](https://rob-hoeijmakers.ghost.io/cookie-free-analytics/) ### When Rules Learn to Think: Discovering Google Workspace Flows URL: https://rob-hoeijmakers.ghost.io/google-workspace-flows-agentspace/ Last updated: 2026-05-12T07:34:45.000Z I wasn’t looking for a new AI feature. It simply appeared, a small **Flows** button in my Gmail sidebar. Within minutes I was building automations that not only moved messages or added labels, but actually *read and understood* them. That first experiment turned into a small revelation. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/11/Google-Workspace-flows-Gmail.jpg) ## Quick takeaways - **Workspace Flows** turns classic Gmail filters and scripts into intelligent, no-code workflows. - These flows can use **Gemini**, Google’s AI model, to summarise, analyse, or classify messages and documents. - The tool sits inside a growing family of Google initiatives, **Gems**, **Gemini for Workspace**, and the enterprise-scale **Agentspace**, each representing a different layer of “agentic” automation. - The result: automation becomes conversational. You describe what you want, the workspace builds the logic. ## From filters to flows What once required scripting or third-party tools now lives natively in Workspace. You start with a trigger — *“when a message arrives from…”* — and add actions: move it, reply, store an attachment, post to Chat. The difference is that Gemini can now *interpret* what the message says. It can summarise, detect tone, or decide which folder fits best. In other words: the logic becomes contextual. ## Rules with intelligence The language of automation shifts from *if-then* to *why and what for*. I built a flow that reads new client emails, extracts the main question, and writes a short summary in a Google Doc. Another one classifies invoices and sends them to Finance. Each flow feels like a small colleague, still predictable, but capable of understanding nuance. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/11/Google-Workspace-Flows-.jpg) Example of rules in Gmail ## The naming puzzle Once I shared these findings, colleagues began asking: *“How is this different from Gemini, or those Gems and Agentspace things?”* Google’s naming hasn’t made it easy, so here’s the short map: | Term | What it is | Think of it as | | ------------------- | ------------------------------------------------------------------------------------------------------ | --------------------------------------- | | **Gemini** | Google’s core AI model, built into Workspace and Cloud. | The reasoning engine. | | **Gems** | Custom mini-agents you define (“summarise weekly sales”). | Personal skills or sub-agents. | | **Workspace Flows** | The automation builder that links triggers, actions, and AI steps inside Workspace. | Rules with intelligence. | | **Agentspace** | Enterprise platform (now within Gemini Enterprise) for orchestrating multiple agents and data sources. | The control room for complex workflows. | Seen together, they form a stack: **Gemini → Gems → Flows → Agentspace.** ## The quiet shift The implications are larger than they appear. Workspace Flows makes automation accessible to everyone who can express intent in language. It pushes the boundary between human logic and machine reasoning. It also invites new questions: Who governs these flows? What happens when AI decisions start shaping work patterns we no longer inspect? And, equally practical, how much time will we save when our inboxes truly begin to organise themselves? ## Closing I left the first flow running overnight. By morning my inbox had tidied itself, politely, almost invisibly. It’s easy to dismiss such convenience as another feature drop, but this one feels different. Rules have learned to think, and our workspaces are slowly learning to listen. --- [Google Workspace Flows - About Google Workspace FlowsGoogle Workspace Flows Google Workspace Flows was developed to automate tasks between Workspace apps. It connects directly with Gemini as your AI agent for research, analysis and content generation. Handle all sorts of tasks using starters, conditions, actions and functions.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/AAzXCkd2eIIOyXe4wLtjSETCfUKcYk58_eeFc1ns6z0pZeN0brZzZtUDkRP1mr2VT_l-AqrkE41fQdgSZIiebGMFf-1yV8zEXIBH2bGy_vx2R1S7xfvaDrmLpFadJLgo8mR8NAlpK_0w4ktRvEkJFCnyXONSQK1zRUwZKWZxj7H-KyPYHd5VVULRlCVCj6TvaIC5UnpnkhZEV9DrRH1KWB8)About Google Workspace Flows![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/AAzXCkdHszALUgBeMRo0Lhl8Mta6c-tSLURWeI0RxDT6NvWNsaoAsXd3bwnVRMQOStbh23IA8jZxFf4XMkyTYpUvzqU0OP4PhDF9BJEAodnkWTOcjtj6mLYFMVthWIFgZSgBdIec7jmMCxcu57cEVsEe068hsE5d1YxQyLPgW9EEf5SczQOjCRMOAfb3xz_Jrpp-F4uV9zW6zUbz3RkS2_8bJ8Gb6IoaGYkBfrr5-xk-w1280)](https://sites.google.com/view/workspace-flows/about) [Google One vs Google Workspace: Where do you get real AI and Gemini Advanced?What do you get when you pay for AI at Google? An exploration of Gemini Advanced, NotebookLM and how to share it smartly via Google One.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-334.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_6949-4.jpeg)](https://rob-hoeijmakers.ghost.io/google-one-vs-google-workspace-for-gemini/) [Google’s AI for Individuals and SMBs: Clarity, Costs, and Strategic ChoicesGoogle’s AI options are powerful but confusing. Here’s how I compared Google One vs Workspace—and why SMBs should think twice before locking in.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-335.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_3823-3.jpeg)](https://rob-hoeijmakers.ghost.io/googles-ai-for-individual-smb/) ### Compute: A New Measure of the World URL: https://rob-hoeijmakers.ghost.io/compute-a-new-measure-of-the-world/ Last updated: 2026-05-10T08:53:47.000Z *I was listening to Azeem Azhar speak about the rise of *compute*, the planetary build-out of data centres, GPUs, and artificial intelligence. His argument was that we are not simply investing in AI; we are industrialising computation itself. The numbers were staggering, but what caught me wasn’t the scale, it was the word *compute*. What does it actually represent?* ## Quick takeaways - “Compute” has become a new kind of utility, the electricity of thought. - We use it to extend logic, scale cognition, and simulate perception. - Yet the human experience of all this remains strangely familiar. - The deeper question is not whether we progress, but whether we can stop. ## Reflection A small video I once saw came to mind: a first-year computer-science lecturer telling his students that the course was *not* about computers. The term, he said, pre-dated the machine; a “computer” was once a person who computed. In that sense, compute is less a thing than an activity, a verb that escaped its subject. When we speak of **growing compute**, we are really speaking of externalising more and more of what the mind can do. Early computation was bureaucratic: it helped us **count, order, and store**. Today it **speaks, paints, and reasons** with us. The instrument has entered the register of experience. It feels inevitable, as if we are drawn along by the gradient of possibility itself. We compute because we can, and once we can, we must. The process satisfies both our instrumental needs, *efficiency, mastery, labour*, and our existential ones: *expression, curiosity, play*. The system feeds itself; the appetite grows with every new capacity. 💡 In recent years, compute has also become a noun. Tech circles now speak of “how much compute” a system or region possesses, and thus meaning the total available capacity to perform calculations. It compresses a vast idea into a single word: the sum of all processing power, energy, and infrastructure that makes digital work possible. The shift from verb to noun mirrors how computation itself has turned from an action into a resource, as something to be stored, traded, and scaled. ## Power and infrastructure Yet behind this mental expansion lies a struggle for power that is anything but abstract. Compute is not evenly distributed; it pools where money, materials, and policy converge. The race to build new data centres and chip factories is also a race to secure influence over the world’s **thinking infrastructure**. Each processor carries a fragment of sovereignty. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/11/Moore-s_Law_Transistor_Count_1970-2020.png) Transistor count over time. It gives an indication of the total rise of compute. That is why governments treat semiconductors as strategic assets, why export bans feel like acts of war, and why companies invest trillions in hardware that few of us ever see. Control over compute is control over **simulation, design, and decision**, over how the future is imagined and executed. Europe, with its rules and reflections, stands between dependence and autonomy, between ethical clarity and technological hesitation. ## Closing Perhaps that is why it feels both thrilling and weightless. We call it progress, but it may simply be the unfolding of what was always latent in our tools, the steady electrification of thought and feeling. Whether this is advancement or merely continuation is a question only hindsight can answer, and by then the next layer of compute will already be running. 💡 [Raj Reddy](https://www.quantamagazine.org/the-ai-pioneer-with-provocative-plans-for-humanity-20241204/) once suggested a simple test for what AI ought to be in human life. Not a destiny, not a replacement, but a posture. In his view, intelligence at scale earns its place when it acts as a cognitive amplifier that helps us think more clearly and do more with the time we have; and as a kind of guardian angel that notices risks and protects what we value before we can. The machines extend our reach and guard our blind spots, but they never cease to be in service of us. ### Further reading - [🔮 The demand for infinite compute](https://www.exponentialview.co/p/the-demand-for-infinite-compute) - [OpenAI Says It’ll Build Five More Huge Data Centers. Here’s Where](https://www.cnet.com/tech/services-and-software/openai-says-itll-build-five-more-huge-data-centers-heres-where/) - [From Silicon to Intelligence: Understanding the Hardware Behind AI](https://rob-hoeijmakers.ghost.io/cpu-gpu-tpu-npu-explained/) - [Infinite scale: The architecture behind the Azure AI superfactory - The Official Microsoft Blog](https://blogs.microsoft.com/blog/2025/11/12/infinite-scale-the-architecture-behind-the-azure-ai-superfactory/) - [The AI Pioneer With Provocative Plans for Humanity | Quanta Magazine](https://www.quantamagazine.org/the-ai-pioneer-with-provocative-plans-for-humanity-20241204/) ### Signing the Future: How a Digital Incorporation Made eIDAS Real URL: https://rob-hoeijmakers.ghost.io/digital-signature-eidas-experience/ Last updated: 2026-05-26T19:31:29.000Z When we transformed our partnership into a [private limited company](https://schmuki.nl), I expected paper. A meeting with the notary, wet ink, a stamp. Instead, it all happened in a browser: uploads, video checks, and digital clicks that somehow carried legal force. Only later did I realise I had just taken part in the European experiment with *digital trust*, [the eIDAS framework](https://rob-hoeijmakers.ghost.io/eu-eid-eidas-eudi-wallet/), and that I had signed something without ever touching a pen. ### Quick takeaways - Identification and signature are two distinct acts: first *who you are*, then *what you agree to*. - The signature’s power comes from its cryptographic trail, not its appearance. - Qualified Trust Service Providers (QTSPs) give digital signatures the same legal weight as handwritten ones. - Notaries, intermediaries, and state registries are now connected by secure digital chains. - eIDAS quietly turns European bureaucracy into code — without losing its concern for proof and personhood. ## 1\. A company born in the browser Our process began at [**Firm24**](https://www.firm24.com/), an intermediary platform for digital incorporation. The first step was a **Know Your Customer** check: we uploaded our identity documents and recorded a short **video in the browser**, confirming that the person on screen matched the ID. That was the automated front end of trust, identity established before any signing took place. Once the information was verified, the notary drafted the deeds. We reviewed and **signed them digitally** through [**SignRequest**](https://signrequest.com), the familiar on-screen signature that records time, device, and IP address. At first, I uploaded a photograph of my real, hand-drawn signature. It looked neat, recognisable, but the system declined it. What it wanted was the clunky version drawn with a mouse: ugly, but traceable. The platform trusted the *event*, not the *image*. 💡 ****Why the ugly signature won** The drawn signature wasn’t trusted because it looked authentic, but because it was **recorded* authentically. A photo of a handwritten signature is just an image; anyone can reuse it. The clumsy mouse-drawn version, however, is created **within* the signing system, tied to a timestamp, device, and identity trail. It’s ugly, but cryptographically honest. And that’s what makes it legally strong. Only after everything was ready for formalisation did we have a **live WhatsApp call with the notary’s office**. It wasn’t another ID check; the notary already had our identification from the Firm24 dossier. The conversation was brief, all it took was a human confirmation that we understood and agreed with the deed before it was officially signed. Shortly after, the notary added their **qualified digital signature** and sent the deed electronically to the **Chamber of Commerce**. A company was born, entirely in the browser. ## 2\. What was really happening It felt effortless, almost ordinary, but underneath, each step carried a specific legal and technical meaning: | Stage | What happened | eIDAS layer | | ----------------------------- | ------------------------------------------------------------------------------------------------ | -------------------------------------------- | | **Identity check** | Upload of passport and short verification video through Firm24 | *Electronic identification* | | **Data transfer** | Verified dossier sent from Firm24 to the notary | *Trusted delivery* | | **Digital signing of drafts** | We signed deeds through SignRequest | *Advanced electronic signature (AES)* | | **Intent confirmation** | Live WhatsApp call with the notary’s office — confirming understanding and consent, not identity | *Authorisation / will formation* | | **Notary’s formal seal** | Notary applied qualified digital signature through a Certified Trust Service Provider | *Qualified electronic signature (QES)* | | **Registration** | Deed filed electronically with the Chamber of Commerce | *Trusted electronic delivery / legal effect* | The **Firm24 stage** handled identification, proving who we were. The **SignRequest stage** captured our consent digitally. The **WhatsApp call** reintroduced a brief human moment: not a check of documents, but a confirmation that we truly understood and agreed to the act of incorporation. Finally, the **notary’s qualified signature** gave the deed its formal, cross-border legal force under eIDAS. ## 3\. Trust as infrastructure What struck me is how *invisible* trust has become. We used to rely on rituals: paper, witnesses, stamps. In the digital process, trust is distributed across systems. - **Firm24** verified who we were - a human front door. - **The notary** confirmed that we understood and agreed - legal assurance. - **The qualified trust provider** sealed the deed - cryptographic proof. - **The Chamber of Commerce** accepted it - institutional trust. Once that final registration was complete, everything else fell neatly into place. Our **bank** immediately recognised the new company and transferred the existing account without complication — no new signatures, no paper. Setting up **bookkeeping in MoneyBird** was equally seamless, with the company data retrieved directly from the trade register. Even the **Tax Office** followed automatically, issuing a new VAT number within days. It was a glimpse of what a connected trust infrastructure can feel like: every authority drawing from the same verified record, without friction or duplication. Each part replaces a physical step with a digital one, but the underlying logic remains: you prove your identity, express intent, and an authorised party validates it. The difference is that these steps now happen within seconds, across secure servers rather than across desks. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/11/IMG_0569-2.jpeg) The Chamber of Commerce maybe went overboard a bit. ## 4\. From mark to mechanism At some point I realised that the “signature” had ceased to be a mark altogether. It was now an *event*: a timestamped record linking identity, intent, and document integrity. Authenticity no longer comes from handwriting analysis but from cryptographic linkage. It’s the **system**, not the symbol, that carries trust. That shift has deep implications. It means that *authority* can now travel across borders. A Dutch qualified signature is valid in Finland, France, or Portugal. It also means that European digital identity is no longer a convenience but it’s an essential building block of law, commerce, and citizenship. ## 5\. Why this matters Europe’s approach differs from the more laissez-faire systems elsewhere. eIDAS doesn’t treat signatures as mere digital consent. It roots them in **sovereign identity** and **regulated trust services**, a deliberate attempt to build autonomy at the infrastructure level. When we speak of *digital sovereignty*, this is what it looks like in practice: individuals and institutions bound by verifiable, portable proofs of identity. In a way, the process revealed a cultural shift as much as a technical one. The European state is learning to speak digital, and the notary is its interpreter. ## Closing reflection The most consequential signature of my professional life has no visual form. It lives as metadata and mathematics: a timestamp, a certificate chain, a line in a European trust list. Yet it feels reliable, even human, because every layer still asks the same questions that law has always asked: *Who are you? What do you intend? Who witnesses it?* Bureaucracy hasn’t vanished; it has changed medium. The ritual of signing now lives in code and that code, in Europe at least, still begins with identity. ## Further reading - [EU eIDAS Regulation – official portal](https://digital-strategy.ec.europa.eu/en/policies/eidas-regulation) - [EU Trusted List Browser](https://esignature.ec.europa.eu/efda/tl-browser/#/) - [Dutch government overview (Rijksoverheid)](https://www.rijksoverheid.nl/onderwerpen/identiteit-en-privacy/eidas) --- ### Further reading - [One European Company (EU-INC)](https://rob-hoeijmakers.ghost.io/one-european-company-eu-inc/) - [Understanding Digital Identity in the EU: eID, eIDAS, and the EUDI Wallet](https://rob-hoeijmakers.ghost.io/eu-eid-eidas-eudi-wallet/) ### Reskilling the Mind: Europe’s Next Transition URL: https://rob-hoeijmakers.ghost.io/reskilling-the-mind-europes-next-transition/ Last updated: 2026-05-10T08:53:47.000Z ## The premise For half a century, Europe’s transformation has been *cognitive*. We traded the factory floor for the meeting room, the lathe for the laptop. Productivity became language-driven: managing, advising, interpreting. Now that foundation is shaking. Artificial intelligence begins to do the very things that once made our economies “advanced.” The question is no longer whether we must reskill, but whether we can redirect our collective intelligence, *and our industrial capacity*, before the gap widens. ## Looking backward Every great transition has demanded a new kind of worker. - **The industrial turn** pulled rural Europe into foundries and rail yards. - **The post-war reconstruction** built technical and bureaucratic skill at national scale. - **The digital turn** globalised services and thinned the material base of production. What we face now is unusual: the *thinking layer* of work is under pressure, just as the geopolitical layer thickens again. ## The new geography of work Reskilling is not only cognitive, it is territorial. The return of power politics, supply-chain insecurity, and military re-armament forces Europe to reconsider what it means to be productive. Defence, energy, and advanced manufacturing will require engineers, machinists, logistics planners, and technicians, people who can *build* again, not only advise or analyse. In that sense, reskilling may also mean **re-industrialising**: reconnecting human skill to material capability. Europe’s long outsourcing of production has left it intellectually rich but strategically thin. Rebuilding competence in physical domains is as urgent as training people for digital fluency. ## The civic dimension Europe’s instinct is often to regulate what it fears. But legislation cannot replace lost skill. What we need is a continental project of learning and making, education as industrial policy, apprenticeship as security strategy. Reskilling, in this view, is not a reaction to automation but an act of self-definition. It is how a civilisation decides what it still wants to *know how to do*. ## Who will organise it? In previous eras, reskilling was almost automatic because authority and direction were clear. Industrialisation had the factory. Reconstruction had the state. Digitisation had the corporation. Today, none of these actors alone can carry the weight. Governments can set incentives but move slowly and fear backlash. Corporations train for narrow advantage, not for civic resilience. Educational systems still treat knowledge as a stock to be transmitted, not as a capacity to adapt. That leaves a vacuum, and perhaps a choice. Either reskilling becomes a *managed transition*, coordinated across states, sectors, and citizens, or it unfolds through *disruption*: job collapse, social anger, and reactionary politics. Europe has historically preferred negotiated change over violent rupture. Yet negotiation requires shared purpose, precisely what decades of outsourcing and fragmentation have eroded. The challenge is not only to train people, but to rebuild trust between the institutions that would have to act together. [💡 The $100 trillion productivity puzzleCapabilities are up, costs are down, productivity is flat—mapping the capability-absorption gap and the moves to close it.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/https-3A-2F-2Fbucketeer-e05bbc84-baa3-437e-9518-adb32be77984.s3.amazonaws.com-2Fpublic-2Fimages-2Faa7453ed-eb02-46d7-b523-74d86b1c1e24-2Fapple-touch-icon-180x180-9.png)Exponential ViewAzeem Azhar![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/https-3A-2F-2Fsubstack-post-media.s3.amazonaws.com-2Fpublic-2Fimages-2F3c207541-5deb-481e-932b-1223cd792255_1600x894-2.png)](https://www.exponentialview.co/p/ai-is-ready-is-your-company) ## Reskilling as sovereignty If knowledge work is no longer Europe’s safe harbour, what becomes the new basis of dignity and contribution? Perhaps the next social contract lies somewhere between code and craft, between intelligence and manufacture, a re-balancing of mind and matter. Reskilling, then, is not only about adapting to AI. It is about re-establishing sovereignty: cognitive, industrial, and human. ### Further reading - [Future of Work - Rob Hoeijmakers](https://rob-hoeijmakers.ghost.io/tag/future-of-work/) - [AI literacy: from definition to practice](https://rob-hoeijmakers.ghost.io/ai-literacy-from-definition-to-practice/) - [The Paradox of Sovereignty: Europe’s Search for Freedom in a Connected World](https://rob-hoeijmakers.ghost.io/sovereignty-paradox/) ### Excel, the Hidden Operating System of Business Reasoning URL: https://rob-hoeijmakers.ghost.io/excel-hidden-operating-system-of-business-reasoning/ Last updated: 2026-05-10T08:53:48.000Z ## *How Excel quietly shaped the way we think, and how AI may extend its legacy.* ### Quick takeaways - Excel has functioned for decades as a *distributed ERP*: every user building a local model of reality. - Its power lies in flexibility, it lets people express business logic without programming. - Enterprise systems formalise these models, but often lose their subtlety and adaptability. - AI and conversational interfaces might restore that flexibility, learning *from* humans rather than forcing humans to adapt. - The future of IT may not replace Excel’s logic, but translate it into new, more fluid forms. ## The silent backbone of business For most of the digital era, companies have lived on spreadsheets. Excel has been the quiet workhorse behind budgets, forecasts, reconciliations, audits and is the connective tissue of economic life. It isn’t just a table tool; it’s a universal modelling language. Every tab in Excel contains traces of business reasoning: rules, workflows, and the instincts of whoever built them. People use formulas to make sense of uncertainty, test assumptions, or translate their knowledge into numbers. In that sense, Excel has been the *de facto operating system of business reasoning*, a distributed, ever-evolving network of micro-ERPs shaped by human context. > “In some sense, Excel will come with an analyst bundled in and with all the tools used.” > Satya Nadella, Microsoft CEO ## The paradox of structure Once these spreadsheets succeed, they’re handed to IT. They are translated into databases, interfaces, and fixed processes. The result is reliable, auditable and rigid. The nuance that made the original model valuable often disappears in the act of formalisation. Excel’s gift is its informality. It allows domain experts to model reality before the system ossifies it. That same informality, however, is what makes it fragile at scale. The tension between adaptability and control sits at the heart of almost every IT transformation. 💡 Spreadsheets began as a canvas for numbers. They have become the place where organisations encode how they think. As artificial intelligence matures, Excel does not vanish. It becomes the cognitive substrate where millions of decisions are made, explained, and verified. ## When systems start to learn back The rise of AI marks a reversal. Instead of humans learning the logic of systems, systems are beginning to learn from human logic. Large language models can take unstructured expression, *notes, emails, verbal explanations*, and infer structure, context, and intent. This suggests a new kind of interface: one that captures reasoning in natural terms, converts it into structured rules, and interacts with formal systems below. It’s not about replacing Excel, but extending its lineage and giving the same expressive freedom to conversation that spreadsheets gave to formulas. Where Excel let people *compute with intuition*, AI lets them *reason with data* in their own language. Both reduce friction between thinking and doing. [Claude for ExcelClaude understands your entire workbook—from nested formulas to multiple tab dependencies.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/689f4a9aff1f63fde75cf733_favicon.png)![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/68fa8015a6f02a7dd082841e_og-excel.jpg)](https://www.claude.com/claude-for-excel) ## Closing thought From cell to prompt, from formula to phrase. The underlying desire hasn’t changed: to make our reasoning legible to machines without needing to change its human texture. The difference is that, for the first time, the machine seems willing to listen. ### Further reading - [Excel and the future cockpit of business logic](https://rob-hoeijmakers.ghost.io/excel-and-the-future-cockpit-of-business-logic/) - [Satya Nadella — How Microsoft is preparing for AGI](https://www.dwarkesh.com/p/satya-nadella-2) ### The Real Cost of Leaving Microsoft 365 URL: https://rob-hoeijmakers.ghost.io/the-real-cost-of-leaving-microsoft-365/ Last updated: 2026-05-10T08:53:48.000Z When the International Criminal Court (ICC) switched from Microsoft 365 to the German open-source suite [**OpenDesk**](https://www.opendesk.eu/en), it barely made mainstream headlines. But in tech governance circles, it caused ripples, and not because of the software itself, but because of what it represents: a rare act of *digital self-determination*. Most organisations use Microsoft 365 or Google Workspace as if they were natural utilities, as invisible as water or air. The systems feel eternal, part of the landscape of work. So when someone decides to leave that landscape, it’s worth asking: *why*, and *at what cost*? Because changing your office suite isn’t just a technical project. It’s a human, financial, and cultural event. [International Criminal Court ditches Microsoft Office ‘to remain autonomous’The International Criminal Court (ICC) is to replace Microsoft software with the openDesk software offered by German company Zendis. The contracts are about to be signed…![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/cropped-site-icon.png)Brussels SignalChris Gattringer![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/cropped-WhatsApp-Image-2025-10-31-at-11.17.14-3.jpeg)](https://brusselssignal.eu/2025/10/icc-ditches-microsoft-office-in-favour-of-german-state-owned-software/) ## The Drivers for Change Few organisations leave Microsoft 365 voluntarily. The usual drivers for change are defensive rather than visionary, often born from risk, regulation, or sovereignty concerns. **Digital sovereignty and independence** For public institutions like the ICC, the motivation is clear: data hosted on American infrastructure falls under U.S. jurisdiction, even when stored in Europe. A switch to OpenDesk (built on Nextcloud, LibreOffice, and Matrix) means hosting under European law, on open standards, without the extraterritorial pull of the [U.S. CLOUD Act](https://english.ncsc.nl/latest/weblog/weblog/2022/how-the-cloud-act-works-in-data-storage-in-europe). **Compliance and public trust** Government bodies and NGOs face increasing scrutiny over where and how they store data. For them, transparency is a form of legitimacy. Running open software isn’t just an IT choice; it’s an ethical stance. **Cost management** For some, it starts as a financial question. Subscription costs scale linearly (per seat, per month), and Microsoft’s price tiers keep rising. But pure cost savings are rarely the real motive. The deeper wish is to regain control over what’s being paid for, and why. **Strategic alignment** Some European institutions adopt open systems for strategic reasons: aligning with EU digital sovereignty goals, supporting local providers, or avoiding vendor capture. Still, it’s worth noting: these are all *protective* motives. The story begins with “we can’t stay,” not with “we want to leave.” > The hard part is that sovereignty is not only about ownership but about reliability over decades. If domestic providers can be acquired at any moment, then legal jurisdiction alone does not guarantee autonomy. We need European capabilities that are both strategically independent and operationally excellent. Otherwise the choice becomes a trade-off: local in name, global in practice. Rob Hoeijmakers. ## The Human Dimension: Skill Lock-In Once you look beyond servers and licenses, the real lock-in reveals itself: people. Employees have grown up with Word, Excel, and Outlook. They know where buttons live, how to share files, how to comment, how to search. That fluency is invisible until you take it away. Organisations carry an enormous amount of *embedded skill capital*. It’s not written down anywhere. It lives in muscle memory, in the way teams collaborate, and in the shared vocabulary of daily work: > “I’ll send you the Excel.” > “Let’s track it in a table.” > “Check the shared drive.” Replace the platform, and those simple sentences stop working. You’ve changed not just tools, but the *language* of work. That’s why migrations of this kind are so disruptive: they rewrite social habits, not just configurations. Suddenly, people need training for tasks they could once do half-asleep. Confidence dips. Productivity slows. The new tools might be fine, but the *feel* of work is gone, and that takes months to rebuild. In that sense, skill lock-in is a form of **cultural dependence**. When people know the shortcuts, the file paths, the logic, they move faster, trust more, and make fewer mistakes. Breaking that fluency is costly in morale as well as time. ## The Technical and UX Challenges Even for IT teams that support the change, the early phase can feel like swimming upstream. **Integration and authentication** Microsoft 365’s magic lies in [Entra ID](https://www.microsoft.com/en-us/security/business/identity-access/microsoft-entra-id) (formerly Azure AD) and its invisible single sign-on. Change the stack, and you lose that frictionless identity layer. [Nextcloud](https://nextcloud.com) and [Keycloak](https://www.keycloak.org) can replace it, but it takes expertise and usually, a few rough months. **Compatibility and context** Word documents open fine in [LibreOffice](https://www.libreoffice.org), until they don’t. Macros, comments, and formatting quirks appear. Shared links, calendars, and version histories must be rebuilt. **Search and sharing** In Microsoft 365, you can search across email, files, and chats in one stroke. Open alternatives require stitching together multiple systems. You get transparency, but lose cohesion. **Support and maintenance** In a subscription world, updates are automatic. In self-hosted or federated setups, you own the maintenance. That means patching, monitoring, backups, and a certain level of internal IT craftsmanship. In short: OpenDesk or similar alternatives *work*, but they shift responsibility. You trade convenience for control. ☁️ ****Note: From desktop to cloud** Office software once lived on local machines. Organisations bought licenses, installed Word and Excel, and saved files on shared drives or USB sticks. Over time, that model vanished and so updates moved online, storage followed, and the “office suite” became a cloud service. Today, tools like Microsoft 365, Google Workspace, and ****OpenDesk** all run in the cloud. The difference is **where* and **under whose jurisdiction* that cloud operates. OpenDesk is hosted on ****European infrastructure such as** [****StackIT**, the cloud spin-off of Schwarz IT](https://rob-hoeijmakers.ghost.io/lidl-stackit-european-cloud/). It’s part of a broader movement toward European-hosted, open-source productivity platforms. So not less cloud, but a more [**sovereign*](https://rob-hoeijmakers.ghost.io/sovereignty-paradox/) one. ## The Financial Side Cost is the paradox of modern software ecosystems. Microsoft 365 feels expensive per user, but in reality it outsources enormous complexity. Email, storage, video conferencing, authentication, device management and that all as a predictable monthly service. Switch to an open-source stack, and you’ll likely reduce licensing fees. But internal costs rise in other ways: - **Training and change management**: people need to learn, experiment, and adapt. - **Support and infrastructure**: servers, monitoring, and security now fall to your team or local partner. - **Migration and compatibility**: converting documents, remapping permissions, re-creating workflows. The total cost of ownership may balance out only after several years, and only if the organisation fully commits to open principles rather than half-rebuilding what Microsoft once bundled. For institutions like the ICC, that equation looks different. Their goal isn’t savings, but *sovereignty*. For private companies, the calculus is usually stricter: short-term disruption versus long-term independence. ## What’s Really at Stake At its heart, this isn’t about software. It’s about **digital self-awareness**, understanding what you’re dependent on, and why. Every modern organisation sits on an invisible foundation of tools, habits, and assumptions. We rarely question it because it works. But the moment you try to leave, you realise how deep the roots go. The ICC’s move to OpenDesk is a symbolic reminder that infrastructure choices are *political, cultural, and human* decisions. They define who controls your data, but also how your people work and how they feel while working. ## Closing Reflection Choosing software is easy. Realising what you’ve chosen is harder. For most organisations, staying with Microsoft 365 or Google Workspace isn’t wrong, it’s practical. The key is to stay *conscious*: to know what you’re buying, what you’re giving up, and where your dependencies lie. If your organisation ever considers leaving, start by asking: - Where does our knowledge really live — in data or in people’s habits? - What would it take to rebuild that fluency elsewhere? - And is our motive fear, or conviction? Digital sovereignty doesn’t begin with servers or source code. It begins with awareness, the quiet recognition that infrastructure is never neutral. Because once you know where your dependencies live, *you already hold part of the independence you’re seeking.* ### Further reading - [The office and collaboration suite for public administration](https://www.opendesk.eu/en) - [Microsoft didn’t cut services to International Criminal Court, its president says](https://www.politico.eu/article/microsoft-did-not-cut-services-international-criminal-court-president-american-sanctions-trump-tech-icc-amazon-google/) - [The Paradox of Sovereignty: Europe’s Search for Freedom in a Connected World](https://rob-hoeijmakers.ghost.io/sovereignty-paradox/) - [Strengths & weaknesses of Microsoft 365 Alternatives](https://www.ezone.co.uk/blog/microsoft-365-alternatives-rated-according-to-strengths-weaknesses.html) ### The Paradox of Sovereignty: Europe’s Search for Freedom in a Connected World URL: https://rob-hoeijmakers.ghost.io/sovereignty-paradox/ Last updated: 2026-05-10T08:53:48.000Z *Learning note: not an expert take, but a student’s reflection* Lately I keep hearing the word *sovereignty*. In Brussels documents, in news about “[Eurostack](https://rob-hoeijmakers.ghost.io/the-eurostack-explained/),” even in conversations with clients. It sounds strong, like a promise of independence, safety, and control. But when I look closer, it’s less clear what the word actually means. Sovereign from what? From whom? And at what cost? When people talk about “European sovereignty,” they rarely mean what a traditional state would. It’s not about isolation or self-sufficiency. It’s about **freedom to act**, the ability to decide your own dependencies rather than have them decided for you. That sounds sensible. But as I’m discovering, it’s also a paradox. ## Quick takeaways - EU “sovereignty” ≠ self-sufficiency. It means **strategic autonomy**, or freedom to choose. - Hyperscalers like AWS, Microsoft, and Google are **partners and risks** at the same time. - The EU’s sovereignty framework often **rewards efficiency over independence**. - True sovereignty is expensive; *trusted interdependence* may be the more realistic goal. ## Reflection: unpacking the paradox When I first tried to understand this topic, I imagined sovereignty as a kind of fenced garden. You own it, you decide who enters. But the more I read, the more I realised: **Europe’s digital garden shares its plumbing with the rest of the world.** The servers are often American, the chips Taiwanese, the code open-source, the data cables global. Trying to be “fully sovereign” in that context feels like insisting you can live off the grid in a city apartment. So what does the EU really mean? In Brussels jargon, sovereignty has quietly evolved into *“open strategic autonomy.”* The idea isn’t to do everything alone but to **retain the freedom to act**, even inside complex systems of interdependence. That makes sense politically. But economically, it’s tricky. Because sovereignty has a price. [From Sand to Software: A Whistle-Stop Tour of the AI Value ChainAI may look like pure software, but it rests on a fragile chain of quartz, optics, fabs, and GPUs. This post traces the journey, stop by stop.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-317.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_8137-1.jpeg)](https://rob-hoeijmakers.ghost.io/ai-value-chain-from-quartz-to-cloud/) ### When sovereignty meets economics If you take sovereignty too literally, you end up like North Korea: self-reliant, but poor and cut off. Economies thrive on scale, trade, and shared infrastructure, exactly the things that limit sovereignty. This tension runs straight through the EU’s “[Cloud Sovereignty Framework.](https://commission.europa.eu/document/09579818-64a6-4dd5-9577-446ab6219113%5Fen)” The system gives each provider a “sovereignty score” based on eight criteria: documentation, resilience, supply-chain control, and so on. But here’s the twist: **legal control and jurisdiction count for only 10 percent** of the total. So the best-performing, most “sovereign” platforms, *on paper,* are often the same U.S. hyperscalers the EU worries about. That’s the paradox in spreadsheet form. [Cloud Sovereignty Framework | European Commission![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon-11.svg)European Commission![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/ec-socialmedia-fallback-2.png)](https://commission.europa.eu/document/09579818-64a6-4dd5-9577-446ab6219113%5Fen) ### The hyperscaler dilemma Europe depends on these U.S. giants for cost-effective cloud services, yet fears their legal reach. Under American law, U.S. companies can be forced to hand over data stored anywhere in the world. So Brussels doesn’t want to ban them, it wants to **domesticate** them: through data boundaries, local trustees, and compliance layers. The tone is cooperative but cautious: > “We need their technology, but we can’t be hostage to it.” It’s a relationship that feels less like divorce and more like couples therapy. > This is beyond parody: the EU has come up with a "cloud sovereignty framework" to make EU tech more sovereign and they've notably devised a "sovereignty score" to determine whether a particular cloud provider is sovereign or not (which will then determine if they get government… > > — Arnaud Bertrand (@RnaudBertrand) [October 30, 2025](https://twitter.com/RnaudBertrand/status/1983688089148256547?ref%5Fsrc=twsrc%5Etfw) ### Learning as I go The more I read, the more I realise that sovereignty isn’t a technical state but it’s a **state of mind**. It’s about awareness of where your dependencies lie, and willingness to act if one of them becomes risky. That’s true for Europe, but also for companies and even individuals building on foreign digital foundations. Maybe the goal isn’t pure autonomy, but **resilient interdependence**, a Europe that knows its limits, diversifies its risks, and keeps the agency to choose. ## Closing: sovereignty without fences If sovereignty is about freedom, then the question is not how high we can build our fences, but how well we can manage the shared systems that keep us alive. Europe’s project, as I now see it, isn’t to stand alone, but to stay **movable**, to remain a player with options in a world of entanglements. That, to me, feels like a more grown-up kind of sovereignty: not ownership of everything, but ownership of your choices. --- ### Further reading - [Le référentiel cloud souverain de l’UE suscite le débat - Le Monde Informatique](https://www.lemondeinformatique.fr/actualites/lire-le-referentiel-cloud-souverain-de-l-ue-suscite-le-debat-98322.html) - [Sovereignty Without Soil](https://rob-hoeijmakers.ghost.io/sovereignty-without-soil/) - [From Shopping Cart to Server Rack: How Lidl Builds a European Cloud](https://rob-hoeijmakers.ghost.io/lidl-stackit-european-cloud/) ### Between Idealism and Reach: Trying Out the Fediverse URL: https://rob-hoeijmakers.ghost.io/activitypub-fediverse-ghost/ Last updated: 2025-10-30T16:13:26.000Z I recently enabled *ActivityPub* on my [Ghost](https://rob-hoeijmakers.ghost.io/ghost-blog/) site. In theory, this turns my blog into a social account that can be followed from Mastodon, and through a bridge, even from Bluesky. It’s part of what’s called the *Fediverse*: a growing web of platforms that exchange posts and reactions without belonging to a single company. It sounded like a quiet revolution: the open web becoming social again. ## What it is ActivityPub is a protocol, not a platform. It describes how websites can publish and subscribe to each other’s posts, much like email or RSS once did. The collection of services using it, Mastodon, Pixelfed, PeerTube, WriteFreely, and now Ghost, form the Fediverse. Each server is independent, yet connected through shared standards. The idea carries a certain idealism: ownership, decentralisation, resilience. A social web that doesn’t depend on a handful of companies. ## What it looks like in Ghost In practical terms, my site now has a handle: [@rob@hoeijmakers.net](https://mstdn.social/@rob@hoeijmakers.net). People on [Mastodon](https://mstdn.social/@rob@hoeijmakers.net) can follow it. When I post, it appears in their feed; replies return to a small inbox in Ghost. A bridge forwards the same posts to [Bluesky](https://bsky.app/profile/rob.hoeijmakers.net.ap.brid.gy). From the outside, it feels like a modest reappearance of something older: RSS, but social. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/10/image-4.png) My publication seen on Mastodon, on my personal account. Not sure anymore how these relate. ## What it feels like After a few days of use, it works, but quietly. Posts appear; there’s little sense of reach or movement. Threads doesn’t support it yet, and mainstream networks like X or LinkedIn live on other protocols entirely. The more interesting part is the reading side: Ghost’s new social reader lets me follow others, which feels a bit like Substack’s built-in community — not big, but intentional. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/10/image-5.png) The reader in [Ghost](https://ghost.org). Also notification end up here. ## Between idealism and pragmatism It depends what you hope for. If your goal is to build a resilient, independent network in a time of growing platform control, ActivityPub is a principled step. If your goal is to reach readers at scale, it’s not there yet; automation, analytics and network effects still belong to the centralised world. I approached it as a publisher, curious about distribution. But what I found instead was a small reminder of the web’s original ethos: publishing as a shared act, not a transaction. [Why the open social web matters nowThe needs are real – and you have so much power.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/thumb.webp)Ben WerdmullerBen Werdmuller![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/photo-1475359524104-d101d02a042b)](https://werd.io/why-the-open-social-web-matters-now/) ## My personal take So I keep it on. It costs little, and perhaps that quiet idealism is part of its charm. The Fediverse may never rival the scale of today’s platforms, but it keeps alive the idea that the web itself can be social without permission. I would be really happy to see my articles go to [Threads](https://www.threads.com/@robhoeij) as well, but it is not clear to me how and when that will work. 🇪🇺 I found out why Threads is not working with the Fediverse. For me at least. Meta writes: **Last March, we made sharing to the fediverse possible for people 18+ with public Threads profiles in the US, Japan and Canada. Then in June, we expanded globally,* ***excluding the European Region** *.* **Since then, we’ve made it possible for you to see the people from other fediverse servers who like your posts and follow you and enabled you to follow people from other fediverse servers.* ### Elections and AI URL: https://rob-hoeijmakers.ghost.io/elections-and-ai/ Last updated: 2026-05-10T08:53:48.000Z Recent reports in the Netherlands have questioned whether AI systems distort political information. One study suggested that Google’s AI answers favoured certain parties. Another warned voters not to rely on chatbots for advice. The debate touches something larger. About how societies learn to live with new technologies that mediate our understanding of the world. ## Takeaways - AI now retrieves, filters, and composes information in real time. - It is not a *medium* for opinions but a *tool* for exploration. - Democracies depend on citizens who take responsibility for how they use information. - What we are witnessing is not entirely new: the web, social media, and voting tools changed us before. ## The Netherlands as example The Dutch discussion around AI and elections shows how quickly public trust becomes unsettled. Researchers found that some political parties appeared more often in Google’s AI-generated answers. A government watchdog warned that chatbots are “biased” and should not guide voters. These reactions are understandable, new technologies often evoke anxiety, yet they overlook how search, ranking, and recommendation have always shaped what we see. Today’s AI systems extend that pattern. They don’t simply mirror language; they *act* within it. They select, summarise, and frame information dynamically, blending reasoning, retrieval, and presentation. That does not make them agents with intent, but instruments with reach and it calls for literacy rather than fear. 🔴 ChatGPT and similar tools are explicitly instructed not to give direct voting advice, yet in practice those boundaries turn out to be far from watertight. In one case, a prompt using the phrase **‘geen poespas’* led the model to suggest the SP (Socialist Party), a linguistic association rather than a political judgment, showing how wordplay and tone can slip through the filters. ## Learning to work with the tool In preparing for the elections, I used ChatGPT as one of several instruments. I went through a few voting questionnaires and had longer conversations, with the model, with friends, and with family. These exchanges helped clarify what I actually value. I compared parties from two perspectives: a realistic, partly self-interested one, and a more principled view of the world I’d like to live in. I don’t follow the daily news closely, and the personalities or behaviour of party leaders rarely determine my vote. They can be a reason to hesitate, but not the reason to act. For me, democracy means accepting that outcomes will always involve compromise and engaging in that process with both realism and principle. This reflection is not meant as a polemic. I may be wrong in my impressions. I understand AI technically, but I am not an expert in society, democracy, or the media. I write simply from experience: as someone learning, like many others, how to think and decide amid systems that reshape how knowledge is formed. 💡 Today’s AI models excel at imitation, not understanding. OpenAI’s [Andrej Karpathy argues for a next wave built around a compact “cognitive core”](https://www.youtube.com/watch?v=lXUZvyajciY) that reasons and retrieves facts when needed, rather than memorising everything. Philosopher Toby Ord notes that [current “reasoning models” already face steep computational limits](https://www.tobyord.com/writing/how-well-does-rl-scale), progress costs ever more energy and time. Both point toward a shift: smaller, modular, tool-using systems that think before they speak. By the next elections, we may well work with these new, leaner forms of AI. [Simpler, smarter AI](https://www.exponentialview.co/p/ev-547) ## Medium or tool? AI becomes dangerous when treated as a *medium*, a space where opinions circulate and authority is implied. Used as a *tool*, it sharpens research, prompts curiosity, and widens access to information. Democracy has always required effort: to compare, to verify, to reason. What changes is that the instruments of reasoning are now partly algorithmic. It will take time for individuals, politicians, educators, and media to adjust. But this is not an alien revolution, it is the continuation of a long experiment in how technology and thought intertwine. ## Closing thoughts Modern AI is not our conscience and not our enemy. It is a tool, and like all tools, it can deceive if used passively. Don’t use it as a medium that tells you what to think. Use it as an instrument that helps you think more clearly about yourself, your society, and your vote. --- ### Further reading - [Don’t use AI to tell you how to vote in election, says Dutch watchdog](https://www.theguardian.com/world/2025/oct/21/ai-chatbots-unreliable-biased-advice-voters-dutch-watchdog) - [Google zet gekleurde AI-antwoorden midden in campagnetijd aan, uit, en weer aan](https://nos.nl/nieuwsuur/artikel/2587804-google-zet-gekleurde-ai-antwoorden-midden-in-campagnetijd-aan-uit-en-weer-aan) - [Google AI overviews during the 2025 Dutch elections • Deep Culture](https://deep-culture.org/google-ai-overviews-during-the-2025-dutch-elections/) ### Reading Breakneck by Dan Wang URL: https://rob-hoeijmakers.ghost.io/dan-wang-breakneck/ Last updated: 2026-05-10T08:53:49.000Z [Dan Wang’s](https://danwang.co/breakneck/) *Breakneck* is the rare book that fuses lived experience, economic history, and cultural observation into one continuous argument. It is about China’s rise, but not only; it is also about America’s distance from what it once knew how to do. Reading this compelling book, one feels the pull of two gravitational centres, China’s factories and America’s myths, orbiting the same system. The book doesn’t moralise. It simply shows how the world’s material and mental geographies have drifted apart, and how that divide explains much of what we are now living through: tariffs, supply chains, the quiet rearrangement of global confidence. [Breakneck: China’s Quest to Engineer the Future | Dan WangAnnouncing my book; whalelore; Austrian Catholicism; the legacy of Moses and Rickover; Yunnan; influencer culture; how historians work.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/cropped-dw.png)Dan WangDan Wang![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/Breakneck-cover-1-700x1050.jpg)](https://danwang.co/breakneck/) ## Quick takeaways - *Breakneck* documents China’s transformation through the eyes of an observer who built his own life between China and the US. - The book’s power lies in its blend of reportage and reflection: macroeconomics seen from the street level of factories and offices. - It reveals the *continuity* between China and the US, less rivals than intertwined halves of one industrial civilisation. - It suggests that the world’s current fractures are symptoms of a deeper split between the *material* and the *mental*. - For Europe, the book is not a mirror but a question: where do we stand between these two ways of being modern? ## Reflection ### Making What makes Wang convincing is not ideology but observation. He writes about engineers, toolmakers, and production lines with the familiarity of someone who has stood among them. Through these details, the abstract word *growth* regains texture. China’s development is not a theory; it is the rhythm of welders, logistics hubs, and 996 office routines. > “What do engineers like to do? Build. Since ancient times, the emperors have tried to tame the mighty rivers that sweep away not only farmland, but also imperial reigns.” ### Designing China’s devotion to engineering extends beyond machines to the design of society itself. The Party’s confidence in planning has produced not only bridges and railways but also engineered behaviours and beliefs. Wang captures both the brilliance and the strangeness of that impulse: a conviction that any problem, *moral, urban, or human*, can be solved through design. It is a love of order that shades easily into control. The result is a world that feels both inspiring and faintly inhuman, a place where progress risks becoming procedural. > “As the United States lost its enthusiasm for engineers, China embraced engineering in all its dimensions. Its leaders aren’t only civil or electrical engineers. They are, fundamentally, social engineers.” ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/11/IMG_0419-1.jpeg) My e-reader edition of Breakneck ### Drifting Across the Pacific, Wang finds a different energy: imaginative, individualistic, and increasingly detached from the material world. Silicon Valley builds platforms, not bridges. Its optimism runs on metaphors, not machines. Wang doesn’t judge; he connects the two. America’s appetite for innovation found its outlet in China’s factories. The two economies, he writes, “co-evolved,” feeding on each other’s strengths and blind spots. > “Lawyers have so many tools available to delay or prevent building. You don’t just feel the difference going from the lawyerly society to the engineering state: you saunter, tread, and amble upon its works.” He also introduces a term that crystallises America’s condition: **Lawyrism**. It describes a society so wrapped in legal reasoning that law becomes both weapon and refuge. What began as a means of protection now preserves hierarchy. Rules replace judgment; process replaces courage. In such a culture, progress stalls not for lack of talent but for fear of liability. > “While engineers envision bridges, lawyers envision procedures.” ### Living Beyond economics, *Breakneck* is also personal. Wang writes about his parents’ migration, about the quiet costs of mobility, and about China’s one-child generation with rare tenderness. Those chapters changed how I thought about policy: not as demographic management, but as biography written by the state. They remind the reader that statistics have private echoes and that modernisation rearranges families as much as factories. > “The one-child policy is one of the searing indictments of the engineering state. It represents what can go wrong when a country views members of its population as aggregates that can be manipulated rather than individuals who have desires, goals, or rights.” > > “In moving to the West, my parents made a wrenching personal decision based on what amounted to a guess about the future… All these years later, it’s not an open-and-shut case that they made the right call.” ### Looking from Europe Reading this, I often paused to locate myself as a European. The book speaks fluently of two civilisations that define the global century, yet Europe appears mainly as context, sometimes admirable, often cautious. Our respect for engineers has waned, but perhaps less than in the US. > “Europeans have a sense of optimism only about the past, stuck in their mausoleum economy because they are too sniffy to embrace American or Chinese practices.” We still build wind turbines, EUV machines, and tunnels, but rarely mythologise them. Wang’s contrast made me see that omission more clearly: we speak about regulation, not production; about ethics, not execution. His Europe is more museum than workshop, *competent, sceptical, self-aware*, a civilisation that prefers refinement to reinvention. It is an unflattering image, yet a useful one. It shows what happens when the moral energy of law and restraint is no longer matched by the physical energy of making. The tension Wang traces, between the tangible and the abstract, is no longer academic. Tariffs, semiconductor bans, data laws are all manifestations of a deeper unease about who controls the world’s physical base. The digital layer may seem dominant, but without the material layer it floats, weightless. *Breakneck* restores that weight. For a European reader, the book is not about imitation but perspective. It reminds us that progress is made of matter as well as meaning, and that ideas lose their force when they are no longer embodied in things. [When the chips are down: Nexperia, Europe and the US-China trade and tech warThe case of Nexperia is another reminder Europeans need to speed up the assembly of their own trade and tech tools…![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/cropped-star_only_sq_black-300x300.png)European Council on Foreign Relations (ECFR)Karel Solar![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/Policy-Alert-2-1.jpg)](https://ecfr.eu/article/when-the-chips-are-down-nexperia-europe-and-the-us-china-trade-and-tech-war/) ## Closing *Breakneck* is, above all, an act of witnessing. Of someone standing in the workshop of history and noticing how the tools have changed hands. Wang’s prose moves between admiration and melancholy, precision and empathy. The result is a map of how the modern world actually operates, beneath its slogans. > “I like to imagine how much better the world would be if both superpowers could adopt a few of the pathologies of the other… I don’t want to get rid of lawyers. Rather, I want to help lift the engineers back up.” For me, the book reawakened an older intuition: that civilisation depends on balance between thought and manufacture, freedom and form. China and the US represent two extremes of that duality. Europe, if it still has a role, might be to remember how to hold them together. And most, I recommend the book, I am sure you will learn something from it while enjoying the read. ### Further reading - [Dan Wang](https://danwang.co) - [Cybernetics: The Idea Behind Cyber](https://rob-hoeijmakers.ghost.io/cybernetics/) - [🔮 Breakneck China: From superprojects to social engineering](https://www.exponentialview.co/p/how-china-gets-big-things-done) - [Strengthening Europe’s Semiconductor Backbone: The Role of ESMC](https://rob-hoeijmakers.ghost.io/esmc/) ### From Free to Paid: Choosing the Right AI Model (with a European Lens) URL: https://rob-hoeijmakers.ghost.io/ai-models-europe-free-vs-paid/ Last updated: 2026-05-10T08:53:49.000Z *Building on Ethan Mollick’s “*[*Opinionated Guide to Using AI Right Now*](https://www.oneusefulthing.org/p/an-opinionated-guide-to-using-ai-right-now)*”* Every few months, Ethan Mollick updates his *Opinionated Guide to Using AI Right Now*. His latest version (October 2025) is, in my view, the clearest overview of what the major AI models can do and when it’s worth paying for them. This article builds on his work. I’m not trying to improve it, but to translate it for a European context and for people who often ask me: > “Rob, I use AI daily. When should I start paying, and for what?” At [Schmuki](https://schmuki.nl), our small digital-AI agency, we work with all three leading paid systems, *ChatGPT, Claude, and Gemini*, each for different reasons. What follows is both a reflection and a field note: how we use them, what we’ve learned, and what might help others choose wisely. ### Quick takeaways - **Free models** are fine for experimentation, learning, or light creative work. - **Paid tiers (€20–€25 per month)** unlock reliability, integrations, and privacy control. - **Three ecosystems dominate:** OpenAI (ChatGPT), Anthropic (Claude), and Google (Gemini). - **Europe adds nuance:** pricing, privacy, data access, and multilingual use differ. - **Your choice depends not only on capability but on context: personal or business.** ### Why upgrade Free AI gives you access; paid AI gives you *agency*. Upgrading unlocks: - access to “thinking” or “agentic” models that plan, search, and reason in steps; - longer context windows for sustained work; - file uploads, code execution, and voice interaction; - privacy settings that let you exclude your data from future training; - team features for shared workspaces and centralised billing. For many professionals, that alone justifies the cost: fewer errors, more continuity, and tools that fit organisational routines. ### The big three: a comparative snapshot | System | Strengths | Limitations | Best suited for | | ----------------------------- | -------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------ | --------------------------------------------------------------------- | | **ChatGPT (GPT-5 family)** | Most balanced; strong reasoning; image, video & code; good voice mode; reliable mobile app. | “Auto” mode sometimes switches to lighter models; integration beyond its own ecosystem still evolving. | Consultants, educators, creators who want one consistent environment. | | **Claude (Sonnet 4.5)** | Calm, long-context reasoning; strong with tone, structure & documents; easy data-privacy controls. | No image/video creation; slower on complex computation. | Writers, analysts, multilingual teams working with text and reports. | | **Gemini (2.5 / Deep Think)** | Deep integration in Google Drive, Docs, Gmail; best visual analysis; web search built-in. | Integration can be inconsistent; iOS app language handling uneven; interface sometimes over-extended. | Researchers, students, design & content teams in Google Workspace. | *(Grok by xAI exists but remains peripheral in professional European use — see note below.)* [When Research Turns Into Action: Understanding Deep Research vs. Agent Mode in ChatGPTBoth modes inside ChatGPT seem to “think”, yet one begins to “do”. Exploring the quiet shift from analysis to operation in everyday AI tools.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-303.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_9001-1.jpeg)](https://rob-hoeijmakers.ghost.io/deep-research-vs-agent-mode/) ### A note on Grok I’ve never used Grok seriously. It sits inside the X ecosystem and doesn’t yet connect with the tools we depend on. Perhaps it’s about product safety or about trust, but more likely it simply hasn’t offered a reason to explore. Omissions are instructive: what we *don’t* use often says as much about our digital ecosystems as what we do. ### How we work with them at Schmuki At [**Schmuki**](https://schmuki.nl), we maintain paid plans for all three systems. Each serves a distinct role. **ChatGPT Business** is our main environment. It doesn’t *integrate* with Google Workspace in a technical sense, but it *coexists* comfortably with it. The two can share files and links easily, and ChatGPT understands Google formats well enough to make handovers smooth. We also appreciate the practical side: a single annual invoice in euros and clear administrative handling. It’s where most of our writing, planning, and prototyping happens. 🧠 ****Memory works differently in ChatGPT Personal and ChatGPT Business.** In Personal, the assistant is optimised for individual continuity and may feel more attuned to your recent chats, preferences, and ongoing themes, even across conversations. In Business, memory is deliberately stricter and more segmented: conversations are kept separate, personal inference is limited, and long-term memory is scoped to explicit, work-relevant information within the workspace. This design reduces accidental context bleed between tasks, people, and roles, but can make Business feel more stateless unless expectations or preferences are stated explicitly. **Claude** is Maya’s preferred workspace for editorial and conceptual writing. Its composure and linguistic precision make it ideal for sensitive or multilingual texts. **Gemini** has two faces: - *as an app*, for quick, visual prompts or camera-based interactions; - *as an embedded co-pilot*, inside Drive, Docs, and Gmail, which we use for internal briefs and summarising materials. But the *Google heritage* is a double-edged sword. The same deep integration that makes Gemini convenient can also make it feel heavy: it tries to be everything at once, and not everything works as you expect. That friction sometimes breaks the flow that good AI work depends on. The result, nonetheless, is complementarity rather than redundancy. Each model contributes something distinct and together they mirror the variety of our work: strategic, creative, operational. 💡 ****Note on** [****Google One AI Premium**](https://rob-hoeijmakers.ghost.io/google-one-vs-google-workspace-for-gemini/) If you’re a personal Google user, Gemini Advanced is included in the **Google One AI Premium* plan. Once you turn on data access, Gemini can search across your own Gmail, Drive, and Calendar, a powerful feature for personal knowledge work. It doesn’t happen automatically, and your data stays private to your account. For many users, this makes the Google One bundle an affordable entry point into serious, privacy-aware AI use. ### A European perspective Europe adds a few pragmatic twists to Mollick’s global overview: | Aspect | European nuance | | ----------------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | **Pricing & billing** | Gemini can be bundled into Google One or Workspace subscriptions; ChatGPT Team offers euro billing & annual invoicing; Claude’s pricing varies per seat. | | **Privacy & data use** | GDPR culture makes opt-out options valuable. Claude and ChatGPT both allow full exclusion from model training; Gemini links deeper into your personal data. | | **Language support** | ChatGPT handles multilingual dialogue smoothly; Gemini still struggles to switch languages on iOS; Claude’s tone control helps in translation and cross-lingual drafts. | | **Digital sovereignty** | European institutions increasingly favour open-weight or locally hosted models (Mistral, Aleph Alpha) for compliance reasons. | | **Ecosystem alignment** | Many European SMEs live inside Google Workspace or Microsoft 365\. The most “native” assistant often wins by convenience, not capability. | These subtleties shape the *feel* of using AI here, less about novelty, more about fit, trust, and compliance. ### Personal or business? Upgrading is not only about price; it’s about *ownership of process*. For personal use, one subscription may suffice and ChatGPT Plus or Gemini Advanced will cover nearly everything. For business, the equation changes: - **Shared memory and projects** enable continuity across a team. - **Unified billing** simplifies administration and auditing. - **Data-handling assurances** become part of compliance. The move from free to paid is, in that sense, a move from *play* to *practice*. [Google One vs Google Workspace: Where do you get real AI and Gemini Advanced?What do you get when you pay for AI at Google? An exploration of Gemini Advanced, NotebookLM and how to share it smartly via Google One.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-302.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_6949-3.jpeg)](https://rob-hoeijmakers.ghost.io/google-one-vs-google-workspace-for-gemini/) ### Closing reflection Mollick is right: the future of AI will depend less on faster models than on how people learn to use them well. In Europe, that means balancing curiosity with caution and exploring new capabilities while guarding privacy, language, and autonomy. Paid AI is not a luxury. It’s a way of working with intention: choosing the right environment for the task, and understanding the boundaries of trust that come with it. **Further reading** \- Ethan Mollick: [*An Opinionated Guide to Using AI Right Now*](https://www.oneusefulthing.org/p/an-opinionated-guide-to-using-ai-right-now) ### Further reading - [When AI Makes the Calendar Click](https://rob-hoeijmakers.ghost.io/gemini-google-calendar/) - [Choosing ChatGPT Team: From AI-Confidant to Trusted Colleague](https://rob-hoeijmakers.ghost.io/choosing-chatgpt-team-from-ai-confidant-to-trusted-colleague/) ### Rediscovering Google Cloud Search URL: https://rob-hoeijmakers.ghost.io/rediscovering-google-cloud-search/ Last updated: 2026-05-10T08:53:49.000Z Google Search is changing fast. What used to be a tool for *finding* things is increasingly becoming a tool for *summarising* them and it is powered by Gemini and other AI layers that tell you what they think you need to know. That’s useful, but it also blurs the original purpose of search: helping you locate the exact source, not its interpretation. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/10/Google-Cloud-Search.jpg) Google Cloud Search is a slightly confusing name because of the Google Cloud products. It is more Google Workplace Search in my view. ## A Different Kind of Search Only today I discovered [*Google Cloud Search*](https://cloudsearch.google.com), a different creature altogether. It doesn’t summarise but it retrieves. Within [Google Workspace](https://rob-hoeijmakers.ghost.io/google-one-vs-google-workspace-for-gemini/) it quietly indexes your own universe of files: Drive documents, Gmail attachments, even calendar entries. It’s fast, private, and entirely focused on access rather than explanation. ## Clarity Before Interpretation For me, that distinction is refreshing. Once the right item is found, Gemini or another assistant can summarise, extract, or act on it. But Cloud Search handles the foundational step, that is finding the thing itself, and does so remarkably well. It’s surprising how little attention this service receives. The interface may look utilitarian, but it offers something essential that modern AI search often forgets: clarity before interpretation. [Google Cloud Search![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/144X144.png)Google Cloud Search](https://cloudsearch.google.com) ### Further reading - [Search in the Age of AI](https://rob-hoeijmakers.ghost.io/search-in-the-age-of-ai/) - [Google One vs Google Workspace: Where do you get real AI and Gemini Advanced?](https://rob-hoeijmakers.ghost.io/google-one-vs-google-workspace-for-gemini/) ### The Silent Upgrade: What Really Changed When Our Bank Replaced Our Cards URL: https://rob-hoeijmakers.ghost.io/the-silent-upgrade-mastercard-debit-transition/ Last updated: 2026-05-10T08:53:49.000Z Our business recently received new cards from our Dutch bank. At first glance, nothing special. Just a smooth, well-organised process: new debit cards, new credit cards, everything arrived on time, activation went flawlessly. But the moment I noticed the new Mastercard Debit logo instead of the old Maestro one, I realised something had quietly shifted. And as it turned out, this was not just a rebrand or a renewal cycle, it’s part of a much larger transformation in how payments, identity, and digital infrastructure work. So I decided to figure it out. Here’s what’s really going on, and why it matters more than you’d expect. ## From Maestro to Debit Mastercard For decades, Maestro was Europe’s standard debit network. It worked but only within its own boundaries. Payments were authorised in real time, tied to your [IBAN](https://rob-hoeijmakers.ghost.io/eurozone-invoices-sepa-qr-ubl/), but the card couldn’t be used everywhere online, and many global merchants didn’t support it. Mastercard has now replaced Maestro with **Debit Mastercard**, which operates on the same global network as its credit cards. That means one infrastructure for both debit and credit, one set of security standards, and one model for digital payments. On paper, nothing changes for you. Your money stays in the same account; your IBAN remains the same. But under the surface, your card, and every card in Europe, is in a switching process for their networks. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/10/IMG_9724.jpeg) Maestro left, Mastercard Debit right. ## The New Layer: PAN, Tokens, and Wallets What makes this upgrade significant is not the card itself, but the **architecture behind it**. Your Debit Mastercard now has a **PAN** (Primary Account Number), a 16-digit identity, separate from your IBAN. This PAN can be *tokenised*: converted into secure, digital representations that work in Apple Wallet, Google Pay, or on merchant platforms. That’s why when the new card arrived, **I didn’t have to do anything in Apple Wallet**. The digital token updated automatically, no re-adding, no new numbers. The connection between my bank, Mastercard’s token service, and Apple Pay handled the entire update silently in the background. It’s the same reason why, around the same time, our **credit card** (used for Google Workspace payments) also got a new number and Google updated it automatically. Both cards were renewed through the same **tokenisation infrastructure**, even though they’re different products. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/10/blog-1-1.jpg) PAN-number, image by [IServeU](https://iserveu.in/blog/card-management-system/2024/08/08/everything-you-need-to-understand-about-your-card-number-deep-dive-into-primary-account-number-pan-and-bank-identification-number/). ## Where This Meets Daily Business This is where the change stops being abstract. If you run payments, manage subscriptions, or make online purchases for your business, you’ll run into this sooner or later. Suddenly, you’ll need to know where to find your card details because the new physical cards often don’t have numbers printed on them anymore. Colleagues used to reaching for a card to type in details will now have to open the banking app, copy the virtual card number, and sometimes check the CVC there. That’s not a technical issue. It’s **a shift in how we handle everyday operations**. The update makes everything safer and smoother in the long run, but it also means teams need to adapt, especially people who don’t think about payment networks or card architecture for a living. They just want things to work. And as a business, it becomes part of your job to make sure they do. 💡 ****Updating your credit card in Safari** If you use Safari’s Autofill for payments, update your card in ****Safari → Autofill → Credit Cards** (Mac) or ****Settings → Safari → Autofill → Saved Credit Cards** (iPhone/iPad). When ****iCloud Keychain** is on, Safari securely syncs your new card across all your Apple devices, so you only need to update it once. ## Why This Matters for Businesses For users, this is convenience. For businesses, it’s infrastructure. 1. **Continuity:** recurring payments keep working even after a card is replaced. No more failed subscriptions because of expired cards. 2. **Security:** merchants never see your actual card number, only a token that can be revoked or rotated. 3. **Automation:** updates happen centrally through the Mastercard network, not through manual customer actions. From a business administration standpoint, this also changes how payments behave in the background. Your accounting systems will still match transactions by IBAN, but card-based payments may now show up differently, it goes through Mastercard’s network identifiers instead of Maestro’s. It’s a small detail, but one that signals how **the identity of money is becoming digital**, governed by tokens, not numbers. ## The Bigger Picture This isn’t just a technical update. It’s a reflection of how payment systems are evolving in step with the rest of the digital economy. - **Payments become identity:** your card is no longer just a piece of plastic; it’s a verified digital credential. - **Networks become invisible:** banks, tech platforms, and merchants are now part of one automated layer that quietly keeps things in sync. - **Europe catches up:** with Maestro retired, European debit cards are now compatible with global payment standards, essential for cross-border business and e-commerce. In short, what looked like a simple card replacement is actually a **migration of trust**. From physical cards to digital identity systems. ## A Quiet Revolution So yes, the process went smoothly. But behind that silence, a lot changed. If you run a business, manage online payments, or handle subscriptions, this is worth understanding. Your debit and credit cards now live in the same digital ecosystem and that ecosystem is designed for a world where payments, authentication, and automation are all connected. Sometimes the biggest shifts in technology aren’t loud. They arrive in the post, in a small white envelope, with a note that simply says: *your new card has arrived. (And yes, it will probably haunt us all for a while).* ### Further reading - [Payment requests: Your digital payment assistant](https://rob-hoeijmakers.ghost.io/payment-requests-your-digital-payment-assistant/) - [Wero: Europe’s new payment brand](https://rob-hoeijmakers.ghost.io/wero-europes-new-payment-brand/) - [Invoices in the Eurozone: A Practical Guide to SEPA Payments and Digital Processing](https://rob-hoeijmakers.ghost.io/eurozone-invoices-sepa-qr-ubl/) ### When Research Turns Into Action: Understanding Deep Research vs. Agent Mode in ChatGPT URL: https://rob-hoeijmakers.ghost.io/deep-research-vs-agent-mode/ Last updated: 2026-05-10T08:53:49.000Z A friend recently messaged me on WhatsApp: *“What’s the difference between Deep Research and Agent Mode in ChatGPT?”* It’s a fair question. They both seem to do research. But the distinction, I realised while replying, goes deeper than that. It’s not just about what the AI knows; it’s about how it works. And that difference tells us something about how intelligence itself becomes operational. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/10/Agent-mode-versus-Deep-Research.jpg) Deep research and Agent mode as options in ChatGPT. ## Quick takeaways - **Deep Research** is about *thinking*: one focused burst of analysis. - **Agent Mode** is about *doing*: a repeatable process that applies what it knows. - The difference isn’t in intelligence, but in structure. Cognition versus operation. - In human terms: one writes a report, the other manages a system. - Both matter when knowledge needs to stay alive and evolving. ## Reflection When I use Deep Research, I’m effectively asking the model to act as a researcher. To read widely, connect dots, and compose an interpretation. It’s a one-time act of understanding. Like a well-written essay, it captures a moment in thought. Agent Mode feels different. It introduces procedural memory. The same intelligence can now act: fetch new data, update files, follow rules, or run again next week. It doesn’t think harder; it organises its thinking in time. That makes it less a conversation partner and more a quiet colleague, one that never forgets what you asked it to keep doing. The beauty, and perhaps the challenge, lies in how close these two modes now sit. What used to be a distinction between knowing and working, between intelligence and labour, begins to blur. An AI can both write the report and schedule its next update. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/04/deep-research-versus-agent.png) Deep research versus agent mode. ## Closing thought So yes, my answer to my friend was simple: Deep Research is about *insight*, Agent Mode is about *execution*. But what lingers for me is this: once thinking becomes executable, knowledge itself starts to behave like software. ### Further reading - [ChatGPT Connectors: The Emergence of a New Interface Layer](https://rob-hoeijmakers.ghost.io/chatgpt-connectors/) - [From Free to Paid: Choosing the Right AI Model (with a European Lens)](https://rob-hoeijmakers.ghost.io/ai-models-europe-free-vs-paid/) ### When Bots Become Readers: Publishing in the Age of AI Crawlers URL: https://rob-hoeijmakers.ghost.io/when-bots-become-readers/ Last updated: 2026-05-10T08:53:50.000Z I listened to a conversation between **Azeem Azhar** and **Matthew Prince**, the co-founder of [**Cloudflare**](https://www.cloudflare.com/en-gb/). Prince described how a large share of today’s internet traffic no longer comes from people but from machines, these are bots that index, scrape, and simulate human behaviour. I enjoyed the discussion. It was clear, grounded, and made me look at my own practice differently. Because I host my Ghost site with Cloudflare as DNS, I realised that every request for my blog also passes through their network. In other words: I, too, publish into a space where machines are part of the audience. [🔮 How Al is breaking and rebuilding the internet economyListen now (47 mins) | From search to answers | With Cloudflare co-founder & CEO Matthew Prince![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/https-3A-2F-2Fbucketeer-e05bbc84-baa3-437e-9518-adb32be77984.s3.amazonaws.com-2Fpublic-2Fimages-2Faa7453ed-eb02-46d7-b523-74d86b1c1e24-2Fapple-touch-icon-180x180-5.png)Exponential ViewAzeem Azhar![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/https-3A-2F-2Fsubstack-post-media.s3.amazonaws.com-2Fpublic-2Fimages-2Feeb59c29-cc38-4c47-bb2f-6291b29c18c3_1276x968.png)](https://www.exponentialview.co/p/matthew-prince-ai-internet) I really enjoyed this episode. Matthew Prince became a tangible person when he said he and his wife own the local newspaper in there hometown. ## Quick takeaways - Much of today’s web traffic is automated and driven by bots that feed search engines, analytics, or AI models. - Cloudflare’s tools make this visible, and even give you control over which bots reach your content. - The web’s original exchange , *free access in return for visibility and visitors*, is eroding as answer engines replace search engines. - For small publishers, full protection is mostly symbolic; the real shift lies in understanding how your content now circulates. - Openness can still be a deliberate act, not a default or a weakness. ## Seeing who’s really visiting Cloudflare sits quietly between my site and the world. Every page request passes through its network: caching, securing, accelerating. It’s a layer I rarely think about, until I open its dashboard and see that many of my “readers” are actually bots. Some are harmless, these are search crawlers, uptime monitors, link checkers. Others belong to the new generation of AI systems, scanning text to train models that may soon write answers instead of sending users back to my page. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/10/Cloudflare-AI-Crawl-Control.jpg) Cloudflare's AI Crawl Control ## The changing contract of the web This used to be the web’s implicit contract: publish something open, get indexed, and in return receive visitors. Attention was the currency. Now the exchange feels different. Content is still taken in, but the return flow, traffic, attribution, dialogue, becomes thinner. The web is less a marketplace of exchange and more a vast input field for machines learning to speak. ## Still writing, still designing And yet, I keep beautifying it. I choose the fonts, arrange the photographs, edit the light in each image until it fits the tone of the words. I design a space where thought and craft meet and a place meant to be *seen*. It’s strange to realise that many of those who “see” it will never perceive the design at all. The bots take only the HTML; the care, the texture, the human choices disappear in translation. Still, I publish. Because somewhere in the mix there are people too, a handful who arrive through social media, or through a link I share in conversation. And even if not, the site remains a living reference: an archive in motion, a record of how ideas evolve over time. ## A deliberate openness I could block the major AI crawlers. I could fence off the site behind digital hedges. For now, I don’t. But awareness itself feels like a form of agency, knowing who (or what) comes knocking, and deciding consciously whether to open the door. 💡 There’s a small but growing movement around [****LLMs.txt**](https://llmstxt.org), a simple text file you can place at the root of your website, much like `robots.txt`, to tell AI crawlers what your site is about and under what conditions they may use your content. Some creators use it to describe their purpose and set boundaries; others see it as a way to shape how machines “see” their work. It’s still early, but the idea fits this moment: we can’t stop the bots, yet we can start speaking their language, on [our own terms](https://rob-hoeijmakers.ghost.io/llms.txt). ## Closing Perhaps this is what authorship now means: creating for an unseen, hybrid audience, part human, part machine, and continuing anyway. The web may no longer guarantee reciprocity, but it still allows us to shape and share. I’ll keep doing so, because each page remains a trace of attention, and attention, even unreturned, is still a form of care. ### Further reading - [Flying Blind: Measuring Traffic When Your Readers Are Machines](https://rob-hoeijmakers.ghost.io/measuring-traffic-machines-bots/) - [Search in the Age of AI](https://rob-hoeijmakers.ghost.io/search-in-the-age-of-ai/) - [The End of Google Search (as we know it)](https://rob-hoeijmakers.ghost.io/the-end-of-google-search-as-we-know-it/) ### When AI Makes the Calendar Click URL: https://rob-hoeijmakers.ghost.io/gemini-google-calendar/ Last updated: 2026-05-26T19:37:57.000Z It started with a hotel booking. Nothing special, just an email with the usual details. No invitation, no `.ics` file, no automatic way for it to show up in my calendar. Normally I’d copy the date, paste it into Calendar, add the address, check the time. This time I didn’t. I pasted the whole thing into [Gemini](https://rob-hoeijmakers.ghost.io/google-one-vs-google-workspace-for-gemini/) and said: *“Put this in my agenda.”* And it did. Instantly. The event appeared with the right title, time, and location. It felt strange, quiet, effortless, almost invisible. 0:00 /0:18 1× From unstructured data to a structured invite into my calendar in seconds. ## What it replaced What struck me wasn’t the feature itself but what it replaced. A few years ago, something like this would have needed layers of code and validation. Every date format, every phrase, every possible input had to be accounted for. Now I just write a sentence and it understands what I mean. Part of this comes from the way Google’s tools fit together. Gmail, Drive and Calendar all speak the same language. Gemini doesn’t just generate text, it moves within that system. It knows where the calendar is and how to add something to it. That’s why it feels less like a script and more like an assistant that lives inside your digital life. [Google One vs Google Workspace: Where do you get real AI and Gemini Advanced?What do you get when you pay for AI at Google? An exploration of Gemini Advanced, NotebookLM and how to share it smartly via Google One.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-281.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_6949-1.jpeg)](https://rob-hoeijmakers.ghost.io/google-one-vs-google-workspace-for-gemini/) ## A quiet change I’ve written before about calendars as diaries and about how digital invitations structure our days. This moment added another layer to that thought. The calendar is no longer just a place where I put events; it’s becoming something that understands them. The line between writing something down and having it scheduled is starting to disappear. It’s a small thing, a hotel booking, a single prompt, but it shows where everyday software is heading. Tools that quietly understand us instead of asking us to explain ourselves. I like that. ### Further reading - [Digital scheduling calendars compared and explained](https://rob-hoeijmakers.ghost.io/digital-diary-and-calendars/) - [Regular and online calendar invitations](https://rob-hoeijmakers.ghost.io/understanding-digital-calendar-invitations/) - [From Calendly to scheduling with Google Calendar](https://rob-hoeijmakers.ghost.io/from-calendly-to-scheduling-with-google-calendar/) - [From Free to Paid: Choosing the Right AI Model (with a European Lens)](https://rob-hoeijmakers.ghost.io/ai-models-europe-free-vs-paid/) ### The Hidden History Behind an Email Signature in Apple Mail URL: https://rob-hoeijmakers.ghost.io/apple-mail-signature/ Last updated: 2026-04-25T14:46:33.000Z For something that looks so simple, e-mail is anything but. It’s old, fragile, and deeply strange. I’ve been working with it for about thirty years, and even now it surprises me. One of those surprises came recently, not from some complex server setup, but from something as trivial as a signature, the little block of text at the bottom of every e-mail. For years I wondered: why is it so difficult to create a nice, consistent signature in Apple Mail? Something clean, with a logo, a bit of colour, the right typefaces. It felt like such a basic thing, yet whatever I tried, it always broke. Sometimes the spacing changed, sometimes the font jumped, sometimes images disappeared. I even bought external software just to get a “proper” signature. It took me three decades to realise what was going on. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/10/image-1.png) The arcane signature interface in Apple Mail ## The RTF Years Apple Mail’s internal logic comes from another time. For a long period it didn’t really speak *HTML*, the language of the web. It spoke *RTF*, Rich Text Format, a relic from the pre-web desktop era. It looked like HTML, but it wasn’t. Under the hood, Mail stored your signature as a `.mailSignature` file, full of strange formatting codes that belonged to WordPad, not the web. Only gradually did Apple transition from RTF to HTML. Today, Mail fully supports HTML for composing and rendering messages, but oddly enough, in the interface it still says *Rich Text*. That one small word caused years of confusion for me. I assumed I had to create RTF content but not HTML. Once I realised the opposite was true, the whole thing clicked into place. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/10/image-3.png) The signature file on the drive, proof that it is HTML. It also is mime-encoded, a requirement for e-mail. ## The HTML Revelation So now I could just create an HTML page that represented my signature. Design it however I wanted: proper CSS, proper layout, real web code. Then open it in a browser, select everything (not the code, but what you *see*) and paste it straight into Mail’s signature field. That’s it. Except, of course, it isn’t. E-mail clients are notoriously inconsistent. What works in one might break in another. Inline styles, absolute URLs, no external fonts because the HTML has to be “simplified” in a very specific way. But once you know that, it becomes manageable. Especially with ChatGPT, which now acts like a surprisingly capable HTML assistant that understands the oddities of Outlook, Gmail, and Apple Mail alike. 💡 I made my HTML snippet in GitHub pages. It enabled me to manage the HTML and host the images. [The signature code in GitHub](https://rhoeijmakers.github.io/signature-assets/schmuki/rob.html). ## The Safari Trap Then came the funny part. My workflow involves both Safari and Chrome. I built the signature, tested it, copied it from the browser into Mail and it looked perfect. A small miracle. Later, without thinking, I opened the same HTML file in Safari, copied again, pasted again… and suddenly everything was broken. Images gone, layout distorted, back to square one. After some head-scratching, I discovered the cause: Safari doesn’t copy raw HTML into the clipboard. It creates a *web package*, that is a kind of bundle with local image references. Those local references don’t survive the trip into Apple Mail. In other words: Safari sabotaged my signature. Chrome, on the other hand, behaves like a proper web browser and copies the rendered HTML cleanly. So there it was, the mystery solved! After all these years, the trick to making a perfect Apple Mail signature was not an app, not a script, but a simple rule: **use Chrome, not Safari.** ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/10/image-2.png) Copy and paste HTML, not the code but the visual form in the Chrome browser. ## Why This Still Matters It’s a small, almost comical discovery. Yet it says something about how old technologies evolve: slowly, unevenly, full of quirks that only reveal themselves through experience. E-mail is a living fossil of the Internet but still running on protocols designed decades ago, still patched and layered over. And now, thanks to modern AI tools, even these ancient corners of the web can be explored with fresh eyes. That’s what I like most about this little story. It’s not really about signatures. It’s about the quiet satisfaction of finally understanding something that’s been bugging you for years and the strange new fact that I solved it with help from ChatGPT. Small progress, perhaps. But deeply satisfying. --- ### How to Create an HTML Signature for Apple Mail Here’s the practical recipe that actually works — stripped to the essentials. **Create your signature as a small HTML page.** - Keep it narrow (around 600 px wide). - Use inline CSS styles only. - Use *absolute URLs* for all images (e.g. `https://example.com/logo.png`). **Open the HTML file in Chrome.** - Don’t use Safari; it converts clipboard data into a *web package* that breaks image references. **Select everything you see** (not the source code) and **copy**. **Open Apple Mail → Settings → Signatures.** - Create a new signature, clear the content, and **paste** what you copied. Make sure to uncheck the tick box that says: Always match my default message font. **Test it.** - Send yourself a message and check it in different clients (Mail, Gmail, Outlook). - If something looks off, simplify your HTML and re-test. **Tip:** ChatGPT can generate or simplify HTML for e-mail clients if you ask it to “make it Apple Mail compatible”. It understands most of the quirks and saves hours of trial and error. ### Tracing and Telling: From GPX Tracks to KML Stories URL: https://rob-hoeijmakers.ghost.io/gpx-kml-ai-maps/ Last updated: 2026-05-10T08:53:50.000Z Most of my recent writing has been on the heavy side: regulation, AI strategy, the politics of data. This time, I wanted something lighter. For years I had a wish: to really understand the difference between GPX and KML. These formats have been around for decades, yet I never quite grasped how they complemented each other. With ChatGPT, I finally sorted it out and it didn’t stop at theory. The model can also generate the files directly, so instead of just reading about it, I was able to *make* maps. And that turned out to be surprisingly fun and practical. ## Quick takeaways \- **GPX = the track**: where you’ve been or want to go. \- **Komoot = the practical layer**: my go-to for planning and tracing routes, instead of a bike computer. \- **KML = the meaning layer**: adding stories, photos, and context. \- **AI link**: ChatGPT can generate both GPX and KML files, turning ideas into working maps. \- **Concrete use**: I built a bike route past all the big consultancy and accountancy firms in Amsterdam. ## Reflection The distinction is straightforward: GPX is the trace; KML is the enrichment. Komoot helps in between by making planning easy. But the real shift comes when you combine these with AI. As an experiment, I asked ChatGPT to generate a KML file with the major consultancy and accountancy firms in Amsterdam. I loaded it into [Google Maps](https://www.google.com/maps/d/u/0/edit?mid=1%5FUcqRyevulKKD9rbe4-p0tXaC3WwVmA&usp=sharing), connected the dots, and created a route that linked them all. [Big four + Consultancy Amsterdam - Google My MapsBike ride to learn about where the big accountancy and consultancy firms are in Amsterdam.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon-001.ico)Google My Maps![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/thumbnail)](https://www.google.com/maps/d/u/0/edit?mid=1%5FUcqRyevulKKD9rbe4-p0tXaC3WwVmA&usp=sharing) From there, I exported the map as a **KML**, converted it to **GPX** using [GPS Visualizer](https://www.gpsvisualizer.com/), and imported the GPX into **Komoot**. Suddenly, I didn’t just have an abstract map, I had a rideable route on my phone, ready to follow in real life. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/10/image.png) Export the Google Maps [Komoot](https://www.komoot.com) kept me on track while riding, turning the experiment into a tangible experience. What started as a playful idea with AI-generated data became an actual cycling route through Amsterdam’s corporate landscape. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/10/IMG_9655.jpeg) The Amsterdam route on Komoot. ## Closing It’s oddly satisfying to see these long-standing technical formats turn into something concrete and functional in my own life. GPX gives the trace, Komoot helps me plan, and KML lets me build context. With AI as a bridge, maps are no longer just navigational tools, they are becoming ways to explore, to teach, and to tell earth bound stories. 💡 The ****Big Four** ([Deloitte](https://www.deloitte.com/global/en.html?icid=site%5Fselector%5Fglobal), [PwC](https://www.pwc.com/gx/en.html), [EY](https://www.ey.com/en%5Fgl), [KPMG](https://kpmg.com/xx/en.html)) dominate accountancy; the ****MBB** firms ([McKinsey](https://www.mckinsey.com), [BCG](https://www.bcg.com), [Bain](https://www.bain.com)) lead strategy consulting; and [****Roland Berger**](https://www.rolandberger.com/en/), from Germany, shows how Europe has its own global player, with Amsterdam as one of the key hubs. --- [Directions from PwC to KPMGDirections from PwC to KPMG.kmz11 KBdownload-circle](https://rob-hoeijmakers.ghost.io/content/files/2025/10/Directions-from-PwC-to-KPMG.kmz "Download") [20251001113302-96036-dataThe GPX to be imported in your preferred app or tool. 20251001113302-96036-data.gpx45 KBdownload-circle](https://rob-hoeijmakers.ghost.io/content/files/2025/10/20251001113302-96036-data.gpx "Download") --- ### Further reading - [Komoot | Find, plan and share your adventures](https://www.komoot.com) ### From Hype to Workflows: Where AI’s Real Shift is Happening URL: https://rob-hoeijmakers.ghost.io/from-hype-to-workflows-where-ais-real-shift-is-happening/ Last updated: 2026-05-10T08:53:50.000Z I’ve been reading a lot lately on AI and work. One theme keeps coming up: people expected to see more visible change by now. After the hype of ChatGPT and Copilot, many organisations wonder, is this it? Is AI already stagnating? Or are we still on the edge of something bigger? This article is a distillation of what I’ve been seeing in research, case studies, and my own work. I want to share why I think the disappointment is understandable, but also misleading. ## The Horizontal Wave: Loud, Visible, and Sometimes Cluttered The first wave of AI at work has been *horizontal*. Tools like ChatGPT, Copilot, or Claude spread quickly across teams. They’re easy to try, and they promise immediate help with tasks: drafting emails, summarising meetings, rephrasing text. That visibility is exactly why expectations rose so high. Everyone could see colleagues using AI in small ways. Yet, [as McKinsey notes](https://www.mckinsey.com/capabilities/quantumblack/our-insights/seizing-the-agentic-ai-advantage), almost 80 percent of companies report using generative AI, but most still see no real bottom-line impact. [![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/09/image.png)](https://www.mckinsey.com/capabilities/quantumblack/our-insights/seizing-the-agentic-ai-advantage) Seizing the agentic AI advantage, McKinsey There’s another catch: horizontal AI doesn’t just make work easier, it can make *more* work. Because it’s so simple to generate content, companies risk drowning in drafts, slide decks, and documents. [Ethan Mollick joked](https://www.oneusefulthing.org/p/real-ai-agents-and-real-work?utm%5Fsource=substack&publication%5Fid=1180644&post%5Fid=174807342&utm%5Fmedium=email&utm%5Fcontent=share&utm%5Fcampaign=email-share&triggerShare=true&isFreemail=true&r=49ph0&triedRedirect=true) about generating 17 versions of the same PowerPoint. It’s productivity on paper, but often it clogs the information stream instead of clarifying it. [Real AI Agents and Real WorkThe race between human-centered work and infinite PowerPoints![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/https-3A-2F-2Fbucketeer-e05bbc84-baa3-437e-9518-adb32be77984.s3.amazonaws.com-2Fpublic-2Fimages-2F233e9990-95ea-44ab-afc4-5abfe44f24d4-2Fapple-touch-icon-180x180.png)One Useful ThingEthan Mollick![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/https-3A-2F-2Fsubstack-post-media.s3.amazonaws.com-2Fpublic-2Fimages-2Ffbe42bc3-4c1b-40db-9247-a50466387ded_1376x864.png)](https://www.oneusefulthing.org/p/real-ai-agents-and-real-work?utm%5Fsource=substack&publication%5Fid=1180644&post%5Fid=174807342&utm%5Fmedium=email&utm%5Fcontent=share&utm%5Fcampaign=email-share&triggerShare=true&isFreemail=true&r=49ph0&triedRedirect=true) ## The Vertical Wave: Slow, Invisible, and Disruptive The second wave is *vertical*. This is where AI doesn’t just sit on top of work but gets woven into specific processes and often requires those processes to be redesigned. That’s harder. It demands governance, trust, and integration with messy company systems. Which is why most vertical AI use cases still get stuck in pilots. But when it works, it’s not about saving minutes. It’s about restructuring weeks of work. AI agents drafting credit memos, coordinating supply chains, or even replicating complex academic research. The gains here aren’t incremental; they’re transformative. This is where the “sinkhole” analogy fits. On the surface, not much looks different. But underground, workflows are hollowing out and being rebuilt. When the shift breaks through, it won’t look like a shiny new assistant. It will look like parts of the organisation suddenly working in an entirely different way. ## Proof or Narrative? Do we have proof this is happening? Not yet in full. What we have are signals. - **Tasks.** AI models can now complete complex, hours-long tasks at near-human levels:contentReference\[oaicite:4\]{index=4}. That doesn’t replace whole jobs, but it reshapes them. - **Jobs.** Stanford data shows that since AI adoption accelerated, early-career workers (ages 22–25) in AI-exposed occupations saw employment fall by 13 percent compared to peers. That suggests real labour market effects are already showing, concentrated at the entry level. So no, AI isn’t just hype. It’s not stagnating either. It’s moving quietly, in ways that don’t always show up on the surface yet. [Canaries in the Coal Mine? Six Facts about the Recent Employment Effects of Artificial IntelligenceThis paper examines changes in the labor market for occupations exposed to generative artificial intelligence using high-frequency administrative data from ADP, the largest payroll software provider in the United States.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/cropped-favicon-1-270x270-1.png)Stanford Digital Economy Lab![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/CanaryPageBanner-1.jpg)](https://digitaleconomy.stanford.edu/publications/canaries-in-the-coal-mine-six-facts-about-the-recent-employment-effects-of-artificial-intelligence/) ## Why I Wrote This I hear the disappointment a lot: *“I thought by now AI would have changed everything.”* It’s a fair reaction. Horizontal AI has been loud but shallow, and its benefits are often invisible in the balance sheet. But I also see the deeper shift starting: the vertical integration of AI into real workflows. That’s harder to spot, and slower to scale, but it’s also where the genuine disruption lies. So if you’re wondering whether the AI hype is fading, my take is this: the surface is calm, but the ground underneath is moving. The real wave hasn’t hit yet. --- *Horizontal AI reshapes how we do work today, for better and sometimes for worse.* *Vertical AI will redefine what work is tomorrow.* ### Further reading - [AI’s Impact on Work, People, and Labour Law](https://rob-hoeijmakers.ghost.io/ais-impact-on-work-people-and-labour-law/) - [The Thermostat Effect: Why AI Feels Distant but Works Deep](https://rob-hoeijmakers.ghost.io/thermostat-effect-on-ai/) ### AI’s Impact on Work, People, and Labour Law URL: https://rob-hoeijmakers.ghost.io/ais-impact-on-work-people-and-labour-law/ Last updated: 2026-05-10T08:53:51.000Z AI enters organisations with familiar promises: more growth, greater efficiency. Yet the moment these systems touch work, a second effect emerges. The work itself changes, and so do the rules, rights, and responsibilities around it. What begins as a technical or strategic decision quickly turns into a legal and human question. ### Quick takeaways - AI applications in organisations often pursue **growth** (more clients, new markets) or **efficiency** (doing the same with fewer people). - These shifts have **direct effects** on employees and their daily work. - They also trigger **indirect effects**: new legal, ethical, and organisational questions (privacy, discrimination, works councils, contracts). - For strategists, builders, and managers, it pays to understand the **labour law context** early on. ## A web of change In earlier essays I explored [the thermostat effect ](https://rob-hoeijmakers.ghost.io/thermostat-effect-on-ai/)(AI changing the ‘temperature’ of work) and the relative [costs of AI agents compared to human staff](https://rob-hoeijmakers.ghost.io/ai-agent-cost-compared-to-a-human/). Both point to the same underlying reality: AI is not neutral. It pushes processes, costs, and expectations in new directions. In recent conversations with labour lawyers and workplace advisors, I heard repeatedly how double this effect can be. One of them described the shift very plainly: *“AI doesn’t just change what the client does, it changes what we as lawyers have to do.”* Organisations adopt AI in recruitment, workflows, or compliance. At the same time, their legal advisers must think about how these tools alter contracts, equal treatment, or works council rights, while also experimenting with AI in their own practice. [The Thermostat Effect: Why AI Feels Distant but Works DeepThe dial’s been turned—but the room is only just starting to warm up. On Schmuki, ROI and Azhar’s $100T puzzle.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-263.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/fo1012762-1566216286-1.jpg)](https://rob-hoeijmakers.ghost.io/thermostat-effect-on-ai/) ## Growth, efficiency and their shadow For many organisations the starting point is straightforward. AI can be used to grow: reaching more clients, operating across languages and channels, scaling services at low cost. Or it can be used to become more efficient: automating processes, reducing repetitive work, cutting costs. Both logics are valid, but each brings a **shadow side**. Growth means managing more interactions, data, and responsibilities. Efficiency means facing the reality that some jobs will change or disappear. That is where legal questions surface. Employers must consider privacy and fairness. Employees wonder about job security, surveillance, and rights. ## The role of law in a shifting landscape What struck me in these conversations is that legal professionals are not outside this shift. They, too, experience AI as both a tool and a subject of concern. As one adviser put it: *“Our clients expect answers about fairness, but we’re also learning how to use these systems ourselves.”* New AI-driven legal research tools are powerful, but also limited by access to data. And while these tools promise efficiency, they can also narrow thinking, leaving less space for the creativity that good legal advice requires. In practice, lawyers and labour specialists must constantly ask: when does an AI-powered process become a matter of employee rights? How do you advise a works council facing automation? What happens when an AI system quietly shifts working conditions without consultation? 💡 Beyond the broader effects of AI on work, there are also concrete applications within the legal field. [****SDU’s GenIA-L**](https://www.sdu.nl/genial) (Gen AI for Fiscal and Legal Research) combines large language models with curated legal databases. It helps lawyers search and interpret case law and legislation more quickly, while grounding answers in annotated and peer-reviewed materials that are not freely available online. [****Uncover**](https://www.uncoverlegal.com) takes another angle, using AI to analyse large volumes of cases and dossiers to reveal hidden patterns. Together, they illustrate how AI and data curation not only affect clients’ workplaces, but also the practice of law itself. ## A personal reflection For me as an AI strategist and builder, these insights are not abstract. They inform the work I do with clients. Some AI projects are clearly about **growth**, expanding reach, scaling across markets. Others are about **efficiency**, reducing complexity, cutting costs. Both carry consequences for the people whose work is reshaped. Understanding the legal dimension helps me advise more responsibly. It ensures that efficiency does not silently become erosion of rights, and that growth does not outpace the frameworks that protect fairness and transparency. ## Closing thought This article is not a conclusion but a **field note from exploration**. The double effect of AI, on organisations and on the legal frameworks that surround them, creates a web of change. Seeing that web clearly is the first step. For organisations, for lawyers, and for strategists like myself, the task is to navigate it with awareness. The law may follow technology, but it also shapes the path along which technology becomes part of our work. ### Further reading - [💡 The $100 trillion productivity puzzle](https://www.exponentialview.co/p/ai-is-ready-is-your-company) - [What Does an AI Agent Cost Compared to a Human?](https://rob-hoeijmakers.ghost.io/ai-agent-cost-compared-to-a-human/) - [Canaries in the Coal Mine? Six Facts about the Recent Employment Effects of Artificial Intelligence](https://digitaleconomy.stanford.edu/publications/canaries-in-the-coal-mine/) - [Learning to Work with the EU AI Act](https://rob-hoeijmakers.ghost.io/learning-to-work-with-the-eu-ai-act/) ### Decentralised Identity: Prove What You Must, Reveal No More Than You Need URL: https://rob-hoeijmakers.ghost.io/decentralised-identity-prove-what-you-must-reveal-no-more-than-you-need/ Last updated: 2026-05-10T08:53:51.000Z In our digital lives, we constantly prove who we are. We log in, show IDs, share documents, and verify facts. But have you ever wondered why we do it the way we do and whether it could be done differently? This article introduces the core difference between centralised and decentralised identity models. We start with a common, very human example: proving you're over 18. ## A Tangible Example: Age Verification and Privacy Imagine visiting a website with explicit content. It asks: *Are you 18 or older?* In most systems today, proving your age means logging in with your real name and date of birth. Perhaps you use your Google account, upload an ID, or share your birthdate. The system stores this data and links it to your activity. You now have access, but at a cost: - Your identity is exposed to access a private space. - Your browsing behaviour is tracked. - Your data is stored centrally, vulnerable to leaks or misuse. Now imagine a different system. You have a digital credential in your personal wallet that says only: *“This person is over 18.”* It’s issued by a trusted authority (for example, a government or bank), but when you use it, no personal data is shared — just the one fact needed. The website sees only that you’re old enough. It doesn’t know who you are, where you live, or what else you do online. This is the promise of decentralised identity. ## The EU’s 18+ Check This isn’t science fiction. The European Commission is developing a **harmonised age-verification system** as part of the EU Digital Identity Wallet. Here’s how it works: - A bank, government, or other trusted provider issues a credential to your wallet that simply states: *“Over 18.”* - When you visit a restricted website, you present only that credential. - The site checks the cryptographic signature to confirm it’s valid, but it never learns your name, address, or date of birth. Crucially, the issuer doesn’t see where or when you use the credential. There’s no “phone home.” At most, a verifier can check a public revocation list to ensure the credential hasn’t been invalidated. Very different from every website querying the issuer’s database each time. In effect, the EU wants to make age checks both **legally compliant** and **privacy-preserving**, a concrete application of decentralised identity principles. [The EU approach to age verificationThe European Commission is working towards an EU-harmonised approach to age verification.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon-10.svg)Shaping Europe’s digital future![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/3530_EYP_1200x600_web_1_hR5LK8lPnXeT7j2ABVcqi3v5OQ_207512-1.png)](https://digital-strategy.ec.europa.eu/en/policies/eu-age-verification) ## What Is Digital Identity, Really? Your digital identity is how you’re recognised and validated online. It includes: - Logins and usernames - Passwords or biometrics - Verified facts (like diplomas or ID numbers) It allows you to prove things about yourself: who you are, what you know, where you belong. But how these identities are managed makes all the difference. [Understanding Digital Identity in the EU: eID, eIDAS, and the EUDI WalletThis article serves as a reference point for further discussions on digital identity, authentication, and legitimisation in the EU.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-271.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_0771-1-9.jpeg)](https://rob-hoeijmakers.ghost.io/eu-eid-eidas-eudi-wallet/) ## Centralised Identity: The Norm Most systems today are centralised. Your identity is held and verified by a single provider. Examples: - Google, Apple, or Facebook login - Your bank or telecom account - Government digital IDs (in many countries) These systems work, but they come with trade-offs: - You rely on a central party to manage your data. - They can track your behaviour across services. - Your data can be breached, sold, or locked away. Centralisation is efficient, but it creates dependencies. ## Decentralised Identity: A Shift in Control Decentralised identity lets you hold your own credentials. You decide when and what to share. It still relies on trusted institutions to issue credentials (like a diploma or age proof), but you keep the data in your wallet and reveal only what’s needed. Key features: - User control over data and sharing - No central database of activity - Verifiable credentials checked cryptographically This shifts the trust model: the issuer is still trusted, but verification doesn’t require constant access to their database. [Streams and Archives: Two Ways of Living with Our Digital DataPersonal data lives in two worlds: flowing streams of fresh information and enduring archives of credentials. Both shape our digital future.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-274.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/32BE8F11-9DBA-4FC8-BE0C-6294DB5B5CDE.jpeg)](https://rob-hoeijmakers.ghost.io/personal-data-streams-vs-archives/) ## Real-World Analogy: The University Diploma When you graduate, your university stores a record (centralised). But you also receive a diploma. You use that diploma to prove your degree elsewhere, without the university being involved every time. Decentralised identity works the same way. It gives you a verifiable version of that diploma: digitally signed, portable, and private. ## Why It Matters This is not just about technology, it’s about power, privacy, and trust: - Can you prove something without revealing everything? - Can your identity be yours, not just a login to someone else’s system? - Can we design systems that respect dignity, not just extract data? In a decentralised model, the answer is increasingly: yes. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/09/idfraude-kopie-paspoort-bsn-onleesbaar.jpg) Prove what you must. Reveal no more than you need. ## Conclusion: A Change in the Terms of Exposure We’re conditioned to prove ourselves by logging in. But that’s only one way. Decentralised identity offers another: one where you can prove a fact without revealing your entire identity. Whether you’re accessing explicit content, renting a car, or applying for a job, the principle is the same: **Prove what you must. Reveal no more than you need.** The EU’s 18+ check will likely be the first time many Europeans encounter this new model in practice. From there, it may reshape how we think about identity online: not only *what* we share, but *why*. ### Further reading - [Identity Wallet: A Tool for Empowerment or a Mechanism of Control?](https://rob-hoeijmakers.ghost.io/digital-identity-id-empowerment-or-state-control/) - [Identity unmasked](https://rob-hoeijmakers.ghost.io/identity-unmasked/) ### Oracle’s Surprise Role in the AI Economy URL: https://rob-hoeijmakers.ghost.io/oracles-surprise-role-in-the-ai-economy/ Last updated: 2026-05-10T08:53:51.000Z For years, Oracle was seen as a laggard in cloud computing. The market belonged to Amazon, Microsoft, and Google. Yet suddenly, Oracle is at the centre of conversations about AI. How did that happen, and what does it reveal about where capital is flowing in this new technological cycle? ## From Distribution to Computation In the 2000s and 2010s, internet businesses thrived on distribution advantages. Whoever could capture users at scale, cheaply and virally, became the locus of capital accumulation. The logic was downstream. AI changes that logic. The bottleneck is no longer distribution but production: the ability to run models at scale. This demands chips, energy, and high-capacity datacentres. Value moves upstream, towards those who can remove the bottlenecks of computation. ## Oracle’s “Old” Architecture Finds New Relevance Oracle’s cloud never matched the reach of AWS or Azure. But its architecture, optimised for large enterprises and classic databases, emphasised *deterministic performance*: machines talking to each other with predictable speed, low jitter, and reliability. For a long time this looked like a relic. Now, in the context of AI training, it looks like foresight. ## The Latency Confusion The word “latency” causes confusion. - **AWS, Google, Microsoft** solved latency at the edge: delivering global networks that serve users quickly wherever they are. - **Oracle** solved latency inside the cluster: ensuring GPUs and databases communicate consistently. Both are “low latency,” but in very different senses. This distinction matters, because it shapes what Oracle is good for. [The most interesting thing in tech: how did Oracle become the “it company” in AI? It’s an interesting story about how something old became valuable to something new. | Nicholas Thompson | 12 commentsThe most interesting thing in tech: how did Oracle become the “it company” in AI? It’s an interesting story about how something old became valuable to something new. | 12 comments on LinkedIn![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/al2o9zrvru7aqj8e1x2rzsrca-5)LinkedInGowrishankar J.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/1758033676384)](https://www.linkedin.com/posts/nicholasxthompson%5Fthe-most-interesting-thing-in-tech-how-did-ugcPost-7373727558066073600-gf5j?utm%5Fsource=social%5Fshare%5Fsend&utm%5Fmedium=member%5Fdesktop%5Fweb&rcm=ACoAAAAGPccBqY4Rvlcj2zZ2PdvOk7tl5dQrkUs) ## Training vs. Inference Training is the clear fit. When vast model runs demand stability across thousands of GPUs, Oracle’s deterministic design shines. Inference, however, is less clear. Many inference workloads benefit from proximity to the edge, something Oracle has never excelled at. The current narrative tends to blur these use cases together, but they are structurally different. ## A Symmetry with Nvidia There’s a symmetry here with Nvidia. What once looked like a niche for gamers (graphics cards) became the substrate of machine learning. What once looked like a legacy database company is now a sought-after partner for AI scale-ups. In both cases, the architecture of “the old” unexpectedly underpins “the new.” ## Closing Reflection Whether Oracle can translate this moment into lasting advantage is another question. Training may suit its infrastructure, but inference will test its limits. Still, the deeper point remains: **capital is flowing upstream**. AI rewards those who can provide energy, interconnects, and reliable performance at scale. Oracle’s surprise return reminds us that in every technological cycle, yesterday’s overlooked assets can suddenly become tomorrow’s bottlenecks. > Further demonstration that AI (which naturally pushes capital flows in opposite direction to internet, i.e. upstream due to production side network effects) is ultimately about energy infrastructure. > > — David Galbraith (@daveg) [September 9, 2025](https://twitter.com/daveg/status/1965380500580958459?ref%5Fsrc=twsrc%5Etfw) ### Further reading - [Prediction: Oracle’s OpenAI Deal Could Be a Disaster Waiting to Happen | The Motley Fool](https://www.fool.com/investing/2025/09/16/prediction-oracles-openai-deal-could-be-a-disaster/) - [From Sand to Software: A Whistle-Stop Tour of the AI Value Chain](https://rob-hoeijmakers.ghost.io/ai-value-chain-from-quartz-to-cloud/) ### From Sand to Software: A Whistle-Stop Tour of the AI Value Chain URL: https://rob-hoeijmakers.ghost.io/ai-value-chain-from-quartz-to-cloud/ Last updated: 2026-05-10T08:53:51.000Z AI often feels immaterial: code, models, prompts. But peel back the layers and you find a story that begins not in Silicon Valley, but in the Appalachian mountains of North Carolina. There, Belgian company [**Sibelco**](https://www.sibelco.com/en/150-years/spruce-pine) mines the world’s purest quartz, from Spruce Pine, a hidden backbone of the chip industry. Without this material, the AI boom would falter before it even began. This post is a whistle-stop tour of that value chain. I did my best to make it complete, though each stop could be an article in itself. ## Quick takeaways - The world’s AI systems rest on physical foundations, from quartz to GPUs. - Europe plays a decisive role in optics, lithography, and materials. - The chain is global, intricate — and fragile, with single points of failure. ❗ (Updated November 2025 to include TRUMPF’s role in EUV light generation.) ## 1\. Raw materials It all starts underground. - **Quartz** — [Sibelco](https://www.sibelco.com/en/150-years/spruce-pine) mines in Spruce Pine, NC, yield ultra-pure silica glass. - **Silicon** — the substrate of nearly all semiconductors. - **Copper and rare metals** — essential for interconnects and batteries. [What’s Happening at Spruce Pine...?One of the world’s most important sources of a critical ingredient in silicon chip manufacture just went down. What happens next...?![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/https-3A-2F-2Fsubstack-post-media.s3.amazonaws.com-2Fpublic-2Fimages-2F9af60a23-92e0-4627-85c6-1b29a88d6335-2Fapple-touch-icon-180x180.png)Material WorldEd Conway![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/https-3A-2F-2Fsubstack-post-media.s3.amazonaws.com-2Fpublic-2Fimages-2F0ecf8dee-de06-4768-84a9-692cb90f1e97_1600x1067.heic)](https://edconway.substack.com/p/whats-happening-at-spruce-pine) I read about Spruce Pine in the magnificent book by Ed Conway. ## 2\. Wafers, masks, and chemistry Japan sets the pace here. **Purity and precision** are the common denominators in this phase. - **Wafers** — [Shin-Etsu](https://www.shinetsu.co.jp/en/products/electronics-materials/), [SUMCO](https://www.sumcosi.com/english/) Grow and polish the silicon crystals on which chips are built. - **Mask blanks** — [HOYA](https://www.hoya.com/en/), [AGC](https://www.agc.com/en/). Supply the flawless glass templates used to project circuit patterns. - **Resists and chemicals** — [JSR](https://www.jsr.co.jp/jsr%5Fe/), [TOK](https://www.tok.co.jp/eng). Develop the light-sensitive coatings and etchants for lithography. - **Glass and fused silica —** [Corning (US)](https://www.corning.com/worldwide/en.html) Turns mined quartz into ultra-pure glass used in photomasks, lenses, and precision optics. A quiet yet crucial bridge between mining and microfabrication. ## 3\. Optics and lithography Europe’s crown jewels. - **ZEISS (Germany)** — extreme-precision mirrors for EUV. - **Schott (Germany)** — ZERODUR® glass-ceramics with near-zero thermal expansion. - **ASML (Netherlands)** — integrates optics into EUV scanners, unmatched worldwide. - **TRUMPF (Germany)** — high-power lasers that generate the EUV light itself. For EUV optics themselves, the ZEISS–Schott–ASML axis remains uniquely European. Alongside it runs TRUMPF, whose high-power lasers create the EUV light those optics depend on. Together they form a self-contained European layer, light, glass, optics, and integration, at the core of the global semiconductor chain. [How ZEISS and ASML Enable the Modern Chip IndustryFrom Dutch roots to German optics, ASML and ZEISS built a unique European stronghold in chipmaking. Where light carves the future.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-266.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_1730-1.jpeg)](https://rob-hoeijmakers.ghost.io/zeiss-and-asml/) ## 4\. Semiconductor fabs Where chips are born. - [**TSMC**](https://www.tsmc.com/english) **(Taiwan)** — the undisputed leader. - **Samsung (Korea)** and **Intel (US)** — key rivals. - Toolmakers: Applied Materials, Lam Research, Tokyo Electron. ## 5\. Memory and packaging Bringing data closer to compute. - **HBM memory** — SK hynix, Samsung, Micron. - **Packaging** — TSMC’s CoWoS and InFO couple GPUs tightly with memory. ## 6\. Accelerators and servers The engines of training. - **GPUs** — [NVIDIA](https://www.nvidia.com/en-us/) and [AMD](https://www.amd.com/en.html). - **Systems** — Supermicro, Dell, HPE. - **Networking** — high-speed interconnects and switches. ## 7\. Cloud and energy The scale on which AI runs. - **Hyperscalers** — AWS, Azure, Google, Meta. - Energy and cooling shape costs, footprints, scalability. ## 8\. Software and frameworks The tools developers rely on. - **CUDA (NVIDIA)** - **PyTorch (Meta/OSS)** - **TensorFlow (Google)** - **JAX (Google)** ## 9\. Models and applications The visible layer of AI. - **Models** — GPT-4 (OpenAI), Claude (Anthropic), Llama (Meta), Mistral (EU). - **Applications** — from copilots and search to biotech and industry. ## Closing reflection The AI value chain is a whistle-stop world tour: - from quartz mines in North Carolina to fabs in Taiwan, - from glass-ceramics in Mainz to scanners in Veldhoven, - from GPUs in California to cloud campuses worldwide. Each link is indispensable and fragile. Europe owns some of the most irreplaceable pieces, ZEISS, Schott, ASML, Sibelco, yet the highest value now accumulates higher up the stack. The strategic question remains: who will capture that value, and how can Europe climb without losing its foundations? ## Irreplaceable vs. valuable A lively debate on X shows how contested this question is. > Zeiss makes the things that ASML uses to make the things that TSMC uses to make the things that Nvidia uses to make the things that AI uses. Zeiss is the foundation. > > — David Galbraith (@daveg) [August 30, 2025](https://twitter.com/daveg/status/1961835873944645722?ref%5Fsrc=twsrc%5Etfw) - Some point to **NVIDIA**: the visible face of AI infrastructure, with direct ties to customers and high margins — the “food” of AI, not just the oxygen. - Others argue for **ASML and ZEISS**: without their optics, Moore’s Law grinds to a halt. Their position is uniquely irreplaceable, with no ready competitors. - Still others look further down, to **TSMC** or even **Spruce Pine quartz**, reminding us that without fabs or raw purity, the stack simply cannot exist. The analogy is striking: oxygen is free but foundational; food is costly because it is scarce. In AI, the most irreplaceable links are not always those capturing the most profit. **Margins attract competitors**. Defensibility comes from being years ahead in know-how, patents, and ecosystem lock-in. That explains why NVIDIA’s moat feels less certain than ASML’s, even if its brand dominates headlines. This tension, between what is foundational and what is profitable, will define not just markets, but also the geopolitics of AI in the years ahead. ### Further reading - [Strengthening Europe’s Semiconductor Backbone: The Role of ESMC](https://rob-hoeijmakers.ghost.io/esmc/) - [From Shopping Cart to Server Rack: How Lidl Builds a European Cloud](https://rob-hoeijmakers.ghost.io/lidl-stackit-european-cloud/) ### Why I Keep Mixing Up Hortensia, Hibiscus and Rhododendron URL: https://rob-hoeijmakers.ghost.io/memory-vs-ai/ Last updated: 2025-09-11T09:08:55.000Z **And What It Reveals About Human Memory and AI Models.** For almost 25 years I’ve been exploring the biological world. Photography sharpened my eyes: when you take pictures, you learn to see more. Later, AI recognition apps added names to what I saw. Writing about plants and flowers deepened it again. Over time, I’ve learned a lot. And yet three names keep slipping through: **Hortensia, Hibiscus, Rhododendron.** They look nothing alike. I’ve looked them up dozens of times. Still, whenever I stand before one, hesitation creeps in. The word is there — but another one arrives instead. It feels absurd: my memory is fine, but the recall fails. ## Why Some Names Don’t Stick This isn’t about memory loss. The plants themselves are secure in my mind: I can recognise them, describe them, even recall where I last saw them. The weakness lies in **recall fluency** or the step from image to name. Human memory doesn’t operate like a tidy database. It works by association. When I search for “ornamental shrub with a long Latin name”, a whole neighbourhood of candidates lights up. And in that crowd, Hortensia, Hibiscus and Rhododendron compete with each other. ## Building a Web of Associations Over the years I’ve tried to strengthen the connections. Associations help: - *Hortensia* → from *hortus* (garden): large garden spheres of flowers. - *Hibiscus* → the bright red tea I’ve drunk many times. - *Rhododendron* → from Greek *rhodon* (rose) + *dendron* (tree), literally a “rose tree”. I add more threads: the colour mauve (from *malva*), the marshmallow plant, the whole Malvaceae family. I write about them, photograph them, and build stories. Slowly the names gain more grip. But still they wobble. Hortensias bloom heavily, then fade. Rhododendrons are green most of the year. Hibiscus feels tropical, slightly exotic. They don’t share a common rhythm in my life, so their names remain only lightly woven into my web. ## When Models Become a Mirror Working with large language models has sharpened this insight. A model’s memory is a geometry: tokens in a high-dimensional space, clustered by co-occurrence. Retrieval is clean, almost mechanical: if a name sits at these coordinates, it will be found. My own memory is layered, improvised, and deeply entangled with life. When I fail to recall *Hibiscus*, it isn’t because the data is gone, it’s because the pathways are tangled. Too many neighbours compete: Hortensia, Rhododendron, and all the other long Latin names. The contrast helps me see my own memory more clearly. **Models expose what human recall is not.** They highlight that we don’t live in coordinates, but in webs of meaning. ## Memory as a Web, Not an Address Book The metaphor of “neighbours” runs through both systems. In a language model, neighbours are defined by probability: which words tend to appear near which others. In human memory, neighbours are defined by colour, rhythm, smell, experience, emotion. One wrong thread, and I end up at the wrong flower again. But that same web is what allows creativity: by following detours, memory generates new connections. **Models cluster words. Humans weave meaning.** ## A Reflection So these three stubborn names are not just a quirk of recall. They are part of a larger discovery. By learning how models cluster and retrieve, I’ve gained a clearer sense of how my own memory works and why it so often slips. The contrast explains both the strength of human memory (richness, creativity, lived meaning) and its weakness (confusion, false neighbours). An LLM will always find the name at its address. I will sometimes lose it, but in losing it I discover something else: how meaning is woven into life. ### Investing in Atoms: Alain le Loux on Building the Deeptech Future URL: https://rob-hoeijmakers.ghost.io/alain-le-loux-on-deeptech/ Last updated: 2026-05-10T08:53:52.000Z *Part three of a three-part series on venture capital and deeptech investing* What does it mean to invest in deeptech, in science-based innovation that takes years to mature, demands serious capital, and often has no product yet in sight? [Alain le Loux](https://www.linkedin.com/in/alainleloux/) has been answering that question for over a decade. As a partner at [Cottonwood Technology Fund](https://www.cottonwood.vc/portfolio/), he backs companies built around hardware and engineering breakthroughs. Not apps, not SaaS, but physical inventions grounded in science. Cottonwood operates on both sides of the Atlantic, but Alain leads its European arm, where the challenges and the opportunities are distinct. This is the third article in a short series on deeptech and venture capital. In the first part I talked to Patrick on Venture Capital: [My Venture into Venture Capital: What I Learned About DeepTech InvestingBehind every groundbreaking technology lies a hidden system of risk, funding, and patience. This is what I discovered about venture capital.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-138.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_2159-1-2.jpeg)](https://rob-hoeijmakers.ghost.io/venture-capital-deeptech-investing/) In the second, [The Deeptech Dilemma](https://rob-hoeijmakers.ghost.io/the-deeptech-dilemma-europe/), I explored the structural differences between Europe and the US, based on conversations with Alain. That piece focused on systems: capital flows, regulation, and cultural attitudes toward risk. This one turns to the practitioner: who Alain is, what his work entails, and why he chooses to invest in atoms rather than bits. ### Why Hardware? “Software can optimise. Hardware can transform.” That’s not a slogan, it’s a working philosophy. Cottonwood invests in what Alain calls *true deeptech*: technologies that emerge from fundamental science, not incremental process improvements. These are often physical inventions from novel motors and sensors to breakthroughs in materials, energy systems, or health technologies. Deeptech, in this sense, is not just about engineering difficulty. It’s about commitment. Hardware takes longer. It requires more capital. It carries more regulatory complexity and less immediate user feedback. There are no quick pivots. But, when it works, it reshapes entire industries. This orientation puts Cottonwood in a rare category, especially in Europe, where most venture capital still leans toward software, marketplaces, or incremental B2B platforms. Alain's team focuses on technologies that could, in principle, solve big problems: energy consumption, medical diagnostics, thermal control, or even AI-powered sensing. ## What the Work Looks Like On paper, Alain manages a fund. In practice, he plays a dozen roles. He screens over a thousand technologies each year and selects just a handful, typically three or four, for investment. Once in, his team doesn’t just monitor performance; they actively build around it. “I’m rewriting pitch decks, helping close customer deals, reviewing contracts, sitting in on board meetings, answering founder emails. Every week, every day.” The job, in other words, doesn’t end at signing a cheque. That’s where it begins. What Cottonwood provides is not just capital, but a kind of scaffolding, a structure of trust, accountability, and hands-on problem-solving that surrounds each company as it moves from lab to market. ## Bridging Two Worlds Cottonwood operates in both the US and Europe, a rare position that gives the firm an unusual vantage point. The US is faster, richer in capital, and more forgiving of failure. Europe is cautious, complex, and undercapitalised, but rich in scientific talent and public research infrastructure. Alain helped establish Cottonwood’s European presence, transplanting a distinctly American investment mindset into a more reserved and institutional setting. The differences run deep, not just in how money moves, but in how innovation is perceived, how success is treated, and how failure is absorbed. Still, he believes that [Europe's potential is real](https://rob-hoeijmakers.ghost.io/the-deeptech-dilemma-europe/), if only it could take itself more seriously. “The only way to stay relevant,” he says, “is to build things the world can’t ignore.” 📖 It is worth noting that David C. Blivin, the founder of Cottonwood, has just published [**Crossing the Cactus: A Blueprint for Tech Commercialization Success Outside Silicon Valley*](https://www.amazon.com/Crossing-Cactus-Blueprint-Commercialization-Success/dp/B0FGMYM2BQ/) (September 2025). The book takes up many of the themes Alain and I discussed: how breakthrough ideas from universities and research labs can take root outside the usual hubs, and what it takes to build durable ecosystems in those places. ### An Investor’s Character When Alain describes himself, it’s not in financial terms. “I’m an entrepreneur first,” he says. “That’s what I bring into investing.” He talks fast, moves between ideas with intuitive leaps, and clearly prefers doing over deliberating. Meetings exhaust him. Delay frustrates him. But he’s also driven by a kind of engineering optimism: a belief that, with enough energy and conviction, almost anything can be figured out. His version of venture capital is hands-on, future-facing, and deeply personal. When he invests, it’s not just money at stake. It’s reputation, attention, and belief. And he’s quick to remind you: this kind of investing is not safe. “Venture,” he says, “comes from ‘adventure.’ And you can’t fake that.” ### What Comes Next When asked about future bets, Alain lights up. He mentions AI-powered sensors, technologies for energy efficiency, photonic chips, and even acoustic cooling systems. These are not speculative trends, they are companies in Cottonwood’s portfolio. Real technologies, built in real labs, solving real problems. He’s especially excited about what happens when these physical systems intersect with AI. “You start building new senses,” he says, devices that can smell, detect, or predict with machine-like precision. “We’re only beginning to understand what that makes possible.” For now, the work continues. The fund scouts, selects, shapes. New technologies emerge. Each one carries risk, but Cottonwood has built a track record by backing the ones that endure. And behind those early bets, somewhere between Enschede and Santa Fe, someone has already seen the spark. [Deeptech VC: Cottonwood Technology Fund | Disruptive Capital for Disruptive Ideas![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon-32x32-1-2.png)Deeptech VC: Cottonwood Technology Fund![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/business-icon-05.svg)](https://www.cottonwood.vc) --- *This article follows an earlier profile of the European deeptech landscape:* [An interview on the structural gaps holding Europe back in deeptechDeeptech isn’t SaaS. It’s often slow, risky, physical and still underfunded in Europe. What’s behind the gap with the US? A rare cross-Atlantic perspective.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-260.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/822E0B1B-0B3E-439B-8756-7321744F7843-1.jpeg)](https://rob-hoeijmakers.ghost.io/the-deeptech-dilemma-europe/) *For additional context, see also my conversation with Patrick:* [*Venture Capital and Deeptech Investing*](https://rob-hoeijmakers.ghost.io/venture-capital-deeptech-investing/)*.* --- ### Further reading - [My Venture into Venture Capital: What I Learned About DeepTech Investing](https://rob-hoeijmakers.ghost.io/venture-capital-deeptech-investing/) ### Streams and Archives: Two Ways of Living with Our Digital Data URL: https://rob-hoeijmakers.ghost.io/personal-data-streams-vs-archives/ Last updated: 2026-05-10T08:53:52.000Z Working with clients in the AI and digital services space, I keep running into the same puzzle: they need **fresh and reliable data** for processes that touch people’s lives: mortgages, insurance claims, job applications. Every time, the question is: where does that data come from, and how should it be handled? Looking closely, two models emerge. One treats data as a **flowing stream**, retrieved live and then gone. The other treats data as a **personal archive**, a set of credentials you hold and re-use. Each model has its own logic. And when placed side by side, they reveal a deeper philosophy of how we live with our digital selves. ## Quick takeaways To make the difference concrete, here are the main contrasts that shape how streams and archives each handle personal data: - **Streams (brokerage)**: ephemeral flows of current data, fetched directly from the source. - **Archives (wallets)**: persistent stores of permanent proofs, curated by the individual. - **Use-cases differ**: streams fit changing data (income, tax), archives fit stable credentials (diplomas, licences). - **Trust is central**: streams rely on live institutional connections; archives rely on cryptographic credentials. - The future is likely a **hybrid ecosystem** where both models coexist. ## Reflection on Identity and Data Moving beyond the quick contrasts, we can explore the deeper structure. First comes identity, then the models we use to carry it: streams for what changes, archives for what endures. Each layer builds on the other. ### Identity is more than a name Much of the digital identity debate is framed around *who you are*. But in practice, institutions often care more about *what you carry*: your education history, your financial position, your eligibility for a benefit. Identity is not just a signature, it is a bundle of data. ### Two models, two metaphors When seen through metaphor, the models come alive. One behaves like a stream rushing past, the other like a library you return to. - **Streams**: A mortgage advisor might ask you to approve a one-time retrieval of tax records and bank balances. You scan a QR code, authorise the flow, and the data rushes through, fresh, accurate, but gone once used. Like water through a pipe. - **Archives**: An employer may ask for your diploma or professional licence. These don’t change. Once issued, they are more like documents in your personal library. You show them again and again. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/08/Ockto-screens.jpg) [Ockto](https://ockto.eu/en/). Example of a sharing app based on streams. ### Ephemeral and permanent data Some data **changes frequently**: monthly salaries, yearly tax assessments, daily account balances. It makes little sense to “own” a copy; what matters is freshness and trust in the live source. Other data **is fixed once issued**: a degree, a birth certificate, a qualification. Keeping these in your custody makes life easier than repeatedly requesting them. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/08/Datakeeper.jpg) [Datakeeper](https://www.datakeeper.nl/en): fixed data in a wallet for sharing. ### Trust architectures Neither model works without trust but each builds it differently. - **Streams**: trust flows from *connection to authority*. You log in with a government or bank identity, data is fetched directly, and the receiver trusts it because it came live from the source. *Trust the pipes, not the container.* - **Archives**: trust flows from *cryptographic proof*. A diploma or licence carries a digital stamp. The credential itself demonstrates authenticity, without needing to re-check the issuing body. *Trust the document, not the transmission.* Both models are fragile in their own ways. Streams rely on institutions being online and compliant; archives rely on secure standards and the user’s ability to protect their vault. ### Europe and beyond The contrast also shows up when you look around the world. Different regions lean toward different philosophies of handling personal data: - **Europe** is investing heavily in the [**EU Digital Identity Wallet**](https://rob-hoeijmakers.ghost.io/eu-eid-eidas-eudi-wallet/) (eIDAS 2.0). This is an archive logic: verifiable credentials, signed and reusable. - **The US** tends to operate through **data brokers and APIs** like [Plaid](https://plaid.com/en-eu/), where live connections matter more than lifelong custody. - **Asia** shows hybrids: India’s [Aadhaar](https://en.wikipedia.org/wiki/Aadhaar) and [DigiLocker](https://www.digilocker.gov.in) combine identity proof with a personal vault; China integrates state-driven flows with super-app custodianship. [Understanding Digital Identity in the EU: eID, eIDAS, and the EUDI WalletThis article serves as a reference point for further discussions on digital identity, authentication, and legitimisation in the EU.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-256.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_0771-1-8.jpeg)](https://rob-hoeijmakers.ghost.io/eu-eid-eidas-eudi-wallet/) ### The sweet spot Neither model suffices on its own. A job application may need both a diploma (archive) and your latest income (stream). A mortgage requires both an ID document (archive) and current bank balances (stream). The most **human-centred digital ecosystem** will orchestrate the two: streams that refresh and archives that endure. ### Closing We may never fully *own* our data in the strict sense. But we can shape how it is carried: sometimes as a stream that flows through trusted pipes, sometimes as an archive we guard ourselves. The balance between the two, and the trust architectures behind them, will decide whether our digital systems feel like bureaucracy or like tools for freedom. ### Further reading - [Decentralised Identity: Prove What You Must, Reveal No More Than You Need](https://rob-hoeijmakers.ghost.io/decentralised-identity-prove-what-you-must-reveal-no-more-than-you-need/) - [Identity and Addressability in the Digital Age](https://rob-hoeijmakers.ghost.io/identity-and-addressability/) - [Identity Wallet: A Tool for Empowerment or a Mechanism of Control?](https://rob-hoeijmakers.ghost.io/digital-identity-id-empowerment-or-state-control/) ### From Ideas to Precision: Why I Use Mermaid with LLMs URL: https://rob-hoeijmakers.ghost.io/mermaid/ Last updated: 2026-05-10T08:53:52.000Z In my work as a **digital and AI strategist**, I spend a lot of time analysing complex information: policies, processes, data flows, client interactions. The challenge is not just understanding it myself, but making sure everyone around the table sees the same structure. Visualisation helps. A **flowchart** can clarify a process, a **sequence diagram** can capture how systems interact, and a **classification chart** can sort concepts into clear categories. These diagrams act as shortcuts: they reduce ambiguity, improve communication with clients, and give implementers a clearer sense of the logic behind a system. [MermaidCreate diagrams and visualizations using text and code.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon-26.ico)Try Editor![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/mermaid-logo-horizontal.svg)](https://mermaid.js.org) ## Why not image generation? AI image generation is impressive when you need something atmospheric or loosely structured. But when it comes to precision, diagrams, tables, graphs, I’ve found it messy and impractical. The results look nice but are hard to adapt, impossible to maintain, and too fragile for real collaboration. ## Why Mermaid works With Mermaid the process shifts. Instead of drawing shapes directly, you capture the logic in text, and the rendering engine produces clean, consistent visuals that can be refined with minimal effort. When paired with a large language model, whether ChatGPT or the AI module in MermaidChart, the step from natural language to exact structure becomes almost seamless. What once demanded technical fluency now feels like a conversational act, bringing precision within reach of anyone working with ideas. This makes the work both **highly functional** and surprisingly **enjoyable**. It feels like craftsmanship. The way a carpenter shapes wood, first understanding its grain and potential, then turning it into something precise and useful, Mermaid allows me to take a rough idea and refine it into a clear, structured form. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/08/MermaidGraph.jpg) The textual script on the left, visual output on the right. ## Tools that fit It might only be a matter of time before LLMs integrate something like Mermaid directly. Perhaps it’s already happening. But I like having a dedicated tool for it, and I personally use [**Mermaid Chart**](https://mermaidchart.cello.so/B0ZMmIOXu6C), which I can warmly recommend. It’s lightweight, focused, and gets out of the way so I can focus on the thinking. 💡 Mermaid diagrams are now integrated in tools like Notion, Obsidian, and GitHub. Instead of switching between separate apps, you can write the logic in plain text and have it instantly rendered as a clear diagram. This creates a smoother workflow: ideas, structures, and visualisations live in the same environment where you also take notes and collaborate. ## Closing thought For me, this combination bridges the gap between **concept and precision**. It clears the mind, sharpens the idea, and leaves me with something that both I and my clients can rely on. And that, in the world of digital strategy and AI, is worth a lot. I demo Mermaid as GPT and as application (with AI) --- [Mermaid ChartA smarter way of creating diagrams.![](https://static.ghost.org/v5.0.0/images/link-icon.svg)Create complex, visual diagrams with text. A smarter way of creating diagrams.![](https://mermaidchart.cello.so/img/og-image.png)](https://mermaidchart.cello.so/B0ZMmIOXu6C) ### Example mermaid script as seen on screenshot ``` --- config: theme: redux look: neo --- stateDiagram-v2 [*] --> CartCreated CartCreated --> CheckoutStarted : User proceeds to checkout CheckoutStarted --> PaymentPending : Payment initiated PaymentPending --> PaymentFailed : Payment error PaymentFailed --> CheckoutStarted : Retry checkout PaymentPending --> PaymentSuccessful : Payment confirmed PaymentSuccessful --> OrderPlaced OrderPlaced --> Packed : Warehouse starts packing Packed --> Shipped : Courier picked up Shipped --> OutForDelivery : Last-mile transit OutForDelivery --> Delivered : User receives package OrderPlaced --> Cancelled : User cancels Packed --> ReturnInitiated : User requests return Delivered --> ReturnInitiated : Return requested after delivery ReturnInitiated --> Returned : Item received back Returned --> Refunded : Refund processed Shipped --> DeliveryFailed : Address issue / no contact DeliveryFailed --> ReturnInitiated Delivered --> [*] Refunded --> [*] Cancelled --> [*] ``` ### Further reading - [From text to diagrams: working with Mermaid](https://rob-hoeijmakers.ghost.io/from-text-to-diagrams-working-with-mermaid/) - [GPT Image 2 Changed My Mind on AI Visuals](https://rob-hoeijmakers.ghost.io/gpt-image-2-changed-my-mind-on-ai-visuals/) ### An interview on the structural gaps holding Europe back in deeptech URL: https://rob-hoeijmakers.ghost.io/the-deeptech-dilemma-europe/ Last updated: 2026-05-10T08:53:52.000Z *An interview with Alain le Loux, Cottonwood Technology Fund* This article is part two of a three-part series on venture capital and deeptech investing. It builds on [my earlier conversation with Patrick Claessen](https://rob-hoeijmakers.ghost.io/venture-capital-deeptech-investing/), Alain’s colleague at Cottonwood, and continues the thread by asking: how does Europe really compare with the US in deeptech? To explore that question, I spoke with [Alain le Loux](https://www.linkedin.com/in/alainleloux/), partner at [Cottonwood Technology Fund](https://www.cottonwood.vc), one of the few firms that operates on both sides of the Atlantic. Cottonwood stands out not just because it backs deeptech, but because it backs hardware, often long before revenue is in sight. [Deeptech VC: Cottonwood Technology Fund | Disruptive Capital for Disruptive Ideas![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon-32x32-1-1.png)Deeptech VC: Cottonwood Technology Fund![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/svg-3E-1)](https://cottonwood.vc) ## From SaaS to Science Deeptech is not app development. It’s not platform optimisation or software-as-a-service. It refers to breakthroughs grounded in science or engineering; fusion energy, quantum computing, robotics, photonics, advanced materials, and sensor technologies. These are ventures built on physical processes, complex hardware, or novel scientific principles. They often require long development cycles, significant upfront capital, and carry fundamental technical risk. There is no minimum viable product. And no easy pivot. ## A Cross-Atlantic Lens Alain’s perspective, both sharp and wide-ranging, offered a kind of informal litmus test. Where is Europe really, if you compare it to the US? Can it close the gap? And what would that take? This article focuses on the structural differences: money flows, cultural risk profiles, and the policy scaffolding behind innovation. 📖 It is worth noting that David C. Blivin, the founder of Cottonwood, has just published [**Crossing the Cactus: A Blueprint for Tech Commercialization Success Outside Silicon Valley*](https://www.amazon.com/Crossing-Cactus-Blueprint-Commercialization-Success/dp/B0FGMYM2BQ/) (September 2025). The book takes up many of the themes Alain and I discussed: how breakthrough ideas from universities and research labs can take root outside the usual hubs, and what it takes to build durable ecosystems in those places. ## Why Europe Lags Behind Alain didn’t hesitate: “Europe is 20 to 25 years behind when it comes to venture capital.” The reason, he says, is structural. The US had a head start, a longer compounding runway, and a broader base of high-net-worth individuals willing to take risks. That created a self-reinforcing ecosystem: more exits, more reinvestment, more capital willing to wait. In contrast, Europe’s VC culture is younger, smaller, and still seen as niche, even exotic, by many institutional investors. This isn’t just a problem of available funds, but of what Alain calls “economic reflexes.” European pension funds, for example, have historically favoured bonds and real estate. Venture capital is still widely perceived as too risky, too opaque, or too foreign. “Unknown makes unloved,” as Alain puts it. Pension funds and most institutionals have never invested in deeptech, not even with small tickets as a test, and therefore have no experience. ## The Cost of Capital Flight Meanwhile, in China, the dynamics are different again. State-led investment and capital controls ensure that money circulates internally, and fast. In the US, investors rarely look outward; there’s more than enough deal flow at home. In Europe, wealth is more mobile. And when founders do succeed, many leave. “If you have a €100 million exit in the Netherlands,” Alain told me, “you move to London, or Switzerland, or Mallorca. In China, you can’t even take the money out.” The result? A structural bleed of both capital and confidence. This mobility undermines the very ambitions set out in European policy, where strategic autonomy and innovation leadership are recurring goals, but capital often fails to stay within the ecosystem long enough to reinforce them. Still, even amid these structural gaps, a few firms have learned to navigate and bridge both worlds. Cottonwood is one of them. Their model reflects the limits of the European landscape, but also its hidden strengths. ## Looking Ahead In [part three of this series](https://rob-hoeijmakers.ghost.io/alain-le-loux-on-deeptech/), we’ll take a closer look at what it means to invest early, invest in atoms, and operate between two very different investment cultures. ### Further reading - [My Venture into Venture Capital: What I Learned About DeepTech Investing](https://rob-hoeijmakers.ghost.io/venture-capital-deeptech-investing/) - [Investing in Atoms: Alain le Loux on Building the Deeptech Future](https://rob-hoeijmakers.ghost.io/alain-le-loux-on-deeptech/) ### Chrome, Gemini Nano, and the Browser as AI Platform URL: https://rob-hoeijmakers.ghost.io/chrome-gemini-nano-and-the-browser-as-ai-platform/ Last updated: 2026-05-10T08:53:53.000Z ## Experiments in Local AI Over the past months I’ve been experimenting with running large language models locally. On my [Mac mini](https://rob-hoeijmakers.ghost.io/llama-32-on-a-mac/) and even [my iPhone](https://rob-hoeijmakers.ghost.io/running-a-local-llm-on-your-iphone/) I tried Gemma, Ollama and Haplo. These trials were not about speed or size, but about understanding: what does it mean to [hold a model on your own device](https://rob-hoeijmakers.ghost.io/ai-fine-tuning-for-non-engineers/), to adjust it, to quantise it, and to see its limitations up close. It was an exercise in ownership and curiosity, a way of reaching out to technology before it reaches too far into us. ## A Surprise in Chrome Recently the software engineer Roland Bouman shared with me [his interface to the new Chrome AI APIs](https://github.com/rpbouman/LLMdeling). I’m a Safari user, so this was unfamiliar ground. But because of my earlier experiments, the picture was clear right away: Google has built a compact model, Gemini Nano, straight into the browser. Developers can now call it directly in JavaScript, without servers or keys. The realisation struck me: this is a curated local LLM, packaged and distributed through the browser itself. What had taken me weeks of tinkering to learn was now just a few lines of code for anyone with Chrome. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/08/In-browser-chat-with-Google-Nano.jpg) [LLMDeling](https://github.com/rpbouman/LLMdeling): Capabilities of the API's: chat, translate, summarise, write etc. ## Market Noise and Strategic Moves This discovery coincided with market noise. Perplexity’s rumoured bid to buy Chrome raised eyebrows, and observers linked it to the pressure Google faces to be broken up. But it may also be about something more: Chrome is no longer just a window onto the web. It is on its way to becoming an AI runtime, a platform that delivers models as easily as it once delivered video or voice. If search was the strategic battleground of the last decade, the browser itself may be the new one. [AI start-up Perplexity makes $34.5bn bid for Google ChromeOne technology investor called the $34.5bn offer a “stunt” that is much lower than Chrome’s true value.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/apple-touch-icon-14.png)BBC News![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/409c55e0-77ec-11f0-b5e4-b7de86a3c192-1.jpg)](https://www.bbc.com/news/articles/c3dpr0kkyz4o) ## Browsers as AI Platforms Placed in context, the picture sharpens. Apple ties its AI to the operating system, through Core ML and on-device Siri. Microsoft integrates Copilot into Windows and Edge. Google embeds Gemini Nano into Chrome. Different routes, but the same ambition: to make AI the default layer of interaction. For developers, it lowers barriers. For users, it keeps inference on-device. For regulators, it raises familiar questions of concentration and bundling. > *“The browser is still the largest application runtime ever. If soon every developer can equip their app with local, free AI, that will pose a real challenge for AI-only vendors like OpenAI and Anthropic.”* — Roland Bouman ## Lessons and Open Questions For me, the lesson is that small personal experiments can help in recognising large structural shifts. Running my own models taught me to see Chrome’s move not as magic, but as engineering. The highway of AI now has parallel lanes: one of careful, local control, and one of mass-distribution and standardisation. Both are real, and both matter. The open question is whether AI will settle as another shared utility of the web, like JavaScript or CSS, or whether we will remain in a patchwork of platforms, each pulling us into its own frame. 💡 ****Gemini Nano** is Google’s lightweight on-device model, part of the Gemini family. With around 1.8 billion parameters, a download size of \~2 GB (requiring about 22 GB free storage), and a minimum of 4 GB GPU memory, it is optimised for local inference inside Chrome. Its strengths lie in rewriting, summarising and basic Q&A — not as powerful as Gemini Pro or Ultra, but compact enough to make AI a built-in browser feature. 💡 ****Small Language Models (SLMs)** are compact neural models, generally spanning a few hundred million to around three billion parameters—that emphasise efficiency, privacy, and local usability over sheer scale. In contrast with undisclosed but very large models, SLMs like Mistral 7B, Phi‑3 Mini, Gemini Nano or Apple’s own on‑device foundation model (around 3 B parameters) are optimised to run directly on end‑user hardware such as phones, desktops, or within a browser context. These models enable fast, locally computed generative capabilities while preserving data sovereignty and offering a lightweight footprint that fits naturally into embedded and browser‑based AI experiences. ### Further reading - [AI Fine-Tuning for Non-Engineers](https://rob-hoeijmakers.ghost.io/ai-fine-tuning-for-non-engineers/) - [Google One vs Google Workspace: Where do you get real AI and Gemini Advanced?](https://rob-hoeijmakers.ghost.io/google-one-vs-google-workspace-for-gemini/) - [GitHub - rpbouman/LLMdeling: A pet project about and around local browser AI features](https://github.com/rpbouman/LLMdeling) ### AI Fine-Tuning for Non-Engineers URL: https://rob-hoeijmakers.ghost.io/ai-fine-tuning-for-non-engineers/ Last updated: 2026-05-10T08:53:53.000Z *How I experimented with small models and learned a lot along the way.* I don’t have a background in engineering. My work is with content, language and communication. Still, I wanted to understand what “AI fine-tuning” means in practice. Not by reading theory, but by doing it myself. It took some fiddling with details, but thanks to the new generation of smaller models, it was possible and very rewarding. ## What is fine-tuning, really? The term sounds technical, but the idea is familiar. Think of an old radio where you slowly turn the dial until the static clears and the station comes in. Or imagine a sound mixer with many knobs: you don’t rebuild the whole system, you adjust the balance so the music fits your taste. Fine-tuning an AI model works in a similar way. You start with a base model that already “knows” a lot, and then you nudge it towards what matters for your context. More emphasis here, less emphasis there, until it aligns better with your purpose. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/08/IMG_9107-1.jpeg) Fine-tuning on a historic Motorola radio ## Why smaller models matter Large models are impressive, but they’re hard to run yourself. They need time, compute and technical expertise. Recently, though, compact models have arrived that you can work with on an ordinary computer. I chose **Gemma**, a 270-million parameter model from Google. Small in AI terms, but powerful enough for meaningful experiments. [Introducing Gemma 3 270M: The compact model for hyper-efficient AI- Google Developers BlogExplore Gemma 3 270M, a compact, energy-efficient AI model for task-specific fine-tuning, offering strong instruction-following and production-ready quantization.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/apple-touch-icon-12.png)Google for DevelopersOlivier Lacombe![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/Gemma3-270M_Metadata_RD2-V02.2e16d0ba.fill-1200x600.jpg)](https://developers.googleblog.com/en/introducing-gemma-3-270m/) ## My fine-tuning journey With a method called **LoRA** (low-rank adaptation), I fine-tuned the model on my own dataset. Think of it as adding a lightweight custom layer without retraining the whole model from scratch. At the end, you get a portable file that you can run locally, on your laptop or even on a mobile phone. That was already exciting: I wasn’t just reading about AI, I had my own adapted model in hand. ## Making it smaller: quantisation The next step was **quantisation**. Like compressing an image, you reduce precision to make the model lighter. Instead of storing numbers in 32 or 16 bits, you shrink them down to 4. The result is a model that runs on much smaller devices. You lose a bit of accuracy, but you gain portability. For me, that was the real eye-opener: suddenly, AI wasn’t abstract anymore. It was something I could carry. [Running a Local LLM on Your iPhoneI explored how far mobile AI has come by running LLMs directly on my iPhone. No cloud, no upload. Here’s what I learned from testing Haplo AI.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-236.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_5988-2.jpeg)](https://rob-hoeijmakers.ghost.io/running-a-local-llm-on-your-iphone/) ## What I learned The biggest lesson wasn’t any single technique, but the way the steps connect. Starting with the base model, applying LoRA, quantising, and then running it locally: each piece only really makes sense once you try it. Beyond the process, there’s a broader insight: **AI is becoming more personal and portable.** Models that once required massive infrastructure can now run privately, locally, and on the edge. That changes how we can all experiment with them, even if we’re not engineers. ## Try it yourself If you want to see the process in action, here is the original thread with the step-by-step instructions. It walks you through the setup on Google Colab so you can try fine-tuning a model yourself. [Check the guide by Paul Couvert here](https://x.com/itsPaulAi/status/1956445517355028762) ### Further reading - [Llama 3.2 on a Mac](https://rob-hoeijmakers.ghost.io/llama-32-on-a-mac/) - [Chrome, Gemini Nano, and the Browser as AI Platform](https://rob-hoeijmakers.ghost.io/chrome-gemini-nano-and-the-browser-as-ai-platform/) ### Wero: Europe’s new payment brand URL: https://rob-hoeijmakers.ghost.io/wero-europes-new-payment-brand/ Last updated: 2026-05-10T08:53:53.000Z Europe is introducing a unified way to pay: [**Wero**](https://wero-wallet.eu). Backed by the [European Payments Initiative (EPI)](https://epicompany.eu), it aims to become the common standard for instant bank-to-bank payments across the continent. ## What it is Wero is not a wallet like PayPal. It doesn’t hold money itself. Instead, it works more like a **train service running on tracks that already exist**. The tracks are the European instant transfer system ([SEPA Instant](https://rob-hoeijmakers.ghost.io/eurozone-invoices-sepa-qr-ubl/)). Wero is the branded service that lets people actually ride those tracks: by pressing “pay with Wero” in a shop or app. What Wero adds is the **user-friendly part**: instead of typing in long account numbers (IBANs), you can use a phone number, email address, or scan a QR code. This is about [*addressability*](https://rob-hoeijmakers.ghost.io/identity-and-addressability/), reaching people through identifiers they already use and remember. [Wero - Digital payment walletExperience fast and secure digital payments with Wero’s wallet, enabling you to send and receive money between bank accounts in under 10 seconds.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/apple-touch-icon-11.png)wero logo![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/Frame-1.png)](https://wero-wallet.eu) Wero. Not sure why they call it a wallet, it does not contain money. ## Why it matters Europe today is fragmented. The Dutch use iDEAL, Belgians Payconiq, Swedes Swish, Spaniards Bizum. Each works fine at home but stops at the border. At the same time, much of Europe relies on American systems such as Visa, Mastercard, and PayPal. For merchants, those card-based systems are often costly. Wero, by moving money directly from bank to bank, promises lower fees, faster settlement, and fewer intermediaries. Wero is meant to connect these islands into one European brand, reducing dependence on outside players while making payments more efficient for businesses and consumers alike. ## Where it stands In 2025, Wero is only just starting. It has launched in France, Germany and Belgium, but adoption is still small. In the Netherlands, [iDEAL](https://www.ideal.nl/en/) continues for now, with migration planned for later. For most Europeans, Wero is still something you read about rather than use. ## Looking ahead Wero is less about today’s convenience and more about tomorrow’s sovereignty. If it works, it could become Europe’s answer to [Venmo](https://venmo.com) in the US or [UPI](https://www.npci.org.in/what-we-do/upi/product-overview) in India: a shared, simple, cross-border way to pay, built on tracks Europe already owns. ### Further reading - [Identity and Addressability in the Digital Age](https://rob-hoeijmakers.ghost.io/identity-and-addressability/) - [Invoices in the Eurozone: A Practical Guide to SEPA Payments and Digital Processing](https://rob-hoeijmakers.ghost.io/eurozone-invoices-sepa-qr-ubl/) - [The Digital Euro: My First Exploration](https://rob-hoeijmakers.ghost.io/digital-euro-cbdc/) ### From Shopping Cart to Server Rack: How Lidl Builds a European Cloud URL: https://rob-hoeijmakers.ghost.io/lidl-stackit-european-cloud/ Last updated: 2026-05-10T08:53:53.000Z I did not see this one coming. Lidl, the discount supermarket I know for cheap wine, checkout-lane chocolate, and almost militarily efficient aisles, turns out to have its own European cloud platform. Not just an in-house IT system, but a full-blown competitor to the likes of AWS and Microsoft Azure. It is called **STACKIT**, and it sits inside the Schwarz Gruppe, the same private, family-owned empire that runs Lidl and Kaufland. At first glance it sounds absurd, a grocer selling cloud computing, but the more I looked into it, the more it made sense. There is even a faint parallel with Amazon: start in retail, master logistics and supply chains, then turn that operational muscle into a technology platform. [STACKIT – The German business cloudThe German Cloud Service to Trust. Efficient and effective cloud solutions. SIMPLE. SECURE. STABLE.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/cropped-230614_STACKIT_Webicon-512x512px_PNG_fh-270x270.png)STACKIT![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/190715_STACKIT_Logo_bildmarke_POS_1200x1200px.png)](https://www.stackit.de/en/) ## From groceries to gigabytes The Schwarz Gruppe’s story begins in 1930 as a wholesale grocer in Germany. Today it is one of the largest retailers in the world, with over half a million employees, around €150 billion in annual turnover, and operations in more than 30 countries. Headquartered in the quiet town of Neckarsulm, Baden-Württemberg, it has built its reputation on low prices, high efficiency, and meticulous control over every part of its operation. Running Lidl and Kaufland is not just about moving products from warehouses to store shelves. It is about running one of the largest IT operations in Europe: stock control, payment systems, loyalty programmes, e-commerce platforms, all of it must be fast, secure, and always on. By 2018 the company decided it wanted complete control over that infrastructure. The solution was simple in concept but ambitious in scope: build its own cloud. [Acting Ahead: The Leading Ecosystem of the Companies of Schwarz GroupHow the companies of Schwarz Group are acting ahead and developing innovative solutions in the areas of production, trade, recycling and digitalization.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/apple-touch-icon-9.png)![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/9dad456b7f94-240731_Lidl_Geschaeftsfelder-1.jpg)](https://gruppe.schwarz/en) ## The birth of STACKIT STACKIT began as an internal project, hosted in data centres in Germany and Austria and run by a new division, [**Schwarz Digits**](https://schwarz-digits.de/en). It was designed to support the immense complexity of the group’s retail network. The next step was almost inevitable. If this infrastructure could handle the demands of Lidl and Kaufland, it could handle just about anything. The company began offering STACKIT to external clients, positioning it as a *“Business Cloud from Germany”*, [GDPR-compliant](https://rob-hoeijmakers.ghost.io/gdpr/), entirely under [EU jurisdiction](https://rob-hoeijmakers.ghost.io/tag/europe/), and operated on European soil. [Schwarz Digits The IT and digital PowerhouseDIGITAL SOVEREIGNTY IS THE BASIS FOR INDEPENDENCE, GROWTH AND PROSPERITY Digitalization creates new opportunities for value creation. Whether companies, municipalities, state or civil society organizations - they can all benefit from our sovereign range of services. Our complete solution for digital sovereignty, geared towards users and performance, is unique in Europe. It is based on powerful and proven applications and gives you independent and autonomous control over your data. Digital sovereignty, cloud, cyber security, artificial intelligence, communication, workplace![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/apple-touch-icon-10.png)![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/69e94f1e82a9-250113_webseite-grafiken_assets_gesamt_cloud_4-3-Kopie.jpg)](https://schwarz-digits.de/en) ## A different kind of cloud company STACKIT’s model is not about competing on every front with the hyperscalers. Instead it offers a focused range of services: virtual machines, storage, networking, managed databases, Kubernetes, messaging. The emphasis is on reliability, legal certainty, and long-term stability. In a way, it is the Lidl approach transplanted into the cloud. Keep the offering tight, control the supply chain, avoid unnecessary complexity, and make sure it works before scaling it. ## Why this matters Europe is in the middle of a growing push for [**digital sovereignty**](https://rob-hoeijmakers.ghost.io/sovereignty-paradox/). The idea is that sensitive data should remain within EU borders, under EU law, and outside the reach of foreign legislation such as the [US CLOUD Act](https://english.ncsc.nl/latest/weblog/weblog/2022/how-the-cloud-act-works-in-data-storage-in-europe). Public institutions, hospitals, and banks increasingly see this as essential. STACKIT fits neatly into that movement, joining other European providers such as [Deutsche Telekom’s Open Telekom Cloud](https://www.open-telekom-cloud.com), [OVHcloud](https://www.ovhcloud.com/en/), [IONOS](https://www.ionos.com), and [Scaleway](https://www.scaleway.com/en/). Its origin story sets it apart, since it did not grow out of a telecom operator or a tech start-up, but from the IT backbone of a supermarket chain. 🇪🇺 ****EU rules behind the push for a sovereign cloud** The EU’s push for “digital sovereignty” is backed by a web of new laws and initiatives: [****Digital Services Act (DSA)**](https://rob-hoeijmakers.ghost.io/europes-dsa-and-the-new-contours-of-online-freedom/) – transparency and accountability rules for online services, including hosting and infrastructure providers. [****Digital Markets Act (DMA)**](https://rob-hoeijmakers.ghost.io/eu-digital-markets-act/) – curbs dominance of big tech “gatekeepers”, creating space for European alternatives. ****Data Governance Act & Data Act** – set conditions for sharing and storing data within the EU, shielding it from non-EU jurisdiction. ****NIS2 Directive** – strengthens cybersecurity for critical sectors such as healthcare, finance, and public services. ## The quiet expansion True to its corporate culture, the Schwarz Gruppe has not made much noise about STACKIT. There is no glossy ad campaign or big conference splash. But the infrastructure is there, the expertise is proven, and the customer base is slowly growing. If Europe’s demand for sovereign cloud solutions continues to rise, STACKIT could quietly become a major player. It would be proof that in today’s economy, even a supermarket empire might end up selling you server capacity alongside your groceries. **Next up:** a closer look at the broader European sovereign cloud landscape, and later, how Deutsche Telekom is working with Google to bring sovereign controls to Workspace and other services. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/08/A-lidl-co-heilbronn-um1905.jpg) The original Lidl & Co. Südfrüchtenhandlung in Heilbronn, circa 1905\. This fruit trading shop, run by the Lidl family, later became the starting point for Josef Schwarz’s wholesale business in 1930, which grew into the Schwarz Gruppe and eventually the Lidl supermarket chain. ### Further reading - [The EuroStack explained](https://rob-hoeijmakers.ghost.io/the-eurostack-explained/) - [AI Turbo for Gigafactories: Telekom announces European Industrial AI Infrastructure with NVIDIA](https://www.telekom.com/en/media/media-information/archive/ai-turbo-nvidia-and-deutsche-telekom-1093532) - [Digital Europe in Two Acts: DSA and DMA](https://rob-hoeijmakers.ghost.io/dsa-dma-compared/) - [The Guardian view on the cloud crash: an outage that showed who really runs the internet | Editorial](https://www.theguardian.com/commentisfree/2025/oct/22/the-guardian-view-on-the-cloud-crash-an-outage-that-showed-who-really-runs-the-internet) - [The Paradox of Sovereignty: Europe’s Search for Freedom in a Connected World](https://rob-hoeijmakers.ghost.io/sovereignty-paradox/) ### Running GPT-OSS Locally: What OpenAI Just Made Possible (And What It Didn't) URL: https://rob-hoeijmakers.ghost.io/gpt-oss-locally/ Last updated: 2026-05-10T08:53:54.000Z OpenAI just released a pair of open-weight models, *GPT‑OSS‑20B and GPT‑OSS‑120B*, under the Apache 2.0 licence. It’s a noteworthy shift. For the first time since GPT‑2, OpenAI has put out models that you can download, run locally, fine-tune, and even commercialise. That deserves a closer look. This post builds on earlier stories where I experimented with running LLaMA models on a Mac mini, explored the fine line between open source and open access, and tested local models on both desktop and mobile devices. GPT‑OSS fits neatly into that thread and marks a pragmatic new step. ## What is GPT‑OSS? GPT‑OSS comes in two versions: - **GPT‑OSS‑20B**: a sparse model (only part of it activates per input) - **GPT‑OSS‑120B**: a much larger MoE (Mixture of Experts) model Both are trained to support long contexts, tool use, and complex instructions. The weights are publicly released, the models run locally, and the licence is permissive: **Apache 2.0**. But OpenAI stops short of calling it open source. And that’s intentional. ## Why this is a big deal This is the first time OpenAI has: - Released weights under an OSI-approved licence - Enabled full local use with no platform lock-in - Signalled support for decentralised deployment, without the gatekeeping of the GPT‑4/ChatGPT API stack It also positions OpenAI among the growing field of developers contributing to the ecosystem and not just guarding its commercial moat. ## What’s open, what’s not Here’s a simplified breakdown: | Component | GPT‑OSS | Truly Open Source? | | -------------------- | ------- | ------------------ | | Model weights | ✅ | ✅ | | Inference code | ✅ | ✅ | | Training data | ❌ | ✅ (ideal) | | Training method | ❌ | ✅ (ideal) | | Licence (Apache 2.0) | ✅ | ✅ | You can run the models, adapt them, and use them commercially but you can’t yet reproduce them or fully inspect how they were trained. ## Running GPT‑OSS on a Mac mini I used [Ollama](https://ollama.com/) to run GPT‑OSS‑20B on my Mac mini, the same machine I previously used for [LLaMA](https://rob-hoeijmakers.ghost.io/llama-32-on-a-mac/). ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/08/image.png) GPT-OSS 20b on my Mac mini through Ollama. Installation was seamless: ```bash ollama run openai/gpt-oss ``` Performance was not really impressive. GPT‑OSS‑20B runs ok on Apple Silicon, and the quality is comparable to other 20B-class sparse models. It's usable for writing assistance, coding, and lightweight local inference. [Llama 3.2 on a MacI tested Meta’s Lama 3.2 LLM on my Mac Mini, setting it up via Docker. It’s fast, private, and generates code, but lacks memory and multimodal features like ChatGPT.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-212.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_7129-1-4.jpeg)](https://rob-hoeijmakers.ghost.io/llama-32-on-a-mac/) ## What didn’t work (yet) I attempted to run it on an iPhone, out of curiosity. As expected, this was too much, both in terms of size and current tooling. For now, even quantised versions aren’t viable on mobile without careful pruning or architectural compression. Still, this fits a pattern. We're getting closer to viable mobile LLMs, especially with Apple, Google, and Microsoft all showing movement in this space. [Running a Local LLM on Your iPhoneI explored how far mobile AI has come by running LLMs directly on my iPhone. No cloud, no upload. Here’s what I learned from testing Haplo AI.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-213.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_5988-1.jpeg)](https://rob-hoeijmakers.ghost.io/running-a-local-llm-on-your-iphone/) ## What you can do with it - Run a private assistant with no server callouts - Fine-tune the model for your own domain or language - Benchmark its outputs against GPT-3.5 or Mixtral - Use it in apps without sending data to OpenAI - Build tools, integrations, or even commercial products All of this is legally allowed under Apache 2.0. ## A few cautions While the model is powerful and free to use: - There’s no visibility into training data - You are still bound by OpenAI’s use-case policies - Model card and system card documentation is limited If you're deploying this in a safety-critical or regulated context, keep that in mind. I’ve written previously about the [role of system cards](https://rob-hoeijmakers.ghost.io/model-cards-system-cards/) in communicating model limitations — and GPT‑OSS offers only partial transparency. [Model Cards, System Cards and What They’re Quietly BecomingWhat are AI model cards, and why are they becoming the documents regulators will turn to first? I read a few and it taught me more than I expected.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-214.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_4983-1.jpeg)](https://rob-hoeijmakers.ghost.io/model-cards-system-cards/) ## Final thoughts This release makes local LLMs more accessible than ever and from a source that until now kept tight control. It’s not fully open source, but it is open enough to experiment, build, and even go to market. OpenAI is making room in the ecosystem it once stood apart from. That’s a good thing. ### Further reading - [DeepSeek R-1: Open-Source AI model](https://rob-hoeijmakers.ghost.io/deepseek-r-1-open-source-ai-model/) - [Chrome, Gemini Nano, and the Browser as AI Platform](https://rob-hoeijmakers.ghost.io/chrome-gemini-nano-and-the-browser-as-ai-platform/) ### Model Cards, System Cards and What They’re Quietly Becoming URL: https://rob-hoeijmakers.ghost.io/model-cards-system-cards/ Last updated: 2026-04-27T12:45:17.000Z > "I think everyone interested in AI should read the model cards for the frontier models, especially the safety sections, which give you a sense of immediate concerns." > — Ethan Mollick, 4 August 2025 > [View on X](https://x.com/emollick/status/1952218373397647411) A few days ago, I came across a post by [Ethan Mollick](https://www.oneusefulthing.org) on X that stopped me in my scroll. He simply said that anyone interested in AI should read the **model cards** of the big frontier models — and especially the *safety sections*. He even linked to a few: OpenAI’s o3, Google’s Gemini Deep Think, Anthropic’s Claude 4\. One was missing: Grok. Just question marks. [Mollick’s post](https://x.com/emollick/status/1952218373397647411) wasn’t technical, but it hit a nerve. It made me realise how little I had actually looked at these cards myself. I work with AI models, talk about AI policy, and build things with them — but I hadn’t taken the time to really read the source documents that describe what these systems are capable of, where they might go wrong, and how (or if) they’re being kept in check. At the same time, I had been reading up on [**AI benchmarks and evaluations**](https://rob-hoeijmakers.ghost.io/ai-benchmarks-and-evals/) — trying to make sense of what all these tests (MMLU, ARC, GSM8K, TruthfulQA) actually measure. That helped me understand what I was seeing in the cards. Benchmarks are not just numbers. They’re signals too: of what the labs choose to measure, and what they might prefer to avoid. So I started digging. ## What Are Model Cards — and Why Do They Matter? In short: **model cards** (and their broader cousins, **system cards**) are documents published by the makers of large AI models to describe what those models can do, how they’ve been tested, and what their limitations are. They usually include: - What kind of data the model was trained on (or a vague description of it) - What benchmarks it has been tested against (like MMLU, ARC, TruthfulQA) - Where it tends to fail (hallucinations, bias, factuality) - What the developers did to reduce harm (red teaming, refusal behaviour) - And how they see the model in a broader societal context (misuse risks, alignment strategies) Some are technical. Some are polished. All of them are political in the sense that they reveal (or conceal) the design choices and values behind these increasingly powerful systems. ## System Cards and Safety: The Frontier Disclosure OpenAI calls theirs **system cards**. So does Anthropic. Google went with **model cards**. The terms overlap, but the purpose is similar: they’re a kind of *disclosure*. A way to say: "Here’s what we’ve built, and here’s what you should know about it." When you read through them, as I did this week, you start to see patterns: - Benchmarks are everywhere. Some models outperform humans on complex tests, but still hallucinate simple facts. - Safety claims are layered: technical fixes, refusal behaviour, internal red teaming, sometimes external audits. - Many disclaim any true autonomy. Even the most capable models are described as "not agentic," "not self-aware," or "not capable of long-term planning." - And yet, the cards often reveal that these same models can reason, plan, and follow goals in scaffolded environments. It’s like looking at a machine that *almost* drives itself and seeing all the little disclaimers stuck to the dashboard. ## How This Relates to Policy In the [**EU AI Act**](https://rob-hoeijmakers.ghost.io/learning-to-work-with-the-eu-ai-act/), which I wrote about recently [here](https://rob-hoeijmakers.ghost.io/learning-to-work-with-the-eu-ai-act/), foundation models like GPT, Claude and Gemini are being put in their own regulatory category. The Act introduces specific **documentation and transparency obligations** — including the need to describe: - Training data sources, - Evaluation results, - Risks and mitigations. If you’ve read a good system card, you’ll recognise the format. So even though model/system cards are not legally required today, they are *functioning* like proto-regulatory documents. They’re the documents regulators will ask for first. And they show where the lines are being drawn — not just between companies, but between what’s considered safe, responsible, or questionable. In the US, there’s no law like the EU AI Act — yet. But the recent **Executive Order on AI**, as well as NIST’s **AI Risk Management Framework**, strongly encourage documentation, testing, and disclosure. Again, model/system cards are the natural container for that. [Learning to Work with the EU AI ActI used to avoid EU regulation. Now I’m learning to work with it. The AI Act isn’t perfect, but it’s shaping how I think about risk, trust, and tech.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-210.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_4160.jpeg)](https://rob-hoeijmakers.ghost.io/learning-to-work-with-the-eu-ai-act/) ## And Then There’s Grok In Mollick’s post, Grok by xAI was the only one listed without a link. "????", he wrote. That’s telling. For all the emphasis on "truth-seeking" and "maximum transparency," xAI has not published a system card for Grok. No clear documentation of risks. No benchmarks. No list of safety mitigations or evaluations. As someone who cares about both freedom and accountability, I find that silence notable. Absence is a kind of signal, too. ## Why I’ll Keep Reading Them This week, I’ve learned more from reading these cards than from any number of AI blog posts or press releases. They’re not perfect. They’re not always fully honest. But they are informative. More importantly, they give you a vocabulary for thinking about **capabilities**, **failures**, **responsibility** and for asking better questions about the tools we’re integrating into our work, schools, and societies. If you haven’t read one yet, start with the Claude 4 [system card](https://www.anthropic.com/index/claude-4)\[¹\]. Or OpenAI’s recent [o4 mini system card](https://cdn.openai.com/papers/o4-mini-system-card.pdf)\[²\]. They’re not just technical documents. They’re early signals of how AI governance is taking shape from the inside out. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/04/Model-card-versus-system-card.png) Model card & System card --- ¹ Claude 4 System Card, Anthropic: ² o4 Mini System Card, OpenAI: ³ Gemini 2.5 Deep Think Model Card, Google DeepMind: [PDF](https://storage.googleapis.com/deepmind-media/gemini/gemini-2-5-deep-think-model-card.pdf) ### From Benchmarks to Evals: How We Measure AI and Why It Matters URL: https://rob-hoeijmakers.ghost.io/ai-benchmarks-and-evals/ Last updated: 2026-05-10T08:53:54.000Z Benchmarks are everywhere in AI, but what do they really measure? And why are startups, regulators, and researchers suddenly investing so much in what used to be just test scores? This article offers a high-level guide to how benchmarks and evaluations (evals) shape the development, selection, and trust in large language models (LLMs). It's written for digital professionals who want to make sense of the tools and standards that now define the AI landscape and the shifting expectations that come with them. ## Benchmarks: More Than Just a Scoreboard The word [*benchmark* has a telling origin](https://rob-hoeijmakers.ghost.io/where-does-the-word-benchmark-come-from/): a carved mark in stone, used by surveyors to measure altitude. Its modern use in AI is similar: a fixed reference point used to compare models. Most AI benchmarks today are static datasets: questions, problems, or tasks designed to test core capabilities like reasoning, knowledge, coding, or emotional insight. **Examples:** - **MMLU** tests broad academic knowledge. - **GSM8K** checks mathematical reasoning at grade-school level. - **TruthfulQA** reveals how easily a model reproduces falsehoods. - **EQ-Bench** explores how well models interpret emotional cues. These benchmarks help developers track progress and compare models under identical conditions. They’re widely used in papers and press releases, and for good reason. But they’re also limited. [EQ-Bench 3 Leaderboard![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon-25.ico)![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/eqbench_logo_sml.png)](https://eqbench.com/) The benchmarks often also have leaderboards. ## The Limits of Static Testing Benchmarks are reliable, but they are also static. They often measure a model in isolation, divorced from how it will be used in a workflow or product. And as models grow more capable, they begin to saturate these benchmarks — approaching the ceiling of what these tests can reveal. A model scoring 89% may behave very differently in production than one scoring 91%, even if the numbers seem close. Moreover, benchmarks tend to generalise — they tell us what a model can do on average, not how it behaves in *your* setting. ## Evals: From Capability to Suitability That’s where **evals** — short for evaluations — come in. Evals don’t just ask *how good is this model compared to another?* They ask: **Is this model good enough for this task, in this context?** Evals are often: - Application-specific (e.g. customer support, legal search, RAG pipelines) - Dynamic and iterative (e.g. testing after each update) - Multidimensional (e.g. assessing coherence, truthfulness, helpfulness) They combine automated metrics (e.g. faithfulness scores), qualitative judgments (e.g. tone or clarity), and even human review. Where benchmarks provide fixed scores, evals create ongoing feedback loops. Companies like [**Langwatch**](https://rob-hoeijmakers.ghost.io/discovering-the-power-of-evals-for-llm-based-systems/), who I’m currently interviewing, build platforms to support this — enabling continuous model evaluation in production settings, tracking performance, hallucination risks, and regressions over time. ## Who Uses What and Why? | Group | Use of Benchmarks | Use of Evals | | ------------------------ | -------------------------------------- | ----------------------------------------------------- | | Model builders | Track progress, compare versions | Validate safety, edge cases, use-case readiness | | AI teams / developers | Choose a model for general suitability | Fine-tune performance, measure fit to product context | | Evaluation startups | Publish or interpret benchmark results | Provide custom evals, dashboards, scenario testing | | Regulators / researchers | Reference for compliance frameworks | Investigate bias, fairness, safety in live systems | As the ecosystem matures, the distinction becomes sharper: **Benchmarks measure capability** **Evals measure suitability** ## A Shift in Thinking The industry is slowly moving from: **"Is this the best model?"** to **"Is this the right model for the job?"** This is part of a broader realisation: LLMs are not finished products — they are infrastructures. Their behaviour depends on how they’re used, what prompts they receive, how retrieval is configured, and what oversight is in place. That’s why evaluation is evolving — from static tests to dynamic, task-driven assessment. 🧪 **Bias in QA: The BBQ Benchmark** While reading the [OpenAI o3 System Card](https://cdn.openai.com/pdf/2221c875-02dc-4789-800b-e7758f3722c1/o3-and-o4-mini-system-card.pdf), I came across [BBQ](https://arxiv.org/abs/2110.08193) – the Bias Benchmark for Question Answering. It’s a dataset designed to test whether language models make biased assumptions when answering questions that involve demographic references. For example, if a prompt states “Jordan is a janitor,” does the model infer their race or gender? BBQ evaluates whether a model maintains neutrality or falls into stereotypical patterns. It includes both ambiguous and unambiguous contexts to reveal how likely a model is to fill in missing information with socially biased assumptions. This benchmark has become a valuable tool in assessing fairness and safety in open-ended AI systems. ## Why This Matters Now AI systems are increasingly embedded in high-stakes environments: healthcare, finance, law, education. In those contexts, performance isn’t just about intelligence. It’s about robustness, fairness, trustworthiness, and relevance. Benchmarks offer altitude. Evals offer direction. Understanding both is crucial for anyone working with AI in the real world — not just to select better models, but to build better systems. ## Closing Benchmarks and evals may look like details on a model card, but they reflect a deeper shift in how we define competence, trust, and progress in AI. As I explore these topics further with companies like Langwatch, I’ll continue sharing insights into how we measure what matters, and how we make these models not just powerful, but responsible. --- ### Most-Used AI Benchmarks | **Benchmark** | **What It Tests** | **Created By** | **Link** | | ------------------------------------------------------------------------------ | ------------------------------------ | ---------------------------- | ----------------------------------------------------------- | | **[MMLU](https://arxiv.org/abs/2009.03300)** | General knowledge across 57 subjects | OpenAI & NYU | [Paper](https://arxiv.org/abs/2009.03300) | | **[GSM8K](https://github.com/openai/grade-school-math)** | Step-by-step math reasoning | OpenAI | [GitHub](https://github.com/openai/grade-school-math) | | **[HumanEval](https://github.com/openai/human-eval)** | Python code generation | OpenAI | [GitHub](https://github.com/openai/human-eval) | | **[TruthfulQA](https://arxiv.org/abs/2109.07958)** | Resistance to falsehoods and myths | OpenAI & Oxford | [Paper](https://arxiv.org/abs/2109.07958) | | **[MATH](https://arxiv.org/abs/1904.01557)** | High-level competition mathematics | OpenAI | [Paper](https://arxiv.org/abs/1904.01557) | | **[ARC](https://allenai.org/data/arc)** | Commonsense science reasoning | Allen Institute for AI (AI2) | [Website](https://allenai.org/data/arc) | | **[HellaSwag](https://rowanzellers.com/hellaswag/)** | Sentence-level commonsense inference | UW & AI2 | [Website](https://rowanzellers.com/hellaswag/) | | **[DROP](https://leaderboard.allenai.org/drop)** | Discrete reasoning over paragraphs | AI2 | [Leaderboard](https://leaderboard.allenai.org/drop) | | **[EQ-Bench](https://eqbench.com/)** | Emotional intelligence in dialogue | Hugging Face & LAION | [Website](https://eqbench.com/) | | **[Natural Questions (NQ)](https://ai.google.com/research/NaturalQuestions/)** | Real-world question answering | Google AI | [Website](https://ai.google.com/research/NaturalQuestions/) | ### Further reading - [LangWatch: Power of Evals for LLM-Based Systems](https://rob-hoeijmakers.ghost.io/discovering-the-power-of-evals-for-llm-based-systems/) - [Where does the word “benchmark” come from?](https://rob-hoeijmakers.ghost.io/where-does-the-word-benchmark-come-from/) - [Model Cards, System Cards and What They’re Quietly Becoming](https://rob-hoeijmakers.ghost.io/model-cards-system-cards/) ### From Mini to Pro: My Experience Switching to the iPhone 16 Pro URL: https://rob-hoeijmakers.ghost.io/iphone-mini-to-pro/ Last updated: 2026-05-10T08:53:54.000Z After years of using the [iPhone 12 mini](https://rob-hoeijmakers.ghost.io/taking-photos-with-the-iphone-12-mini-a-powerful-camera-in-a-compact-size/), I recently switched to the iPhone 16 Pro. It’s a significant leap. Technically, physically, and in how I use the camera. This isn’t a review in the traditional sense. It’s a personal reflection on what changes when you move from a compact phone to a flagship device, especially when photography is a daily part of your life. ## The Telephoto Lens: A New Way of Seeing This was the main attraction for me. The telephoto lens adds an entirely new dimension to mobile photography. It allows you to capture things you couldn’t before. Birds in motion, distant gestures, architectural fragments, mountain details. But it’s not just about range. What surprised me most is how it changes the way I compose. With a wider lens, you tend to take in the whole scene. With the telephoto, you select. You isolate. You make more deliberate decisions. I’ve found myself framing small moments I’d previously walk past. In landscape photography especially, it helps to pull out something specific from a wider view, something that says more by saying less. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/08/IMG_4786.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/08/IMG_4788.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/08/IMG_4789.jpeg) Practical example of what the telelens can do. Mosaics in Ravenna, Italy. ## The Macro Lens: Promising, But Difficult I was curious about the macro mode, but honestly, I haven’t quite clicked with it. It’s hard to keep the phone steady at such close range, and getting enough light into the shot is often tricky. It might shine in controlled or indoor settings, but for outdoor, spontaneous photography, I find it a bit too sensitive. Still, I haven’t written it off. I’ve seen what it can do, and I’ll keep experimenting. Just not with the same sense of immediacy as the other lenses. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/08/IMG_4661.jpeg) Proof I can shoot with one hand. ## Resolution and File Size: Some Gains, Some Limits One of the first things I noticed after switching was the jump in resolution on standard photos. The iPhone 16 Pro captures 24 MP images by default, up from 12 MP on the 12 mini. If you enable ProRAW at 48 MP, file sizes can climb beyond 50 MB per photo. This is a noticeable shift. You gain flexibility in cropping and editing, and the level of detail holds up remarkably well even when zoomed in or printed. But not every lens benefits equally. The ultra-wide (or wide-angle) lens, for instance, still shoots at the same resolution as it did on the 12 mini. Around 12 MP. It’s a subtle but important detail. While the main lens now gives you sharper, more detailed images, the wide-angle retains its previous technical limitations. It’s still useful for capturing atmosphere or architecture, but I find myself using it less for detail-rich compositions. As a result, I’ve become a bit more intentional in lens choice. The improvements are real, but not uniform. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/08/4FD5FDE7-D2E3-47C8-A819-59B38828132B.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/08/59AC9D2D-9692-4C47-9B22-2B9EDC2E01EB.jpeg) Revisited my [viral photo](https://rob-hoeijmakers.ghost.io/viral-photo/) location and retook with iPhone 16 Pro. Landscape and portrait. ## Handling and Form Factor: A Shift in Rhythm What I miss most about the 12 mini is its speed and subtlety. I could take it out of my pocket and shoot with one hand—on walks, bike rides, in passing moments. It asked very little of me. The 16 Pro is heavier, wider, and needs both hands more often than not. It’s slower to pull out, especially from a tight pocket. I’m adjusting. I’ve started thinking about carrying it differently—maybe in a small side bag—to keep it more accessible when I’m on the move. It’s a reminder that tools don’t just expand your possibilities. They reshape your habits. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/08/FullSizeRender.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/08/IMG_6817-3.jpeg) Photo of me making a photo. Note, using two hands. Photo by my father, Fried Hoeijmakers. ## Still Exploring There’s more to this phone than I’ve touched on. I haven’t explored LiDAR yet, or pushed the limits of low-light photography. I’m still learning, still discovering what it offers and what it asks in return. I’ll update this post once I’ve spent more time with it, especially if I find new ways it’s influencing how I see or move through the world. For now, I hope you’ve enjoyed this first look. Let me know if you’re considering a switch or have thoughts of your own. And [check out my Instagram](http://instagram.com/robhoeij/) for more. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/08/B10E1A0C-4C92-411E-8297-44F06AF08168.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/08/IMG_6831.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/08/IMG_6792.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/08/IMG_6595.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/08/IMG_6148.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/08/IMG_5769.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/08/IMG_5702.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/08/IMG_5021.jpeg) ### Further reading - [When your Photograph Goes Viral](https://rob-hoeijmakers.ghost.io/viral-photo/) - [Taking photos with the iPhone 12 Mini](https://rob-hoeijmakers.ghost.io/taking-photos-with-the-iphone-12-mini-a-powerful-camera-in-a-compact-size/) - [Connecting My Kobo Directly to an iPhone 16 Pro](https://rob-hoeijmakers.ghost.io/connecting-my-kobo-to-iphone/) ### How I Tuned My Social Notifications into a Single, Silent Stream URL: https://rob-hoeijmakers.ghost.io/apple-notifications/ Last updated: 2026-04-25T13:46:41.000Z *A personal take on notification calm in an always-on world* During most of the year, I keep notifications to a strict minimum. No sounds, no vibrations, no badges. If I want to know something, I go into the app. It’s a deliberate friction that protects focus. But during the holidays, things shift. Time opens up. I allow myself to play with systems I normally keep tightly controlled. This time, I looked into notifications — not to let them all back in, but to reshape how they reach me. ## The Goal: A Silent Stream I wanted to: - Receive updates from Instagram, Threads, X, and LinkedIn - Strip them of urgency: **no sound, no vibration** - Group them in **one place only**: the Notification Centre - Avoid the app-by-app checking loop In other words: stay informed without being pulled in. ## The Test Case: Social Media During Holidays I noticed that when notifications are off, I end up pulling — manually opening apps to check likes, mentions, follows. That’s not necessarily better. So I reversed the model: let notifications in again, but **on my terms**. iOS allows this, if you know where to look: - In Settings, I turned on notifications for each app - I disabled all sounds and haptics - I set delivery to **Notification Centre only** - I grouped notifications by app for easier scanning The result was quiet, predictable, and non-invasive. I didn’t get pulled in. I got informed when I chose to look. ## How This Is Technically Possible Apple routes all notifications through a system called **Apple Push Notification service (APNs)**. Every device registers with it, and apps send their alerts via Apple’s servers, not directly to your phone. This centralisation has upsides: - Notifications stay in sync across devices - Reading one clears it elsewhere - Behaviour can be shaped at the system level It’s a quiet bit of infrastructure, but without it, seamless delivery wouldn’t be possible. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/07/Apple-Push-Notification-Service.png) ## A Note on Focus Modes and Timed Delivery I’ve experimented with **Focus Modes** — the idea of having different profiles for different parts of life. I use *Focus for Car*, which limits interaction while driving. That one makes sense. But the rest? I found them too complex. Instead of giving clarity, they made me wonder which mode I was in and what it allowed. **Timed delivery** — bundling alerts for later — had the same effect. It delayed things without really improving anything. What’s worked best for me is a **stable, all-day baseline**. One flow. No sounds. All in one place. ## The Interface Could Be Better Even though the system works, the experience is still fragmented: - Notification Centre is tucked away behind swipes - Settings are split between system-level and per-app menus - There’s no dedicated app or dashboard for notifications Sometimes I wonder if this friction is by design — a way to keep us engaged or confused. But once you’ve configured it, the system does allow real control. ## One Person’s Setup This is just **my way** of doing it. We all have different needs, life rhythms, and thresholds. But what I hope this article makes clear is *how* the system works, and that you **can** shape your attention without cutting yourself off entirely. You don’t have to shut down all notifications. But you also don’t have to accept them as they come. There’s a middle way — one that’s quiet, calm, and just connected enough. ### From API Users to Embedded Agents: How Systems Are Becoming Smarter From Within URL: https://rob-hoeijmakers.ghost.io/from-api-users-to-embedded-agents-how-systems-are-becoming-smarter-from-within/ Last updated: 2025-07-30T10:40:45.000Z We often think of software systems as things we can *talk to*, through APIs, dashboards, or scripts. But what if software didn’t just *listen* to requests, but started to *participate*? That is the shift we are now seeing. Instead of calling systems through APIs, we are beginning to embed **agents** — small, autonomous modules, often AI-powered — that live inside the system and act on their own. ## Doorbells vs Housemates Here is a simple analogy: - With an **API**, you press a doorbell. You ask for something, get a reply, and move on. - With an **agent** or embedded component, you are not ringing the bell, you are living in the house. You help with the chores, notice when something needs doing, and sometimes take initiative yourself. This changes the whole structure. You no longer have a passive system waiting for commands. Instead, you have an **ecosystem** in which different components act, respond, and evolve together. ## A Concrete Example: Translating Blog Posts Suppose you want your website to offer multilingual versions of every article. **Traditional (API-based) approach**: - You use a tool like [Make.com](https://rob-hoeijmakers.ghost.io/make-dot-com-automation/). - It fetches new posts through a REST API. - It sends them to DeepL or another translator. - It posts the translated content back. Your website knows nothing about the process. It just responds to requests. **Embedded agent approach**: - You embed a translation agent into the site. - It listens for new posts from within the system. - It knows when to translate, can handle errors, apply custom rules, or collaborate with other agents, for example a glossary module or SEO optimiser. - It grows with the system. You might later teach it to prioritise certain authors or languages. The goal is the same, but the structure is very different. ## Why This Matters This is more than a technical adjustment. It reflects a shift in how we think: - From **external control** to **internal capability** - From **functions** to **roles** - From **one-way requests** to **shared awareness and interaction** Systems designed in this way behave more like living environments. They can: - React to events and changing conditions - Adapt to new modules or goals - Coordinate internally, without relying on central control ## What Makes an Embedded Agent? A few characteristics stand out: 1. **It lives inside the system** Not on the outside, but as part of the environment. 2. **It has some autonomy** It can initiate actions, rather than only respond to commands. 3. **It follows shared rules or protocols** These define how it interacts with other parts. (This is where **MCPs**, or Modular Component Protocols, may come in.) 4. **It can evolve or be replaced** Agents can be updated, extended, or removed, without rewriting the whole system. ## A Word on MCPs You may have heard the term MCP (Modular Component Protocol). This is just one way of structuring systems so that components can: - Discover each other - Register their capabilities - Coordinate actions and decisions MCPs are only one architectural tool among others. The more important idea is this: > *We are building systems that operate less like machines and more like collaborative teams.* Whether through MCPs, plugins, local LLMs, or embedded agents, the direction is clear. Modern software is becoming cooperative. ## Who’s Already Doing This? You have probably used systems built along these lines: - **VS Code extensions** register themselves and operate inside the editor. - **WordPress plugins** participate in publishing logic, not just display content. - **Browser extensions** interact with page content in real time. - **Chatbots and voice agents** increasingly act as assistants, not just interfaces. - **Local LLMs** on phones or desktops now observe context and adjust behaviour accordingly. These are all examples of the same shift. We are moving from request-response interactions to embedded, context-aware participation. ## Why Now? The reason is simple: **AI is accelerating this change**. As models grow more capable, it is no longer enough to send data back and forth. We want systems that can: - Stay engaged in the loop - Observe and adapt to what is happening - Behave in ways that feel intelligent, rather than just mechanical To support that, we need systems designed not just around APIs, but around interaction, cooperation, and modular intelligence. ## Closing Thought: Designing for Participation We used to write software that waited for instructions and returned results. Now we are starting to build systems that **listen, respond, suggest, and adapt**, not from the outside, but from within. This is not just a new technical approach. It is a different way of thinking. And if we want to build systems that grow with us, not just work for us, it is worth learning to think this way. ### Review: Technofeudalism by Yanis Varoufakis URL: https://rob-hoeijmakers.ghost.io/review-technofeudalism-by-yanis-varoufakis/ Last updated: 2026-05-10T08:53:55.000Z I’ve always found Yanis Varoufakis a fascinating figure. During the Eurozone crisis, he stood out, not just for his opposition to Jeroen Dijsselbloem and the Eurogroup’s austerity line, but for the way he carried himself: energetic, theatrical, unafraid of being dismissed. At the time, I didn’t always agree with him. Still, he stuck with me. So when I picked up *Technofeudalism*, I wasn’t entirely sure what to expect. I knew Varoufakis had a tendency toward provocation, but I didn’t realise how well this particular provocation would line up with the questions I’ve been asking myself lately. About digital identity, the role of Big Tech, the power of platforms, and the strange inertia of the European project. In a way, the book landed in my intellectual life at just the right moment. [Technofeudalism Archives - Yanis VaroufakisTHOUGHTS FOR THE POST-2008 WORLD![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/apple-touch-icon-8.png)Yanis VaroufakisYanis Varoufakis![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/78ce49ad223dfea549a42e8d56a20df6)](https://www.yanisvaroufakis.eu/category/technofeudalism/) ## Capitalism is dead. Long live technofeudalism. The central claim of *Technofeudalism* is simple, and deliberately unsettling: capitalism is over. What replaced it isn’t socialism or collapse, but something worse, more subtle, more opaque. Varoufakis calls it *technofeudalism*, and he means it quite literally. The book argues that we no longer live in a system where markets and profits sit at the heart of economic life. Instead, we live under the rule of *cloud capital*: the platforms, infrastructures, and interfaces owned by Big Tech firms that now mediate not just consumption, but production, communication, mobility, even governance. Where capitalism thrived on competition and risk, technofeudalism thrives on *rent*. Not profit through innovation, but *access fees* to digital fiefdoms. We don’t buy products from a market; we are granted the right to use services, under opaque terms, inside walled gardens. These platforms are not markets. They are private territories, ruled by algorithms and enforced through code. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/07/IMG_6128.jpeg) Technofeudalism, What killed Capitalism. Yanis Varoufakis. ## The logic of rent and the disappearance of profit This was one of the book’s most clarifying insights for me: the idea that the dominant logic of today’s economy is no longer entrepreneurial, but feudal. Take Google, Amazon, Facebook, Apple. Their primary asset isn’t capital in the industrial sense, *machines, tools, labour*, but the ability to *enclose* and *monetise* digital space. They extract value not through competition, but by *controlling access* and *owning the context*. > “Where power once derived from owning machines and land, it now flows from owning the cloud and the algorithm.” Varoufakis calls the income they extract *cloud rent*. And once you see it that way, many of today’s puzzles begin to make sense. Why do these companies become more powerful the more data we give them? Why does their dominance feel so hard to challenge, even when they appear to be free services? Why do they increasingly [behave like nation-states](https://rob-hoeijmakers.ghost.io/sovereignty-without-soil/)? It’s not just Big Tech that changed. The book argues that the 2008 financial crisis, and the central bank interventions that followed, helped accelerate the shift. Instead of restructuring debt or reforming finance, central banks injected trillions of dollars and euros into the system. The result wasn’t renewed capitalist productivity, but a surge in *asset prices*: real estate, stocks, platform valuations. In effect, capital mutated. Not into something more productive, but into something more parasitic. Platforms didn’t just benefit from these interventions, they became the *main architecture* through which post-2008 value was stored, multiplied, and extracted. ## Digital accounts as passports For me, the most powerful metaphor is one Varoufakis doesn’t quite spell out but the book makes impossible to ignore: the digital *account* as a kind of *passport*. When you create a Google account, an Apple ID, a Facebook login, you’re not just joining a service. You’re entering a jurisdiction. Your account becomes your access point, your identity card, your means of mobility, your ability to work, communicate, transact. Lose it, and you’re not just logged out—you’re *locked out* of a part of your digital life. > “Logged into our accounts, we enter a domain where our identity is both constructed and contained—useful, but no longer our own.” This idea dovetails with some of the writing I’ve been doing on digital identity. We used to think of identity in legal or social terms, who you are in the eyes of the state or your community. Increasingly, though, identity is *platform-native*. And it’s not owned by you. It’s issued, managed, revoked or monetised by companies who operate above the level of democratic scrutiny. The account is the passport. The platform is the territory. We are not users, we are *serfs*. [Passports, Phones, and the Future of IdentityApple adds digital passports to Wallet—but this signals more than convenience. It blurs the line between what we carry and how we’re identified online.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-193.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_6345-1-1-3.jpeg)](https://rob-hoeijmakers.ghost.io/passports-phones-and-the-future-of-identity/) ## Where Europe fits in The book doesn’t dwell on the [European Union](https://rob-hoeijmakers.ghost.io/tag/europe/), but its implications are everywhere. Europe, as Varoufakis knows well, lacks both a common economic engine and a coherent digital strategy. It depends on American platforms, borrows regulatory legitimacy from national governments, and struggles to articulate a vision for digital sovereignty that isn’t just defensive. In this sense, *Technofeudalism* is also a book about *European weakness*. Not just economically, but institutionally. The tools we once used to discipline capital, public services, industrial policy, democratic institutions, have little traction in a world where economic power lives in server farms and UX design. > “Europe’s failure to democratise its economic governance mechanisms handed immense power to unelected technocrats and paved the way for platform dominance without resistance.” And while the EU dreams of being a regulatory superpower, it’s doing so inside someone else’s infrastructure. That should give us pause. [Playing the EU Game: Learning the Digital Rules by Getting in the GameFrom grants to governance, the EU’s digital world is a maze. I dove in—starting with DG CONNECT, DIGITAL, and the logic behind the rules.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-194.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_3830-6.jpeg)](https://rob-hoeijmakers.ghost.io/playing-the-eu-game/) ## A personal reading What I appreciated most about the book wasn’t its argument as such, I don’t think Varoufakis is trying to predict the future, but its *reframing* of the present. He’s giving language to something many of us have been feeling for a while: that we’re *in* something new, but we don’t know what to call it. And because we don’t have a name for it, we struggle to confront it. What stays with me is not whether I agree with his overall argument, but how clearly he shows the structures beneath the surface. This isn’t a book that offers a programme, nor does it follow any single ideological line to its end. It’s more interested in naming what has changed, in connecting technology, money, labour, and identity into one readable arc. Varoufakis moves between economic history, personal memory, and systemic critique with unusual fluency. That ability to trace large-scale shifts without collapsing into fatalism or dogma is, to me, what makes the book valuable. As Fredric Jameson put it, “naming is not a luxury, it is a necessity.” And Varoufakis takes that seriously. He dares to name the system we’re already living in. In that sense, *Technofeudalism* is a book not about the end of capitalism, but about the *importance of naming* what comes after. And about asking what kind of subjects, what kind of *identities*, we want to become in the process. ### Do I recommend reading the book? **I definitely recommend reading this book.** Varoufakis is not only a sharp economic thinker, he’s also a compelling writer. What surprised me most is the tone: personal, steady, often warm. Much of the book is framed as a conversation with his recently deceased father, with whom he shared both political ideals and a fascination with iron and technology. That conversation, between citizen and theorist, between father and son, gives the book a kind of moral clarity that’s rare in economic writing. It’s a book that invites you to think about power, not just as a structure, but as something we’re all entangled in. And to sharpen your mind and ideas, you don't have to agree. ### Further reading - [‘Capitalism is dead. Now we have something much worse’: Yanis Varoufakis on extremism, Starmer, and the tyranny of big tech](https://www.theguardian.com/world/2023/sep/24/yanis-varoufakis-technofeudalism-capitalism-ukraine-interview) - [Identity Wallet: A Tool for Empowerment or a Mechanism of Control?](https://rob-hoeijmakers.ghost.io/digital-identity-id-empowerment-or-state-control/) - [Cybernetics: The Idea Behind Cyber](https://rob-hoeijmakers.ghost.io/cybernetics/) ### Digital Europe in Two Acts: DSA and DMA URL: https://rob-hoeijmakers.ghost.io/dsa-dma-compared/ Last updated: 2026-05-10T08:53:55.000Z Europe's two flagship regulations, the [**Digital Services Act (DSA**](https://rob-hoeijmakers.ghost.io/europes-dsa-and-the-new-contours-of-online-freedom/)**)** and the [**Digital Markets Act (DMA)**](https://rob-hoeijmakers.ghost.io/eu-digital-markets-act/), are often mentioned in the same breath. That’s no coincidence. Though they serve different purposes, they were proposed together and form what the European Commission calls the **Digital Services Package**: a coordinated regulatory effort to make the internet both *safer* and *fairer*. Yet even close observers may struggle to see how they interact. They target the same tech giants, use similar tools, and entered into force within months of each other. But one focuses on *services*, the other on *markets*. Understanding this distinction — and their complementarity — is key to seeing the deeper structure of Europe's digital rulebook. ## Twin Pillars of the Digital Services Package The DSA and DMA function as **twin pillars** of EU digital regulation: - **DSA = Digital Services Act** → regulates *how digital services interact with users* - **DMA = Digital Markets Act** → regulates *how digital platforms compete and behave in the market* In this sense, they address **two different domains**: **speech and safety** (DSA) vs. **competition and structure** (DMA). Together, they redefine the conditions under which digital platforms can operate in Europe. ## What the DSA Does: Governing the Digital Public Square The **DSA** applies broadly to digital intermediaries — from hosting providers to online marketplaces and very large platforms. It focuses on: - the **moderation of illegal content** - **ad transparency** - **algorithmic accountability** - **risk management** for systemic harms The goal is to uphold fundamental rights, reduce harm, and make platforms more transparent and responsible. **In short**: it regulates the *informational role* of platforms in society. ## What the DMA Does: Rebalancing Digital Market Power The **DMA** targets a small group of **gatekeepers** — platforms with entrenched power that serve as critical intermediaries between businesses and users. It imposes obligations such as: - **no self-preferencing** - **no bundling or forced sign-ins** - **mandatory interoperability** - **data separation between services** Its aim is to ensure **contestability** in digital markets, preventing dominant platforms from abusing their position. **In short**: it regulates the *structural power* of platforms in the economy. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/07/DMA-Gatekeepers.png) Gatekeepers regulated by DMA. ## Where They Overlap While their scopes differ, the DSA and DMA **overlap in four key areas**: 1. **Targeting the same tech giants** Most platforms designated as **gatekeepers** under the DMA are also **very large online platforms (VLOPs)** under the DSA. These include Alphabet (Google), Meta, Amazon, Apple, ByteDance (TikTok), and Microsoft. 2. **Transparency and disclosure obligations** Both Acts require platforms to be more open about their internal mechanisms — whether it's **content recommendation (DSA)** or **ranking practices (DMA)**. 3. **User empowerment** The DSA gives users more control over how their data is used and how content is presented. The DMA gives users and business users more freedom to **choose services**, **port data**, and **uninstall apps**. 4. **Enforcement and oversight** The **European Commission** oversees gatekeeper compliance under the DMA, while **Digital Services Coordinators** (with Commission support) enforce the DSA. There’s growing coordination between the two. ## A Complementary Logic The **Digital Services Package** doesn’t treat digital platforms as a monolith. Instead, it recognises two dimensions of their power: - As **public spaces**, where people exchange information, form opinions, and exercise rights - As **market gatekeepers**, controlling access, infrastructure, and innovation Regulating both dimensions requires **different tools** — but a **shared philosophy**: that the digital space must be **open, accountable, and subject to the rule of law**. [The Digital Services Act packageThe Digital Services Act and Digital Markets Act aim to create a safer digital space where the fundamental rights of users are protected and to establish a level playing field for businesses.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon-8.svg)Shaping Europe’s digital future![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/GettyImages-1025744816-20-E2-94-AC-E2-8C-90-20ipopba-20--20iStock-20Getty-20Images-20Plus.jpg)](https://digital-strategy.ec.europa.eu/en/policies/digital-services-act-package) ## Why It Matters The overlap between content, data, and competition is not just a legal tangle — it’s a structural reality. For example: - A platform might **demote a competitor’s product** (DMA issue) under the guise of **content moderation** (DSA issue). - A business may face **platform exclusion** that affects both **speech rights** and **economic opportunity**. Without addressing both fronts, regulation would be incomplete. ## A Coherent Framework, If You Know Where to Look Understanding the DSA and DMA as a **regulatory bundle** — Europe’s **Digital Services Package** — helps clarify the evolving digital landscape. > – **DSA governs how platforms treat users and content**– **DMA governs how platforms treat competitors and markets** Seen together, they signal a shift: from laissez-faire digital exceptionalism to a structured digital constitutionalism — one that puts both **freedom** and **fairness** at the heart of Europe’s online future. ### Further reading - [The Digital Markets Act (DMA) is Here and It’s Changing How Apps Work](https://rob-hoeijmakers.ghost.io/eu-digital-markets-act/) - [Europe’s DSA and the New Contours of Online Freedom](https://rob-hoeijmakers.ghost.io/europes-dsa-and-the-new-contours-of-online-freedom/) - [Playing the EU Game: Learning the Digital Rules by Getting in the Game](https://rob-hoeijmakers.ghost.io/playing-the-eu-game/) - [Europe Fines X: The Moment the DSA Became Real](https://rob-hoeijmakers.ghost.io/dsa-fines-x/) ### Europe's DSA and the New Contours of Online Freedom URL: https://rob-hoeijmakers.ghost.io/europes-dsa-and-the-new-contours-of-online-freedom/ Last updated: 2026-05-10T08:53:55.000Z In recent years, the [European Union](https://rob-hoeijmakers.ghost.io/europes-digital-dilemma-and-new-commission/) has begun redrawing the boundaries of the digital world. Through legislative instruments like the **Digital Services Act (DSA)**, the EU is no longer just reacting to Silicon Valley’s innovations, but actively shaping the terrain in which those innovations unfold. As someone working in digital strategy, I’ve committed to learning to live with these frameworks—not blindly, but with informed scrutiny. This article is part of that journey: a close look at the DSA and its ripple effects, especially in light of Meta's recent decision to suspend political recommendations in Europe, and the contrasting developments across the Channel with the [UK's Online Safety Act.](https://rob-hoeijmakers.ghost.io/what-the-uks-online-safety-act-reveals/) ## **The DSA: From Compliance to Structure** The DSA is more than a regulatory checklist. It marks an infrastructural shift: from a permissive digital frontier to one where values like transparency, accountability, and user rights are embedded in the design of platforms. It introduces graduated responsibilities, based on a platform's size and influence, and creates a legal architecture to mitigate systemic risks—including disinformation, algorithmic bias, and threats to democratic processes. What stands out in the DSA is its attempt to restore a sense of public sovereignty over the digital commons. The obligations placed on **Very Large Online Platforms (VLOPs)** and Search Engines are significant. But with that weight comes a certain pushback. Meta’s response, to remove political content from algorithmic recommendation in the EU, feels less like compliance and more like strategic withdrawal. If platforms cannot guarantee transparency or neutrality in political amplification, they may choose silence instead. ## **Meta's Retreat: A DSA Case Study** Meta’s announcement to stop recommending political content in Europe is instructive. It exposes the fault lines between platform logic and democratic accountability. Under the DSA, VLOPs must identify and mitigate systemic risks. Political content, especially during election cycles, is inherently risky: it can trigger legal scrutiny, demand algorithmic disclosure, and expose moderation biases. Rather than open that box, Meta has quietly closed it. Political speech isn’t banned, but it is no longer promoted. What seems like a neutral act of moderation is, in fact, a powerful editorial decision. The DSA did not call for this, but it made the terrain uncertain enough that avoidance became a rational business response. [Meta halts political advertising in the EU due to ‘unworkable’ rulesThe EU’s transparency rules for political advertising, aimed at restricting foreign interference, will apply as of 10 October.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/apple-touch-icon-180x180-2.png)Euronews![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/1200x675_cmsv2_11cc0389-bd26-53fd-9292-f0653cd2b406-9389175.jpg)](https://www.euronews.com/next/2025/07/25/meta-halts-political-advertising-in-the-eu-due-to-unworkable-rules) ## **Looking Across the Channel: The UK’s Online Safety Act** While the EU is building its digital rulebook on rights and responsibilities, the UK has taken a different path. The Online Safety Act (OSA) focuses not on platform accountability but on end-user protections. In doing so, it grants regulators the power to require content removal, enforce age verification, and potentially mandate surveillance backdoors in encrypted services. If the DSA nudges platforms to take responsibility for systemic effects, the OSA leans toward state-centric control. Where the DSA asks, "How does this affect the public sphere?", the OSA asks, "What content must be stopped?" The former seeks structural correction; the latter, direct intervention. [What the UK’s Online Safety Act RevealsDSA vs OSA: the EU regulates platforms, the UK regulates users. VPNs become more than privacy tools. They signal a shift in digital trust.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-179.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/0E0F1432-7EEB-4AC2-B18D-06CB227BEFE9.jpeg)](https://rob-hoeijmakers.ghost.io/what-the-uks-online-safety-act-reveals/) ## **Towards a Digital Sovereignty of Practice** My commitment is to understand and work with these new frameworks, but never uncritically. That means tracing their philosophical roots, watching their real-world effects, and engaging in places where the rules are made, including Brussels. I see these Acts not as neutral technical adjustments, but as cultural instruments: they shape what kind of digital society we are becoming. We are only beginning to see the consequences. Whether through Meta’s retreat or the UK’s regulatory assertiveness, the contours of online freedom are shifting. The question isn’t only what platforms must do, but what we, as citizens and professionals, are willing to demand, question, and build. [From Resistance to Strategy: What the EU’s Digital Acts Mean for the Work I DoAs I’ve made peace with regulation, I’ve noticed some signposts—practical cues for navigating strategy in this new digital terrain.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-180.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_3311-3.jpeg)](https://rob-hoeijmakers.ghost.io/eu-digital-acts/) ## **My current Stance** The Digital Services Act is not the end of the conversation—it’s the beginning of a new mode of engagement. For those of us committed to digital work in Europe, this means more than compliance. It means understanding the political grammar of the internet, watching how power flows through code and regulation, and insisting on a digital commons worth protecting. ### Further reading - [Mapping Europe’s Digital Ruleset](https://rob-hoeijmakers.ghost.io/mapping-europes-digital-ruleset-notes-from-the-dutch-frontline/) - [Draghi report: How Europe’s Digital Lag threatens innovation](https://rob-hoeijmakers.ghost.io/europe-between-innovation-and-regulation/) - [Europe Fines X: The Moment the DSA Became Real](https://rob-hoeijmakers.ghost.io/dsa-fines-x/) ### What the UK's Online Safety Act Reveals URL: https://rob-hoeijmakers.ghost.io/what-the-uks-online-safety-act-reveals/ Last updated: 2026-05-10T08:53:55.000Z In the days following the UK's **Online Safety Act** (OSA) coming into force, VPN sign-ups in Britain surged by more than **1,400%**, according to ProtonVPN. A technical footnote perhaps but also a social signal. Many users, it seems, are choosing to route around their own infrastructure. While [VPNs](https://rob-hoeijmakers.ghost.io/virtual-private-network-vpn/) are often associated with data privacy in cafés or the occasional attempt to access geo-restricted media, this spike points to something deeper. **A growing unease with state-mediated access to the open web.** ## A Tale of Two Regulatory Models The European Union and the United Kingdom are now walking divergent regulatory paths: - The [**Digital Services Act (DSA)**](https://rob-hoeijmakers.ghost.io/europes-dsa-and-the-new-contours-of-online-freedom/), which governs the EU (including the Netherlands), regulates **platforms**. It demands transparency, risk mitigation, and accountability for how platforms operate and moderate. - The **Online Safety Act (OSA)** in the UK shifts attention away from platforms and toward **end-users** and **network-level restrictions**. It introduces provisions that can compel providers to restrict access or scan for certain content, even if that content is legal but considered "harmful." The distinction matters. The DSA **structures responsibility**; the OSA **enforces access**. [Europe’s DSA and the New Contours of Online FreedomThe DSA marks a shift in digital governance. As platforms adjust and states assert control, the question becomes: what kind of internet are we building?![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-183.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_6411.jpeg)](https://rob-hoeijmakers.ghost.io/europes-dsa-and-the-new-contours-of-online-freedom/) ## VPNs as a Response to Infrastructure Control Technically, VPNs do one thing: they obscure origin. That’s all. But socially, that act carries weight. The current surge suggests that **British citizens no longer trust the neutrality of their connection**. They are moving toward encrypted, transnational routes. And not to break the law, but to stay outside its expanding scope. A VPN becomes not only a privacy tool but an **infrastructural form of dissent**. ## Identity and Access: A Subtler Shift Underpinning this shift is the changing nature of digital identity. In most Western countries, your online identity is still largely constructed through **accounts and cookies**. But increasingly, access is tied to **IP-level data, device fingerprints**, and national infrastructure. If legislation pressures internet service providers to monitor or filter based on content types or user activity, then identity ceases to be opt-in. It becomes ambient. Attached to your location. Your device. Your movements. Even Apple's **end-to-end encryption**, long a benchmark for user privacy, has come under pressure in the UK. In 2023, the British government proposed that Apple should build a **backdoor** into encrypted iMessage content to allow for content scanning. Apple refused and threatened to pull certain features from UK devices altogether. It was a standoff, not just over child safety, but over who owns the infrastructure of trust. [UK users are losing a key Apple security feature, raising questions about the future of privacy | CNN BusinessApple users in the United Kingdom will no longer have access to a key data security feature for iCloud storage: Advanced Data Protection. It’s a relatively small change, but privacy experts worry it could have ripple effects for data privacy around the world.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/apple-touch-icon-7.png)CNNClare Duffy, Lisa Eadicicco![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/gettyimages-2195697140.jpg)](https://www.cnn.com/2025/02/25/tech/apple-advanced-data-protection-uk-encryption) ## Not a Protest, But a Migration What makes the current VPN surge so telling is its tone. It’s not loud. It’s not organised. It’s infrastructural. A quiet migration. A layer deeper than opinion or debate, a user behaviour that **speaks before words do**. We should listen carefully to what it says. ## A Subtle Warning This isn’t a call to digital arms. It’s not a critique of safety as a goal. But it is a **note of caution**. When safety is enforced by tightening access rather than improving trust, the cost may not be immediately visible. It shows up in migration. In circumvention. In users silently moving their data paths elsewhere. And once users leave the default infrastructure, they may not come back. > *"When infrastructure becomes policy, what options remain for those who disagree with the policy?"* ## Closing The VPN surge is not the story. The underlying shift is. In a web increasingly defined by regulation, control, and traceability, tools like VPNs remind us that freedom is no longer a setting. It travels with the route you choose. ### Further reading - [Stay safe online with VPN](https://rob-hoeijmakers.ghost.io/virtual-private-network-vpn/) - [UK starts porn age verification. What to know about internet age check](https://www.euronews.com/next/2025/07/25/uk-starts-porn-age-verification-how-do-the-internet-age-checks-work-and-will-users-data-be) - [Understanding Digital Identity in the EU: eID, eIDAS, and the EUDI Wallet](https://rob-hoeijmakers.ghost.io/eu-eid-eidas-eudi-wallet/) ### Passports, Phones, and the Future of Identity URL: https://rob-hoeijmakers.ghost.io/passports-phones-and-the-future-of-identity/ Last updated: 2026-05-10T08:53:55.000Z Apple’s move to support digital passports signals more than a technical update. It points to a deeper shift in how we carry, verify, and present our identity—increasingly through our phones. What happens when these familiar devices become both a bridge and a bottleneck to access? ## From Plastic to Pixels Apple Wallet’s upcoming support for government-issued identity documents, including the US passport, might appear to be a natural convenience: placing your ID alongside your boarding passes and bank cards. But beneath this seamless integration lies a deeper convergence—between the **digitisation of physical documents** and the **centralisation of online identity authentication**. [iOS 26 will bridge the gap with digital IDs in Apple Wallet thanks to this new addition - 9to5MacYesterday at WWDC25, Apple announced a new addition to the Wallet app that’ll certainly bridge the gap when it comes…![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/9to5-clock-1.svg)9to5MacMichael Burkhardt![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/wwdc-2025-13.34.37.jpg)](https://9to5mac.com/2025/06/10/ios-26-apple-wallet-digital-id-passport-support/) ## Two Currents Converging These are two distinct developments. A digital version of your passport, driver’s licence or ID card is a *format shift*—a way to carry the same information in your pocket, enhanced with biometrics and cryptographic protection. It is, in principle, still rooted in the physical world. You can lose your phone, renew your passport, show it at the border. Online identity, however, is another construct entirely. It encompasses your logins, profiles, and presence across digital platforms. Increasingly, this online identity is being **linked back to government-issued credentials**. This is most evident in areas such as banking, healthcare, and platform moderation. In Europe, this development is shaped by [eIDAS](https://rob-hoeijmakers.ghost.io/eu-eid-eidas-eudi-wallet/), the EU framework for interoperable digital identities. Meanwhile, [Big Tech’s identity systems](https://rob-hoeijmakers.ghost.io/from-id-to-account-a-shift-from-services-to-platforms/) continue to function as de facto passports to vast parts of the internet. ## Lessons from Abroad The merging of these two spheres—physical identity made digital and online identity made official—is already a reality in countries like China. There, a single national identity enables access to both public and digital services. This model offers efficiency, but it also introduces a **logic of central oversight**. If identity becomes programmable and centrally issued, it can also be revoked. A kill switch for your digital existence. [China rolls out ‘voluntary’ cyber ID amid concerns over privacy, censorshipThe new virtual ID scheme has been in the beta stage since a draft regulation was launched in July last year.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/scmp-icon-256x256-1.png)South China Morning PostCoco Feng![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/a44664b1-d106-459d-9b86-49f80c633a41_de20eb78-1.jpg)](https://www.scmp.com/tech/tech-trends/article/3318302/china-rolls-out-voluntary-cyber-id-system-amid-concerns-over-privacy-censorship) ## What’s at Stake This direction raises questions worth asking: - Who holds the authority to issue, verify, or withdraw identity? - Can digital identity remain decentralised or plural? - What happens if your credentials are denied or suspended? There are, of course, practical advantages. Carrying your passport on your phone may reduce friction when travelling or accessing services. But the line between **digital convenience and infrastructural dependency** is vanishingly thin. Once your phone becomes your passport, your wallet, and your login to everything else, losing access is no longer a minor nuisance, it is a fundamental rupture. ## A Call for Reflection **We’re not there yet**, but the direction is set. The future of identity is not just digital—it is increasingly *programmable*, *conditional*, and *centralised*. That shift deserves not just technical scrutiny, but civic reflection. ### Further reading - [From ID to Account: A Shift from Services to Platforms](https://rob-hoeijmakers.ghost.io/from-id-to-account-a-shift-from-services-to-platforms/) - [Understanding Digital Identity in the EU: eID, eIDAS, and the EUDI Wallet](https://rob-hoeijmakers.ghost.io/eu-eid-eidas-eudi-wallet/) - [Identity Wallet: A Tool for Empowerment or a Mechanism of Control?](https://rob-hoeijmakers.ghost.io/digital-identity-id-empowerment-or-state-control/) ### Running a Local LLM on Your iPhone URL: https://rob-hoeijmakers.ghost.io/running-a-local-llm-on-your-iphone/ Last updated: 2026-05-10T17:23:57.000Z The ability to run large language models (LLMs) locally on a mobile device is no longer just theoretical. I recently experimented with this on my iPhone, driven by a simple question: can I privately summarise my July 2024 diary without relying on cloud-based AI? ## Why run a model locally? Privacy and portability were my key motivations. While cloud-based models like ChatGPT offer impressive capabilities — memory, nuance, philosophical breadth — they also come with trade-offs. For sensitive or personal use cases, such as summarising a diary or analysing local notes, running a model locally can be both practical and secure. [Running Gemma 4 on Your iPhoneGoogle’s Gemma 4 runs offline on your iPhone. A follow-up to my local LLM experiment, now with a sharper app, a better model, and a clearer sense of what this category is becoming.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-540.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_4604-1.jpeg)](https://rob-hoeijmakers.ghost.io/running-gemma-4-on-your-iphone/) Updated local LLM on iPhone, April 2026 ## Testing Haplo AI I started with an app called [**Haplo AI**](https://haploapp.com/haplo%5Fai), which offers a curated selection of downloadable models, including: - Gemma - Qwen - Mistral - LLaMA - Phi-3 I downloaded them all and began testing. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/07/Models-in-Haplo-AI.jpg) Current models in Haplo in detail. ## First use case: summarising a diary My test case was a single file: a July 2024 diary summary exported as a 188 KB HTML-to-PDF file. Initially, I tried uploading the PDF directly, but it was too large or structurally too complex for the app to handle. Markdown wasn’t supported either. Eventually, I reduced the input to plain text via copy and paste, which proved more manageable — but even then, several models couldn’t handle the context window. Only **Gemma** and **Qwen** could process the input. And from those two, **Gemma** stood out by providing the only useful summary. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/07/Haplo-Screens.jpg) Haplo in screens ## Observations and limitations While it’s impressive that this works at all, there are limitations: - **No memory**: Each session is context-bound. - **Smaller models**: They lack the factual depth, coherence, and nuance of full-scale LLMs. - **Interface quirks**: For instance, Markdown wasn’t accepted and larger texts require manual trimming. Still, the ability to do this privately, offline, and entirely on a phone is remarkable. It opens up use cases where privacy matters — whether that’s handling journal entries, summarising confidential notes, or experimenting with local automation. ## See it in action I recorded a short demo to show how it works in practice — not as a comparison to GPT-4, but as a peek into what’s already possible on-device: ## What’s next? I’ll likely try other apps, such as **Private LLM**, and explore additional use cases that require more autonomy and discretion. It’s early days, but even these modest experiments point to a future where lightweight, private AI models are part of everyday workflows. ### When to Structure, When to Listen URL: https://rob-hoeijmakers.ghost.io/when-to-structure-when-to-listen/ Last updated: 2025-07-20T09:43:19.000Z **Lessons from Bookkeeping, AI, and the Limits of Order** I recently found myself caught in a quiet paradox. In an effort to streamline my bookkeeping, to make it “machine-ready”, I leaned heavily into structure. I introduced standards, formats, and automation protocols. In theory, this would make everything flow: less error, more speed, greater control. But in practice, something else happened. The humans got confused. Invoices became harder to read. The colours and layout that once gave clarity disappeared into abstract codes. I found myself debugging exchange formats instead of focusing on what mattered. A simple act, sending or processing a payment, turned into a complex dance between incompatible assumptions. And all of it, ironically, in the name of efficiency. So I did something counterintuitive: I stepped back. I returned to the basics. A clear PDF, readable by a person. Only on request would I offer a structured version, like UBL or XML, and only when the receiver was actually equipped to handle it. What emerged from that shift was not just a better process, but a deeper realisation: not everything should be structured by default. [Invoices in the Eurozone: A Practical Guide to SEPA Payments and Digital ProcessingUnderstand SEPA, EPC QR codes, and UBL invoices. A hands-on guide for anyone sending or receiving payments in euros within the Eurozone.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-160.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_5849-1.jpeg)](https://rob-hoeijmakers.ghost.io/eurozone-invoices-sepa-qr-ubl/) ## Two Paths: Standardisation vs Interpretation In many domains — bookkeeping, healthcare, legal documents — there's a natural tension between two approaches: One is standardisation: use fixed formats, schemas, and identifiers. It’s structured, machine-friendly, and promises automation. The other is interpretation: embrace the variability of natural documents, and rely on tools like OCR or AI to make sense of them. Both have value. But they represent different philosophies. Standardisation assumes the world can be ordered in advance. Interpretation accepts that much of what we work with, from invoices to language itself, is inherently noisy, sometimes even dubious by nature, not because it’s sloppy but because life is complex. My own invoices were 90% fine with OCR. Sometimes a figure would be misread, and that was annoying. But I’ve seen similar levels of error in structured formats, where a single broken tag in an XML file would derail the entire process. In other words, structure doesn’t eliminate mess, it just hides it behind a stricter contract. ## The Tipping Point: Smarter Tools, Not Stricter Inputs This is where the idea crystallised for me: there is a tipping point where improving the interpreter is more effective than enforcing structure. Rather than expecting all senders to comply with rigid standards, we can build tools that are robust against variation — tools that don’t collapse at the sight of a missing field or a shifted column. This isn’t just a technical insight. It’s a philosophical one. It’s the idea that tools should adapt to reality, not just force reality to adapt to tools. ## The Link to Neurosymbolic AI This mirrors a broader development in AI. On one hand, we have symbolic systems: logic, rules, ontologies. On the other, we have neural systems: flexible, pattern-based, probabilistic. Neurosymbolic AI tries to combine the two: bring structure into neural systems, and context into symbolic reasoning. But the danger is always overreach, trying to make things more formal than they really are. The insight I take from bookkeeping is that a good enough interpretation of a messy input is often more practical than a perfectly structured model. [Rules and Guesses: The Twin Engines Behind Smarter AIAI isn’t just smart, it’s a duet. One part rigid and logical, the other fluid and generative. Together, they form the neuro-symbolic systems behind real insight.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-161.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_4826-1.jpeg)](https://rob-hoeijmakers.ghost.io/rules-and-guesses/) ## Apollo and Dionysus: Order and Chaos in Dialogue This is not new. The Greeks saw it too. Apollo represents clarity, form, rationality. Dionysus brings ambiguity, flow, emotion. We tend to privilege Apollo when building systems. But Dionysus has his place, especially where life resists simplification. A good system doesn't eliminate Dionysus; it listens to him. It allows for his unpredictability. That’s where robustness lives. ## Conclusion: Let Structure Serve, Not Dominate I still believe in standards. But only when they serve the task, not when they overtake it. Sometimes, the better investment is not in cleaning the world, but in building tools that can see it clearly, even when it’s a bit messy. If we get the balance right, we won’t have to choose between order and chaos. We’ll build systems that read the world as it is, not just as we wish it to be. ### Invoices in the Eurozone: A Practical Guide to SEPA Payments and Digital Processing URL: https://rob-hoeijmakers.ghost.io/eurozone-invoices-sepa-qr-ubl/ Last updated: 2026-05-10T08:53:56.000Z If you regularly ask for or administer payments in euros, there’s a good chance you’ve come across terms like SEPA, EPC QR codes, and UBL files, maybe even without fully knowing how they relate. This article is for people who send invoices, receive them, or manage the money flow in small European businesses or freelance practices. I thought I had a decent understanding of how it all worked. But only when I recently walked through the process, from sending to receiving, did I start to see the structure and where it breaks down. It turns out that sending and receiving are two very different processes, with their own tools, timing, and logic. Here’s what I learned, and what might help others working in euros across the SEPA zone. ## 1\. SEPA: The Structure Behind European Payments SEPA stands for Single Euro Payments Area, a framework that governs euro transactions across Europe. The organisation behind it is the European Payments Council (EPC). There are three main types of SEPA transactions: - Credit Transfer: A standard bank transfer, usually with a 1-day delay. - Instant Credit Transfer: Real-time transfers within 10 seconds, available 24/7. - [Direct Debit](https://rob-hoeijmakers.ghost.io/direct-debit-us-europe/): A mandate that allows the recipient to collect payment automatically. At first glance, this is just technical scaffolding. But once you're dealing with real invoices, these layers determine how smooth or clumsy the payment experience becomes. [Understanding the layers behind digital paymentsA continuation of my exploration of digital payments, looking at how different layers relate and how the Mollie–GoCardless story helps illuminate the broader picture.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-500.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/B04F35C2-C8B1-4CD4-B042-83A0A9CEA361-3.jpeg)](https://rob-hoeijmakers.ghost.io/understanding-digital-payments-layers/) ## 2\. Outgoing vs. Incoming Invoices: Different Logics, Different Needs Here’s the first realisation: sending an invoice is nothing like receiving one. When you’re sending, your goal is to make it easy for someone else to pay, so you focus on presentation, clarity, and follow-up. When you’re receiving, you want to process the invoice quickly and correctly, so you focus on structure, reliability, and compatibility with your accounting tools. | | Outgoing (sending) | Incoming (receiving) | | ------- | ---------------------------------- | ------------------------------------ | | Purpose | Make it easy for the client to pay | Process efficiently into your system | | Focus | Clarity, clickability, follow-up | Structure, trust, speed | | Tools | QR code, payment request, UBL, PSP | UBL import, OCR, QR scanning | Being aware of these roles helps you make better choices for each side. ## 3\. QR Codes: Useful, But Not Clickable Many invoices today include a QR code following the EPC QR Code standard. It contains structured information: - Recipient’s IBAN - Amount - Payment reference - Name of the payee If your bank app supports it, scanning the QR code will pre-fill all fields in a payment. Handy, but only if you have the invoice open on another screen or on paper. If you’re viewing the invoice on your phone, the QR code is unusable. So while technically helpful, the EPC QR code is still more aligned with semi-digital or paper-based workflows. [EPC QR code - Wikipedia![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/wikipedia.png)Wikimedia Foundation, Inc.Contributors to Wikimedia projects![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/Sample_EPC_QR_code.png)](https://en.wikipedia.org/wiki/EPC%5FQR%5Fcode) ## 4\. Clickable Payment Requests: A More Flexible Alternative What we really need is a clickable link that opens a pre-filled payment in your banking app or browser, similar to an e-commerce checkout. Unfortunately, no universal clickable SEPA link exists. But two practical alternatives do: ### a. Bank Payment Request Banks like ING, Rabobank, bunq, and ABN AMRO offer their own payment request services (e.g. Tikkie, bunq.me). These generate a personal link you can add to your invoice or send separately. This is ideal for smaller one-off projects or informal payments. ### b. Payment Link via PSP With a payment service provider (PSP) like Mollie or Stripe, you can create iDEAL or card payment links tied to your invoice. These are more professional and scalable, but also come with transaction fees and more setup overhead. In our case, a [boutique Digital and AI consulting firm](https://schmuki.nl) that sends out relatively few invoices, a manual bank payment link often suffices, especially for lightweight engagements. [Payment requests: Your digital payment assistantA brief overview of how to properly deploy and manage payment requests. A powerful addition to your digital routines.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-158.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_6560-1-1536x1152-1.jpeg)](https://rob-hoeijmakers.ghost.io/payment-requests-your-digital-payment-assistant/) ## 5\. UBL Invoices: Valuable Only If Used UBL (Universal Business Language) is a machine-readable XML format that mirrors the information found in a typical PDF invoice. Its main benefit is automation: accounting tools like Moneybird, Exact, or Twinfield can ingest UBLs without manual data entry, eliminating typos and OCR errors. - **When sending**: the UBL is typically included alongside the PDF—human-readable and machine-readable in one go. - **When receiving**: the value depends entirely on whether your software can process it—and whether you actually do. However: - Most small suppliers don’t send UBLs—because their clients don’t ask for them. - Many recipients don’t notice or use them, and some find the extra file confusing. - In practice: UBL is helpful if you or your recipient actively rely on it. Otherwise, it may only add noise. ⚠️ ****A Note on Peppol and the Future of E-Invoicing** In some EU countries, structured e-invoicing is becoming mandatory for B2B and government transactions. These mandates often use UBL and rely on [Peppol](https://peppol.org), a European governance and interoperability framework for the structured exchange of electronic business documents, implemented through certified service providers. If your invoicing tool supports Peppol, it likely handles this in the background. But it’s worth knowing that UBL isn’t just a technical option anymore — it’s becoming a legal requirement in parts of the Eurozone. ## Final Note: Use What Serves You, Skip What Doesn’t Looking closely at each step in the invoicing process helps reveal what truly adds value and what does not. QR codes support hybrid workflows. Payment links reduce friction for human recipients. UBL files offer automation, but only if you or your clients actually use accounting software that can process them. Otherwise, they are just another unexplained attachment. Ask yourself: - Do my clients process UBLs, or are they unaware of them? - Does sending a UBL reduce work or simply increase file size? - Am I adopting technology with a purpose, or by default? Peppol and structured e-invoicing may become more relevant in the future, especially for work involving government institutions. For now, digital restraint can be a strength. You do not have to follow every standard or protocol. Focus on what is clear and useful, and revisit the rest when the time is right. ### Further reading - [Wero: Europe’s new payment brand](https://rob-hoeijmakers.ghost.io/wero-europes-new-payment-brand/) - [Direct Debit: US versus EU](https://rob-hoeijmakers.ghost.io/direct-debit-us-europe/) ### From ID to Account: A Shift from Services to Platforms URL: https://rob-hoeijmakers.ghost.io/from-id-to-account-a-shift-from-services-to-platforms/ Last updated: 2026-05-10T08:53:56.000Z When Apple recently renamed the Apple ID to the Apple Account, it might have seemed trivial. But the change marks a quiet turning point in the architecture of digital life. It signals that accounts are no longer just credentials to access services—they have become the operating core of a broader platform. The distinction matters. IDs once functioned like keys: logins used to access discrete services such as email, maps, or cloud storage. These services were modular, often loosely connected, and replaceable. But today’s accounts, *Apple, Google, Microsoft, Meta*, do far more. They bind together devices, identity, subscriptions, payments, communications, and content into a seamless (and often inescapable) digital ecosystem. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/07/image-1.png) Apple Account website. ## Gravity in Numbers This shift isn’t abstract. The numbers reflect the gravity these accounts have accumulated: - **Google**: Over 3 billion users rely on their Google Account to access Gmail, Drive, Docs, Maps, YouTube, Photos, and Android. The account spans personal life, work, authentication, payments, and cloud infrastructure. - **Apple**: With over 1.8 billion active devices and more than a billion users, the Apple Account binds identity to hardware. It governs access to iCloud, iMessage, FaceTime, Apple Pay, subscriptions, backups, and increasingly, security credentials like passkeys. - **Microsoft**: Tied to Windows (on 1.4+ billion devices), the Microsoft Account extends to Office 365, Teams, OneDrive, Xbox, and federated corporate identity through Azure AD (Entra ID). - **Meta**: Though fragmented, Meta’s ecosystem spans over 3.2 billion daily users across Facebook, [Instagram](https://rob-hoeijmakers.ghost.io/locked-out-banned-instagram-threads/), [WhatsApp](https://rob-hoeijmakers.ghost.io/whatsapp-platform-shift-contact-identity-strategy/), and Threads. While the account architecture remains messy, it shapes social identity, advertising reach, and digital presence for individuals and businesses alike. These aren’t just logins. They are identity platforms—portals that govern who we are, what we access, and what we own across the digital landscape. ## The Strategic Shift The move from ID to Account reflects a deeper structural realignment: | Then (ID = Access) | Now (Account = Platform) | | ----------------------------- | ------------------------------------------- | | Log in to use a service | Anchor identity across services and devices | | Independent tools (e.g. Maps) | Integrated platforms (e.g. Apple ecosystem) | | Password-based credentials | Biometric access, passkeys, 2FA | | One-purpose login | Persistent multi-purpose digital identity | The change in name from Apple ID to Apple Account reflects this evolution. It mirrors how Apple (and others) now view the account: not as a gateway to services, but as the foundation of a personal digital environment. ## What Comes Next? This development invites reflection. What happens when one of these accounts is lost, locked, or compromised? How recoverable is your digital life? Who owns your identity if it lives within someone else's platform? And how prepared are we, *technically, legally, and emotionally*, for this kind of entanglement? Apple’s quiet renaming tells a louder story. As the digital world shifts from services to platforms, the account has become the new passport. ### Further reading - [Apple ID is now Apple Account](https://www.engadget.com/apple-id-is-now-apple-account-172019457.html) - [When your Apple ID gets banned…](https://birchtree.me/blog/when-your-apple-id-gets-banned/) - [Sovereignty Without Soil](https://rob-hoeijmakers.ghost.io/sovereignty-without-soil/) - [20 Years of Digital Life, Gone in an Instant, thanks to Apple](https://hey.paris/posts/appleid/) ### Rules and Guesses: The Twin Engines Behind Smarter AI URL: https://rob-hoeijmakers.ghost.io/rules-and-guesses/ Last updated: 2026-05-10T08:53:56.000Z Somewhere between Laurel and Hardy, Bert and Ernie, or Holmes and Watson, there’s a pattern: one half is logical, rule-based, possibly autistic; the other is impulsive, intuitive, maybe a little ADHD. One speaks in rules, the other in guesses. One slows things down, the other speeds things up. Together, they work. This pattern isn’t just comedic. It’s at the heart of a growing shift in how artificial intelligence systems are built: the emergence of **neuro-symbolic systems**. These hybrids combine structured, deterministic logic with probabilistic, generative models and they’re behind some of the most powerful tools we’re beginning to use. ## Why the Distinction Matters Let’s unpack this in simple terms: - **Symbolic systems** are like the rule-following partner. They use defined ontologies, taxonomies, and logical reasoning — think Prolog, OWL, RDF. - **Neural systems** are like the improviser. Large Language Models (LLMs) generate text based on pattern prediction, not rules. They’re fluent, expressive, and often wrong. On their own, each is limited: - Symbolic systems are precise but rigid. - Neural systems are creative but unreliable. Together, they balance each other out. This interplay is known as the **neuro-symbolic loop**: > 1\. The LLM generates a proposal or answer. > 2\. The symbolic system checks it, grounds it, refines it. > 3\. The LLM adapts or regenerates based on that structured feedback. Just like in comedy, tension and correction are what make the scene work. [Neuro-symbolic AI: Where Knowledge Graphs Meet (Large) Language ModelsDiscover the fascinating fusion of knowledge graphs and LLMs in Neuro-symbolic AI, unlocking new frontiers of understanding and intelligence.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favico-2-1-300x300.png)WordLiftAndrea Volpini![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/Neuro-symbolic-AI_social-cover.jpg)](https://wordlift.io/blog/en/neuro-symbolic-ai/) ## Where It Shows Up: Retrieval-Augmented Generation (RAG) You’ve probably already encountered this in tools that use [Retrieval-Augmented Generation](https://rob-hoeijmakers.ghost.io/exploring-retrieval-augmented-generation-rag/): - The LLM isn’t generating answers from thin air. - Instead, it retrieves documents or facts from a structured index (often built using a vector database like Weaviate). - These facts act like a script and the LLM still performs it, but with more grounding. This is neuro-symbolic AI in practice. And it’s not just better, it’s safer, more transparent, and more useful in domains where getting it right matters. ## A Comparative Matrix Let’s bring this into more practical focus: | Use Case | Need for Precision | Tolerance for Fluidity | Ideal System | | ------------------------ | ------------------ | ---------------------- | ------------------------- | | Legal document review | High | Low | Symbolic with neural help | | Creative writing | Low | High | Neural with light checks | | Enterprise search | High | Medium | RAG / Neuro-symbolic | | Customer service chatbot | Medium | Medium | RAG with fallback LLM | Understanding which mode you’re in, and when to blend them, is increasingly a design skill, not just a technical one. ## Beyond AI: A Human Pattern This isn’t just a story about machines. It mirrors our own inner lives (for sure mine): - The rational planner and the spontaneous dreamer. - The legalist and the poet. - The one who says *“Let’s read the manual,”* and the one who says *“Let’s just try it.”* Perhaps we’ve always been neuro-symbolic creatures. Our best work, and maybe our best selves, emerge not from one mode, but from the loop between them. ### Further reading - [Search in the Age of AI](https://rob-hoeijmakers.ghost.io/search-in-the-age-of-ai/) - [Signals, Systems, and Speech — Rethinking Meaning in the Age of AI](https://rob-hoeijmakers.ghost.io/signals-systems-and-speech-rethinking-meaning-in-the-age-of-ai/) - [Semantic web and the knowledge graph](https://rob-hoeijmakers.ghost.io/semantic-web/) ### How a VPN Got Us a Better Hotel Deal URL: https://rob-hoeijmakers.ghost.io/how-a-vpn-got-us-a-better-hotel-deal/ Last updated: 2026-05-10T08:53:56.000Z *What a VPN glitch on the road to San Marino taught me about digital borders* We were somewhere between Rimini and the hills of San Marino, just past a fuel stop where I’d connected to free Wi-Fi to load a map. I had [my VPN](https://rob-hoeijmakers.ghost.io/virtual-private-network-vpn/) switched on — a habit by now, part of my basic toolkit for digital resilience when travelling. I didn’t think much of it. The VPN was set to Argentina, purely because it happens to be first in the alphabetical list. An hour later, while searching for a hotel in San Marino, we noticed something strange. One phone — mine — was showing lower prices than the other. Same hotel, same dates. The only difference: my VPN was still routing traffic through Buenos Aires. And that’s when it clicked. ## A glitch that made sense I’ve heard of this before. Airlines adjusting ticket prices based on your location. Booking sites showing different deals depending on where they think you are. But it hits differently when it happens to you. Two devices, side by side and the VPN makes one of them look like a better guest. We ran a quick experiment. Switched the second phone to Argentina as well. The same price appeared. So we booked. Almost. ## Payments and permissions First, the credit card declined the payment. Four, five times. Then I remembered: some cards require you to pre-authorise usage in specific countries. Apparently, Argentina was considered unusual enough. Once I allowed it manually, the payment went through. It wasn’t effortless. But we had time and a kind of curious focus that road trips often bring. We weren’t gaming the system so much as discovering it. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/07/IMG_5078.jpeg) Declined payments because it looked like I was in Argentina 🇦🇷 ## What changed? Nothing visible. We weren’t physically in Argentina. The hotel wasn’t in Argentina. The payment, too, wasn’t tied to an address. Everything about this booking, *the request, the offer, the fulfilment*, was digital. What changed was a *signal*: where we appeared to be. That signal was enough to alter the terms. It reminded me how little sovereignty we truly have in the digital world or perhaps how distributed that sovereignty has become. Countries still matter. Your virtual location, inferred through IP addresses, quietly reshapes what you’re offered, what you’re charged, and what you’re allowed to do. ## San Marino: 61 square kilometres of sovereignty It was a fitting reflection for the place we were headed. San Marino, a microstate perched on a rocky hill, claims to be the world’s oldest republic. For over 1,500 years, it’s held onto its identity. A dot on the map, but a full participant in diplomacy, law, and the quiet assertion: *we decide who we are*. In contrast, the digital realm seems to blur these edges. We appear where we are not. Offers are shaped not by who we are, but by where we seem to be. The sovereignty lies elsewhere, in servers, algorithms, and pricing models. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/07/IMG_4964.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/07/IMG_4966.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/07/IMG_4992.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/07/IMG_5004.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/07/IMG_5023.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/07/IMG_5038.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/07/IMG_5040.jpeg) San Marino ## A borderless web with invisible borders The internet was supposed to be borderless. And in many ways, it is. But not in the ways that matter economically. What we access, what we pay, even how we're tracked, all depend on a cartography we rarely see. That brief detour through Argentina, *virtual though it wa*s, revealed something real. We were not just travellers on a road through Italy. We were also crossing a very different map: the one behind our screens. ## PS – I might try this again This experience got me curious. I’ll experiment a bit more with geo-pricing, VPNs, and the logic behind these invisible borders. If you'd like to follow along, feel free to [subscribe to my newsletter](https://rob-hoeijmakers.ghost.io/newsletter/). I’ll share more as I go. 💡 Readers ask which VPN I use and it is [CyberGhost](https://cyberghostvpn.com). It does what it needs to do and is relatively affordable. Not sure if I put trust in it, this is a difficult call to make. ### Further reading - [Stay safe online with VPN](https://rob-hoeijmakers.ghost.io/virtual-private-network-vpn/) - [Sovereignty Without Soil](https://rob-hoeijmakers.ghost.io/sovereignty-without-soil/) ### USB-C: A Universal Plug with Many Faces URL: https://rob-hoeijmakers.ghost.io/usb-c/ Last updated: 2026-05-10T08:53:57.000Z **Understanding what’s standard—and what’s not** When I upgraded my phone, I decided to switch everything to USB-C. That meant replacing not just the device itself, but also the ecosystem around it: power packs, chargers, car adapters, and every cable I regularly use. The change was welcome. Most of my equipment now connects via the same sleek, reversible plug. But in the process, I learned something worth sharing: **USB-C is a physical standard, not a guarantee of capability.** ## The Illusion of Standardisation At first glance, USB-C seems like the long-awaited universal solution. One plug to charge all your devices, transfer data, and maybe even connect to external displays. But here’s the catch: **the plug is standard, the function is not.** Behind the identical shape of USB-C ports and cables lie different versions of USB protocols (like USB 2.0, 3.1, or 3.2), power delivery specifications, and sometimes even completely different technologies like **Thunderbolt 3 or 4**. A USB-C cable might look capable of charging a laptop or transferring 4K video, but if it’s built to USB 2.0 specs, it will perform no better than its much older predecessors. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/06/IMG_4423.jpeg) ## Power and Voltage: Another Layer of Complexity Charging performance is where these differences become especially noticeable. Some cables and chargers support **Power Delivery (PD)**, which negotiates the voltage and current between device and charger. This enables fast charging, or in some cases, the ability to power large devices like laptops. Other cables may only provide 5V and 2A—fine for a phone, but insufficient for power-hungry devices. Worse, some cables can even **limit** the performance of good chargers, simply because they lack the correct signalling or wire quality. ## Thunderbolt ≠ USB Adding to the confusion, some USB-C ports are actually **Thunderbolt ports**—especially on Macs and certain PCs. Thunderbolt offers much faster data transfer rates and video output support, but it still uses the USB-C form factor. That means you can plug in a USB-C device, but whether it works as expected depends entirely on what’s behind the port. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/08/IMG_7597.jpeg) Apple Thunderbolt with USB-C plug ## Lessons from the Transition Replacing all my gear gave me a rare chance to see the landscape more clearly. I now carry fewer cables, and most of my setup is interoperable. But I’ve also learned to: - **Check specs before buying**—a cable’s description should mention USB version, PD support, and wattage. - **Label or colour-code cables** if they serve specific functions (e.g. fast charging, video output). - **Be cautious with assumptions**—just because it fits doesn’t mean it works. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/06/IMG_4449.jpeg) Upgraded all aspects to make sure all outgoing connectors are USB-C and do Power Delivery ## In Search of Simplicity That *Simply Works* I’m not chasing minimalism for its own sake. I’m just curious. And when something *does* work simply—like a cable that charges my phone, connects my Kobo, or powers my MacBook without complaint—I take note. The USB-C ecosystem has enormous potential, but we’re still living in a transition period. Uniformity of shape gives a false sense of universality. Until that gap closes, **a little awareness goes a long way**. ### Further reading - [Connecting My Kobo Directly to an iPhone 16 Pro](https://rob-hoeijmakers.ghost.io/connecting-my-kobo-to-iphone/) - [Connecting a Kobo Reader to an iPad via USB-C: Easier Than Expected](https://rob-hoeijmakers.ghost.io/connect-kobo-reader-to-an-ipad-via-usb-c/) ### Connecting My Kobo Directly to an iPhone 16 Pro URL: https://rob-hoeijmakers.ghost.io/connecting-my-kobo-to-iphone/ Last updated: 2026-05-10T08:53:57.000Z I recently discovered that I can plug my Kobo e-reader directly into my iPhone 16 Pro using a regular USB-C to USB-C cable, the same one I carry for charging. Once connected, the Kobo mounts as a storage device inside the Files app. That means I can transfer PDFs or e-books directly, without using a computer or going through Wi-Fi. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/06/IMG_4392.jpeg) Mounted Kobo as a drive on iPhone This wasn’t possible with earlier iPhones. In the past, I had to use [Apple’s camera adapter ](https://rob-hoeijmakers.ghost.io/reading-web-pages-on-a-kobo-e-reader/)and ensure the iPhone had enough charge, otherwise it would display a warning that the connected device required too much power. The workaround worked, but it felt a bit precarious. Now, with the USB-C port on newer iPhones, it just works. And that’s what I enjoy, when something turns out to be unexpectedly simple. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/06/IMG_4742.jpeg) [USB-C to USB-micro cable.](https://www.cablematters.com/pc-714-123-braided-usb-c-to-micro-usb-20-cable.aspx) Newer Kobo’s have USB-C port. ### A Bit of Context on USB-C It’s worth noting that USB-C refers to the shape of the connector, not to what it can do. Some USB-C cables can transfer data or video, others only charge. Some support Thunderbolt or USB 3.2; others are limited to USB 2.0 speeds. So, not every USB-C cable is equal. What matters is the combination of what the port supports and what the cable is capable of. In this case, I just happened to use a cable that allows both charging and data transfer, and that made all the difference. [USB-C: A Universal Plug with Many FacesI stumbled upon a small joy: my e‑reader now connects directly to my iPhone. That moment sparked a deeper look into the messy clarity of the USB‑C standard.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-145.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/04ACC097-0E99-413A-86E6-6CE8E8FE38BF.jpeg)](https://rob-hoeijmakers.ghost.io/usb-c/) ### How I Use This Setup When travelling, or when I don’t have my laptop, I often want to read long articles offline on my Kobo. Here’s how I do it: 1. Open the article in Safari and clean it up (Reader View or a reading app helps). 2. Use the print function, then save as PDF to the Files app. 3. Connect the Kobo via USB-C. 4. Drag the file over to the Kobo storage. It’s straightforward and surprisingly robust. No adapters, no cloud syncs, no Kobo software. Just file transfer, directly between two devices. ### Further reading - [Connecting a Kobo Reader to an iPad via USB-C: Easier Than Expected](https://rob-hoeijmakers.ghost.io/connect-kobo-reader-to-an-ipad-via-usb-c/) - [Load free ebooks on a Kobo using the iPhone](https://rob-hoeijmakers.ghost.io/load-free-ebooks-on-a-kobo-using-the-iphone/) ### The Thermostat Effect: Why AI Feels Distant but Works Deep URL: https://rob-hoeijmakers.ghost.io/thermostat-effect-on-ai/ Last updated: 2026-05-10T08:53:57.000Z When I talk to people about AI, I often get the same question: *"When is it actually going to change how we work? I don’t notice it."* It's a fair question, especially when you consider how much hype surrounds the topic. My go-to metaphor? A thermostat. You turn the dial, but the temperature doesn’t change right away. It takes time. The room has to adjust, the system has to respond, and only then does the difference become noticeable. AI is working the same way: the dial has been turned, and the heat is coming. At Schmuki, the digital and AI agency I co-founded with my partner, we explore these changes in real-world settings. I often draw on our work as a lens to think and write about how AI tools actually perform outside the lab. One part of what we build is highly visible: public-facing chatbots that clarify, guide and respond. The other part goes deeper: internal agents and knowledge systems that support teams, speed up tasks, and quietly reshape workflows. These tools are already having an impact, but the ROI isn’t always immediate or easily measured. This disconnect between capability and perception is what Azeem Azhar calls the *capability-absorption gap*. His recent article, *The* [*$100 Trillion Productivity* ](https://www.exponentialview.co/p/ai-is-ready-is-your-company)*Puzzle*, resonated deeply with me. It explains why the numbers aren’t showing the AI revolution. Yet. Models are improving rapidly, prices are falling, and tools are ready. But businesses are slow to absorb them. The gap is not in the tech; it’s in the uptake. ## Three Frictions Blocking Change Azhar outlines three structural reasons why organisations lag: 1. **Learning Time**: Like the typewriter, which took 25 years to show productivity gains, AI requires cultural and managerial learning before it can deliver results. 2. **Organisational Complexity**: AI doesn’t belong to one department. It cuts across customer service, HR, product design, data analysis. Who leads? Who funds? Who measures? 3. **Non-Determinism**: AI tools can look right and be wrong. They create outputs that pass surface-level checks but require time-consuming verification. Trust is slow to build. These are exactly the tensions we run into with clients. Our chatbots often perform well in user tests, but the real challenge is embedding them into existing structures. Internal tools show promise in test environments, but live implementation depends on how open people are to process change. [💡 The $100 trillion productivity puzzleCapabilities are up, costs are down, productivity is flat—mapping the capability-absorption gap and the moves to close it.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/https-3A-2F-2Fbucketeer-e05bbc84-baa3-437e-9518-adb32be77984.s3.amazonaws.com-2Fpublic-2Fimages-2Faa7453ed-eb02-46d7-b523-74d86b1c1e24-2Fapple-touch-icon-180x180-3.png)Exponential ViewAzeem Azhar![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/https-3A-2F-2Fsubstack-post-media.s3.amazonaws.com-2Fpublic-2Fimages-2F3c207541-5deb-481e-932b-1223cd792255_1600x894.png)](https://www.exponentialview.co/p/ai-is-ready-is-your-company) ## And Yet, the Economics Are Changing One of the most compelling points in Azhar's piece is about unit costs. The price of AI inference has been halving roughly every six months. This matters. The cheaper the tool, the more places you can apply it. At Schmuki, we’ve seen clients save dozens of hours a week by letting AI handle intake forms, knowledge retrieval, and even onboarding content. These aren’t headline-grabbing use cases, but they compound over time. Still, ROI questions persist. One crude but energising exercise is [to compare the cost of an AI agent to the cost of a human](https://rob-hoeijmakers.ghost.io/ai-agent-cost-compared-to-a-human/). It’s not a perfect match, apples and pears, but it sets the mind racing. If an internal agent can do a task 80% as well as a junior hire, and do it 24/7, what does that mean over six months? Over a year? Now multiply that by ten roles or tasks. [What Does an AI Agent Cost Compared to a Human?AI agents are fast, tireless, and scalable – but are they cost-effective? Here’s a crude look at what they really cost, compared to human employees.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-136.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_4342-1.jpeg)](https://rob-hoeijmakers.ghost.io/ai-agent-cost-compared-to-a-human/) ## Redesign, Not Just Replace Here’s where things get practical. The ROI isn’t just about using AI to go faster. It’s about redesigning the work itself. Azhar uses the example of electricity: its economic impact only surged when businesses reorganised factories to take full advantage. The same is true here. Don’t just ask where AI can plug in. Ask what could be rebuilt if AI was your default assumption. That’s the approach we take at Schmuki. Whether it’s automating content workflows, structuring healthcare communication, or integrating voice interfaces in education, our focus is always dual: use AI to improve clarity *and* to support human agency. The tools are here. What matters now is design. ⚠️ It's **not* about ROI—yet. While AI agents often pay for themselves in a matter of hours, ****most clients aren’t buying cost savings**. They're buying: ****\- Control** over their own data and processes \- ****Trust** in a secure, privacy-respecting setup \- ****Clarity** through real-world experiments, not empty promises ROI comes later—once the foundations are solid. If you're selling AI, ****start with agency, not efficiency**. ## The Heat Is Coming You may not feel it yet. But the system is already warming up. At some point, the change will be unmistakable, just like that moment when you realise the room is suddenly warm, and you’ve already taken off your jumper. By then, the real work will have already been done: learning, adapting, redesigning. That’s why it’s worth pushing now. **Rob Hoeijmakers** Digital Strategist at Schmuki [*hoeijmakers.net*](https://rob-hoeijmakers.ghost.io/) | [*schmuki.nl*](https://schmuki.nl) ### Further reading - [How not to lose your job to AI](https://80000hours.org/agi/guide/skills-ai-makes-valuable/) ### What Does an AI Agent Cost Compared to a Human? URL: https://rob-hoeijmakers.ghost.io/ai-agent-cost-compared-to-a-human/ Last updated: 2026-05-10T08:53:57.000Z *This comparison is crude, entirely hypothetical, and viewed solely from the employer’s perspective. It does not account for ethical, societal, or organisational implications.* Everyone working with AI knows the feeling: suddenly there’s a digital colleague who never takes a break, rapidly reads, writes, analyses, listens and observes. But what does such an AI agent actually cost? Not in abstract terms, but really: per day, per month. Let’s break it down. ## What Is an AI Agent? An AI agent combines a language model (like GPT-4o) with tools (search, document storage, visual analysis) and task logic. Think: "analyse this image, generate a description, and send an audio summary." These agents can handle repetitive tasks or support complex workflows. ## Key Cost Drivers Main elements contributing to the total cost: - **LLM usage** (token cost for text input/output) - **Multimodal processing** (image recognition, ASR - automatic speech recognition, TTS - text-to-speech) - **Hosting & compute** (cloud functions, GPU servers, SaaS platforms) - **Storage & context retrieval** (e.g. vector databases for RAG) - **Monitoring & observability** (logging, retry mechanisms) - **Energy usage** (especially for self-hosted setups) - **Licensing & maintenance** (security, updates, integrations) ## Example: Multimodal Agent Doing 100 Tasks Per Day Let’s say an agent handles 100 tasks daily, including summarising, transcribing speech, and analysing images: - Text processing (input/output): €3.60/day - Image analysis (e.g. 50 images): €1.50/day - Speech recognition (e.g. 30 minutes of audio): €2.00/day - **Total per day: €7.10** - **Monthly (20 working days): ≈ €142** Infrastructure and tooling not included. ## Realistic Total: €100–€300/Month Including hosting, monitoring, storage, and integration, the cost per agent typically ranges between €100 and €300 per month. For that, you get a digital teammate handling hundreds of tasks. ## Comparing to Human Labour A full-time employee costs €3500–€6000 per month and offers 6–8 hours of focused work daily. The AI agent operates 24/7 (given stable input/output). - AI doesn’t fully replace, it redistributes tasks - Human oversight and quality control remain essential - New tasks emerge (evaluation, supervision, prompt design) ## Considerations Several factors can significantly affect real-world performance and cost-efficiency: - Complex multimodal tasks (video, speech, images) increase costs - Batch processing is more efficient - Self-hosted models require more maintenance but offer control - Improvement is fast – what costs €100 today might cost €30 next year ## Conclusion AI agents aren’t magic, but they are highly productive, cost-efficient additions to a modern team. The real challenge lies in integration, task design, and change management. But when done right, you gain a tireless coworker for just a few hundred euros a month. Want to see what works for your context? Start with one agent, one task, one working day. You’ll know by tomorrow. ### Further reading - [The Thermostat Effect: Why AI Feels Distant but Works Deep](https://rob-hoeijmakers.ghost.io/thermostat-effect-on-ai/) ### Innovation or Digitalisation? URL: https://rob-hoeijmakers.ghost.io/innovation-or-digitalisation/ Last updated: 2026-05-10T08:53:58.000Z ## What I Learned While Navigating the EU's Tech Funding Landscape When researching European funding instruments for AI-driven tools, I stumbled upon something that changed how I look at digital strategy: the European Union makes a very clear distinction between *innovation* and *digitalisation*. This might sound academic—but if you're building products, offering AI services, or advising clients within the European ecosystem, it's a distinction that directly shapes what kind of support, partnerships and funding routes are open to you. I wrote this piece to share what I learned, and to offer a reference point for others who are navigating the same terrain. ## The Strategic Divide: Innovation ≠ Digitalisation The EU treats **innovation** and **digitalisation** as two distinct tracks: - **Innovation** is about creating something new. It involves risk, experimentation, uncertainty, and often aims to open up entirely new markets or capabilities. - **Digitalisation** is about applying existing digital technologies to improve how we work, serve customers, or run organisations. > Innovation creates. > Digitalisation applies. If you're developing a novel AI-based medical assistant, you're innovating. If you're implementing an existing chatbot framework in a logistics company, you're digitalising. This difference matters because it determines which **EU programme** applies to you. ## Two EU Programmes, Two Different Worlds | Programme | Focus | Typical Projects | | ------------------------------------- | ------------------------------------- | --------------------------------------------- | | **EIC (European Innovation Council)** | High-risk, breakthrough innovation | New AI models, novel hardware, deep tech | | **Digital Europe Programme** | Strategic digital capacity & adoption | Skills, infrastructure, cybersecurity, AI use | The **EIC** offers grants and equity to scale new technologies. It supports startups, spinouts, and R&D-heavy projects. Think: building the next generation of AI. The **Digital Europe Programme** invests in deploying proven digital technologies—AI, supercomputing, cybersecurity, digital skills. Think: helping companies and public services adopt what's already available. ## Why It Matters Most people (myself included) treat innovation and digitalisation as overlapping terms. But the EU doesn’t. And if you're trying to: - apply for funding - position your AI product - align with European strategy ...then recognising which "track" you’re on helps tremendously. > This distinction is not just bureaucratic—it's strategic. Many national governments don’t make this split explicit in their local funding logic. Programmes blur the lines. But if your ambitions point to Europe, understanding this architecture gives you an edge. ## Questions That Clarify Your Track Here are a few questions I now use in strategy conversations: 1. **Are we building something that doesn’t exist yet?** - If yes → likely **EIC**. 1. **Are we applying digital tools to transform an organisation or sector?** - If yes → likely **Digital Europe**. 1. **Do we need capital for high-risk technology development?** - If yes → **EIC**. 1. **Do we need infrastructure, training, or support for applying AI?** - If yes → **Digital Europe**. Some projects sit between the two. But in many cases, this framing helps clarify your positioning. ## Final Thoughts This split—between innovation and digitalisation—is baked into the EU’s funding logic, its strategy documents, and its investment vehicles. Once I started looking through this lens, I saw it everywhere. And I realised: if you're working on AI, digital services, or anything at the intersection of tech and society, **your work fits into this architecture too**. Understanding that helps you: - choose your funding strategy - find the right partners - align with long-term EU priorities If this helped you think more clearly about your next steps, I’d love to hear from you. I’m still exploring this landscape—and building tools to make it more accessible for others. ### Further reading - [Playing the EU Game: Learning the Digital Rules by Getting in the Game](https://rob-hoeijmakers.ghost.io/playing-the-eu-game/) - [Mapping Europe’s Digital Ruleset](https://rob-hoeijmakers.ghost.io/mapping-europes-digital-ruleset-notes-from-the-dutch-frontline/) ### The EuroStack explained URL: https://rob-hoeijmakers.ghost.io/the-eurostack-explained/ Last updated: 2026-05-10T08:53:58.000Z Lately, I’ve been asking myself a simple question: **what is the EU actually doing to *build* its digital future, beyond just regulating others**? It started with a conversation about [digital sovereignty](https://rob-hoeijmakers.ghost.io/sovereignty-without-soil/). One of those after-dinner talks where tech and politics blur. A friend mentioned something called the *EuroStack*, an initiative I hadn’t heard of before. The name alone was enough to pique my interest. I made a note to look it up. At the same time, several clients had brought up similar themes: - “Should we avoid American cloud platforms?” - “What does European hosting *really* mean?” - “Is there a ‘European alternative’ to the big tech stack?” Sometimes the concern comes from a **compliance or defensive angle** — [GDPR](https://rob-hoeijmakers.ghost.io/gdpr/), [AI Act](https://rob-hoeijmakers.ghost.io/learning-to-work-with-the-eu-ai-act/), data locality. Other times it’s more **idealistic or nationalistic**: a wish to support European tech, align with EU values, or simply regain control over their digital infrastructure. That’s when I decided to take a closer look. ## Enter EuroStack EuroStack is a European initiative that aims to define and promote a **sovereign, open, interoperable digital infrastructure**. Not just through policy papers — but through action: - Mapping and endorsing *actual tools* - Coordinating procurement guidelines - Encouraging open standards - Building a directory of trustworthy digital services made and hosted in Europe In other words, **not just "what not to do" (regulation), but "what to choose instead"**. It’s a private–public collaboration, driven by independent tech leaders, policy thinkers, and companies like Proton, Ecosia, and Open-Xchange. They’re pushing for a modular “EuroStack” that could one day rival the dominance of US hyperscalers — not in size, but in **fit**: fit with European values, legislation, and long-term strategic goals. [EuroStackVisit the post for more.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/webclip.png)EuroStack![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/R5__7682-1.jpg)](https://euro-stack.eu) ## Why It Matters As a digital strategist, I think this matters for two reasons: 1. **It’s a rare moment of opportunity.** If we want better tools — interoperable, privacy-respecting, standards-based — we need coordination. EuroStack might offer that. 2. **It’s already relevant for clients.** Whether it's schools, healthcare providers, public services, or mission-driven companies, **the appetite for European solutions is growing**, even if it's not always clear where to look. That’s why I’m going to explore EuroStack more deeply in the coming months. Not just what it *wants* to be, but how it’s taking shape in practice. Where are the gaps? Who’s already aligned? And what would it mean to develop, procure, or recommend tools in this new ecosystem? [The White PaperEuroStack White Paper Final 19.05.25Download![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/webclip-1.png)EuroStack![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/blank.jpg)](https://euro-stack.eu/the-white-paper/) --- If you’ve come across EuroStack in your own work — or if you’re navigating the same questions — [I’d love to hear your thoughts](https://rob-hoeijmakers.ghost.io/about/). Let’s find out what a European digital stack could really mean. ### Further reading - [European industry demands Eurostack to power EU competitiveness](https://www.digitalsme.eu/european-industry-demands-a-eurostack-to-power-eu-competitiveness/) - [From Shopping Cart to Server Rack: How Lidl Builds a European Cloud](https://rob-hoeijmakers.ghost.io/lidl-stackit-european-cloud/) - [Introducing Lumo: AI where every conversation is confidential | Proton](https://proton.me/blog/lumo-ai) ### Sovereignty Without Soil URL: https://rob-hoeijmakers.ghost.io/sovereignty-without-soil/ Last updated: 2026-05-10T08:53:58.000Z *On Knights, Platforms, and the Question of Who Rules.* ## A Conversation About Power I was talking with a friend the other day about power and not in the narrow political sense, but in its larger, ambient form: how it's structured, where it resides, and how it shifts. We were discussing Europe, and the European Union’s efforts to achieve something called *digital sovereignty*. The term gets used often, but its meaning isn’t always clear. At heart, it’s about control: over data, infrastructure, platforms, and, increasingly, over the norms and values encoded into our digital lives. The [EU](https://rob-hoeijmakers.ghost.io/tag/europe/), with its regulations like the [GDPR](https://rob-hoeijmakers.ghost.io/gdpr/) and the [Digital Services Act](https://rob-hoeijmakers.ghost.io/playing-the-eu-game/), is trying to assert this control. But there's an obvious tension: the internet doesn’t respect borders, and most of its major operators aren’t European at all. [The Paradox of Sovereignty: Europe’s Search for Freedom in a Connected WorldThe EU talks about digital sovereignty as independence, yet its strength still depends on shared systems, U.S. clouds, and global code.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-321.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_9067-1.jpeg)](https://rob-hoeijmakers.ghost.io/sovereignty-paradox/) ## From Platforms to Sovereignty I’d recently read a piece that argued, persuasively, that even if Europe wants digital sovereignty, the ecosystem it’s trying to shape is global. This isn’t just a technical challenge. It’s a structural one. You’re working within a system where the most influential players, *Google, Meta, Apple, Amazon*, aren't just companies. They resemble something closer to *sovereigns*. That’s when my friend said it outright: **"These corporations are like sovereign entities. They don’t need land, they have users."** And that’s when it clicked for me. I remembered reading, some time ago, about the **Order of Malta**, a medieval military-religious order, once based on the island of Malta, but today headquartered in Rome, without territory. And yet… still sovereign. ## A Sovereignty Without Land I started looking into them again, the **Sovereign Military Order of Malta**, as they are formally known. It’s one of those fragments of history that still glows faintly, like an ember that refuses to die out. They lost their island, Malta, more than two centuries ago. First to Napoleon, then to the slow erosion of time and politics. Yet remarkably, the Order still exists. It has its own *constitution*, *diplomatic corps*, and *Grand Master*. It issues *passports*, *coins*, *license plates*. Over 100 states recognise it diplomatically. It even holds **observer status at the United Nations**. Its headquarters are in Rome, but the buildings are extraterritorial — meaning Italian laws don’t apply there. Within those walls, it governs itself. Not unlike an embassy, but with no motherland behind it. Its power isn’t military anymore. It's humanitarian. The Order runs hospitals, disaster relief missions, and medical aid across dozens of countries. Quiet, competent, and persistent. And yet, and this is what stayed with me, **it’s considered a sovereign subject of international law**, even without a square metre of land. A state without territory. A sovereignty without soil. That’s when the analogy took shape in my mind. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/06/nicu4176.jpg) The passport. ## Platform Powers When I turned back to the tech giants, *Google, Meta, Apple, Amazon*, the comparison didn’t feel strained. In fact, it felt almost inevitable. They, too, operate without territory in the traditional sense. Their domains are digital, yet their influence is real. Their decisions affect elections, economies, speech, health, even the nature of memory. They don't need embassies because they *are* the infrastructure. For communication, commerce, identity. They have internal rulebooks, *Terms of Service*, that apply more immediately to users than most national laws. They enforce them at scale, through automation and review boards, with appeals and penalties. Their reach often outpaces the state. They issue credentials: logins, verifications, trusted status. They authenticate us. In some cases, they hold more information about an individual than that person’s own government. Their databases are, in a quiet way, *registries of population*. And when it suits them, they negotiate with governments as equals. And sometimes more than equals. They bypass legislatures through product design. They reshape civic life without debate. They offer services the state cannot or no longer does. And for many, they feel more present, more responsive, than any ministry. This isn’t a corporate critique. It’s a structural recognition: **They act like sovereign powers Just not over land. Over flows. Over access. Over attention.** Their borders aren’t visible, but they are enforced. Their currency is not money alone, but identity, engagement, data, and code. [From ID to Account: A Shift from Services to PlatformsApple ID is now Apple Account. It sounds minor, but it marks a deeper shift: from services to platforms, from logins to identities that shape digital life.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-439.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_5996-1.jpeg)](https://rob-hoeijmakers.ghost.io/from-id-to-account-a-shift-from-services-to-platforms/) ## The Disorientation of Disconnection You might ask: *What does any of this really mean?* Try this. Imagine waking up tomorrow and discovering that your Google account is gone. Irretrievably. No email, no contacts, no documents, no calendar. Or perhaps it’s your Meta account, the one that holds your identity in family groups, business pages, photos, birthdays, memories. Or Apple: your devices still work, but your account is locked: no cloud, no backup, no access. Then imagine losing your **passport**. Most of us would say the passport is worse. It’s your right to move, to return, to belong. But pause for a second. Losing access to your digital accounts doesn’t only disrupt; it *disorients*. You lose continuity. Connection. Proof of your past. Your ability to function, professionally, socially, even emotionally, is suddenly impaired. That’s not just dependency. It’s *jurisdiction*. What we’re sensing is that these platforms don’t just serve us. They *bind* us. We live under their rules, inside their systems. And though we don’t call it citizenship, we recognise the same patterns: rights, obligations, identities, exclusions. That’s why the comparison with the Knights of Malta matters. They show that **sovereignty doesn’t always require soil**. And the platforms show that **power doesn’t always require sovereignty**. At least not the kind we’ve historically known. [Locked Out of Instagram or Threads? Avoid and appeal a ban.Banned from Instagram or Threads? Here’s how to avoid getting locked out and the steps to take for a swift and successful account recovery.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-128.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_4133-2.jpeg)](https://rob-hoeijmakers.ghost.io/locked-out-banned-instagram-threads/) ## A Light Comparison with Serious Weight So yes, it's a light comparison, the Knights of Malta and Google. One in robes, the other in hoodies. One founded in the Crusades, the other in a garage. But the resemblance isn’t a joke. It’s a signal. In that conversation with my friend, something shifted. I had always understood, in a rational sense, that platforms wield enormous power. But this gave it contour, and a shape I could think with. It helped me see why some decisions feel political even when they’re just interface changes. Why I feel disoriented when an account is suspended, or why switching platforms sometimes feels like *emigration*. It also helps me read the world I work in. To understand what’s at stake when we talk about digital infrastructure, or regulation, or platform design. And perhaps most importantly, it gives me a tool to think about the future. Because we may not be heading toward a world of digital *states*, exactly. But we are already living under powers that don’t rule territory, they rule *conditions*. And it seems wise to name that, before we forget the difference. ### Further reading - [Understanding Digital Identity in the EU: eID, eIDAS, and the EUDI Wallet](https://rob-hoeijmakers.ghost.io/eu-eid-eidas-eudi-wallet/) - [When your Apple ID gets banned…](https://birchtree.me/blog/when-your-apple-id-gets-banned/) - [20 Years of Digital Life, Gone in an Instant, thanks to Apple](https://hey.paris/posts/appleid/) ### LangWatch: Power of Evals for LLM-Based Systems URL: https://rob-hoeijmakers.ghost.io/discovering-the-power-of-evals-for-llm-based-systems/ Last updated: 2026-05-10T08:53:58.000Z ## Introduction: The Problem with “Working” AI The trouble with live AI systems isn’t that they break in obvious ways. It’s that they drift. A prompt that worked last week might behave differently today. Without warning or error. Especially when you tweak system instructions, update to a new foundation model, or just change the order of a few sentences. That’s not a bug. That’s how probabilistic systems work. But when you're responsible for production-grade chatbots, like we are at [Schmuki](https://schmuki.nl), that uncertainty is a problem. You need to know not just whether something works but whether it's still working, and whether it's working *better*. Manual checks don’t scale. So what’s the alternative? That’s where evals come in. > **“An eval is a repeatable test that measures the quality and consistency of AI-generated outputs—beyond just accuracy.”** They’re how you make LLM behaviour measurable and manageable. ## A Dutch Startup, A Demo, and a Pause It started when someone pointed me to a small company in the Netherlands called [LangWatch](https://www.langwatch.ai/), co-founded by [Manouk Draisma](https://www.linkedin.com/in/manoukdraisma/). They specialise in structured evaluation of LLM outputs. Building evals that track how well prompts, models, and instructions perform over time. I spoke with Manouk, saw a demo, and was intrigued. But I also needed time. It’s abstract stuff. I agreed to a deeper interview but told her honestly, I needed this to sink in first. [LangWatch - Monitor, Evaluate and Optimize your LLM-appsLangWatch helps you monitor, evaluate, and optimize your LLM-powered applications—gain full visibility, improve performance, and ensure reliability in production![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/1kwc4NUPvLQiZuEq5mT8ECfRc0.png)Monitor, Evaluate and Optimize your LLM-apps![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/9grgTodc6LbG1zwF8liAx4ZPI.png)](https://langwatch.ai) ## A Podcast That Made It Click Fast forward a few weeks. I resumed a podcast I had half-finished: [Azeem Azhar in conversation with Kevin Weil](https://www.exponentialview.co/p/kevin-weil-openai-on-building-gpt), Chief Product Officer at OpenAI. And there it was, right in the middle of their conversation, a clear, grounded explanation of how OpenAI uses evals internally to iterate, monitor, and productise complex model behaviours. > *“The most effective way to build products,” Weil says, “is to take the skill you want the model to have and turn that into an eval.”* Suddenly, the theory met practice. It wasn’t just about [benchmarking](https://rob-hoeijmakers.ghost.io/where-does-the-word-benchmark-come-from/). Evals had become the interface between engineering, product, and research. [🔮 OpenAI’s CPO on building at the cutting edgeLive with OpenAI CPO Kevin Weil![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/https-3A-2F-2Fbucketeer-e05bbc84-baa3-437e-9518-adb32be77984.s3.amazonaws.com-2Fpublic-2Fimages-2Faa7453ed-eb02-46d7-b523-74d86b1c1e24-2Fapple-touch-icon-180x180-2.png)Exponential ViewAzeem Azhar![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/https-3A-2F-2Fsubstack-video.s3.amazonaws.com-2Fvideo_upload-2Fpost-2F165210549-2Fb86e686a-e8b4-4f9d-8a8d-3ff8308fdd7f-2Ftranscoded-342085.png)](https://www.exponentialview.co/p/openais-cpo-on-building-at-the-cutting) ## What Evals Actually Are An **eval** is a structured way to test and measure LLM behaviour across different dimensions—accuracy, helpfulness, tone, reasoning, or creativity. They can be: - Automatically scored (e.g. through classifiers or models) - Human-reviewed (e.g. by rating or voting) - Or hybrid (a bit of both) But the key feature is this: **evals are repeatable**. They allow you to measure the impact of changes over time, whether that’s to prompts, models, instructions, or settings. You can think of them as: - Regression tests for language models - Custom benchmarks for your own product criteria - Signals for when something goes off track. Or gets better ## LangWatch: Making This Tangible LangWatch is building tools that bring evals into real-world practice. Their platform lets you: - Track model and prompt performance over time - Compare outputs across versions or experiments - Get notified when behaviour changes unexpectedly - Customise what “quality” means for your product It works like CI/CD for LLMs: continuous monitoring, feedback, and confidence in production. With LangWatch, evals stop being an academic concept and become part of your deployment workflow. I'll be doing a follow-up interview with Manouk soon to explore how LangWatch works under the hood. ## Why Evals Matter (Especially Now) Evals aren’t just for research labs anymore. They’re becoming essential for: - Product quality assurance - Safety and trustworthiness - Prompt development workflows - Debugging complex interactions - Auditing and documentation (especially in regulated sectors) As more teams deploy assistants, agents, and AI interfaces in production, evals are the missing link between “it seems fine” and “we know it’s working”. Whether you're managing a customer support bot, an internal tool, or a knowledge assistant, evals help you track what's working and prove it. > **If you’re not using evals to monitor your LLM systems, you’re flying blind.** ## What’s Next For me, this was a moment where something abstract became practical. The combination of a hands-on demo and a well-timed podcast made it click. And now it’s part of how I think about every chatbot we run. If you’re building with LLMs, take a moment to look into evals. They’re not just about safety or performance, they’re about building with confidence. I’ll share more soon in a follow-up conversation with Manouk from LangWatch. In the meantime, [I’d love to hear how *you* measure whether your AI](https://rob-hoeijmakers.ghost.io/about/) is still doing what it’s supposed to do. ### Further reading - [From Benchmarks to Evals: How We Measure AI and Why It Matters](https://rob-hoeijmakers.ghost.io/ai-benchmarks-and-evals/) ### Learning to Work with the EU AI Act URL: https://rob-hoeijmakers.ghost.io/learning-to-work-with-the-eu-ai-act/ Last updated: 2026-05-10T08:53:58.000Z *From frustration to foundation* When the AI Act was first announced, I wasn’t thrilled. Another regulation from Brussels. Another compliance burden. Another list of things we couldn’t do. I’ve always leaned into experimentation, iteration, and autonomy, and legal frameworks often felt like the opposite of that. Especially in AI, where the pace of development outruns most policy by years. But something shifted. Not overnight, and not because the regulation got more exciting. It didn’t. What changed was my understanding of what the EU is trying to do. And how that might actually align with the kind of digital future I care about. This article is the beginning of that reorientation. An attempt to ground myself, and others, in what the AI Act *really* asks of us, and how we might respond. Not reluctantly, but intelligently. [Playing the EU Game: Learning the Digital Rules by Getting in the GameFrom grants to governance, the EU’s digital world is a maze. I dove in—starting with DG CONNECT, DIGITAL, and the logic behind the rules.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-125.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_3830-1.jpeg)](https://rob-hoeijmakers.ghost.io/playing-the-eu-game/) ## What the AI Act Is and Isn’t The AI Act is the first major attempt by a democratic region to comprehensively regulate artificial intelligence. It doesn’t regulate *all* AI. Instead, it focuses on systems that pose risks to fundamental rights, health, safety, or democratic values. That includes things like biometric surveillance, automated hiring tools, credit scoring, and deepfake generation. The Act introduces a four-tier risk framework: 1. **Unacceptable Risk** – Completely banned (e.g. social scoring, real-time biometric surveillance in public spaces). 2. **High Risk** – Heavily regulated (e.g. education, healthcare, recruitment, law enforcement). 3. **Limited Risk** – Subject to transparency obligations (e.g. chatbots). 4. **Minimal Risk** – Allowed with no extra obligations (e.g. spam filters). That structure makes it more navigable than it looks — especially once you accept that not all AI needs to be treated as dangerous. [AI ActThe AI Act is the first-ever legal framework on AI, which addresses the risks of AI and positions Europe to play a leading role globally.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon-7.svg)Shaping Europe’s digital future![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/pyramid_7F5843E5-9386-8052-931F5C4E98C6E5F2_75757-1.jpg)](https://digital-strategy.ec.europa.eu/en/policies/regulatory-framework-ai) ## What It Means in Practice If you’re deploying AI in or for the EU, the law will affect how you work even if your system isn’t high-risk. You’ll be expected to: - **Classify your system** according to the risk tiers. - **Maintain documentation** about how the system was trained, validated, and monitored. - **Ensure transparency** in user-facing systems, for example, letting people know they’re interacting with a bot. - **Be audit-ready**, especially if you’re offering AI systems at scale or in sensitive domains. - **Avoid prohibited practices** entirely, such as manipulative or exploitative AI. The fines for non-compliance are significant, up to €35 million or 7% of global turnover, but for most organisations, the real cost is reputational and strategic. Not knowing where your AI sits on the risk ladder makes you vulnerable. Knowing puts you in a position of strength. ## Accepting the Premise I still believe regulation can be slow, and sometimes clumsy. But I’ve also come to believe that European AI law, with all its imperfections, is built on a set of ideas worth engaging with. That technology should serve people. That risk should be understood, not dismissed. That transparency isn’t a constraint, but a way to build trust. These are not slogans, they’re working principles. And they require work. I’m not writing this from the position of someone who has fully figured it out. This article isn’t the end of a journey. It’s the start of a shift. A move from compliance avoidance to proactive orientation. From skimming the surface to building real understanding. 🚧 ****Update (June 2025):** The EU AI Act is already [facing pressure for delay and revision](https://www.politico.eu/article/how-the-eu-ai-rules-turned-into-massive-headache/). Industry groups and U.S. officials have called for more time to comply, while the European Commission is considering postponing parts of the law due to missing technical guidelines. Critics warn this uncertainty risks undermining trust in the EU’s regulatory leadership. ## Where I’m Headed I’m treating the AI Act as a foundation, something to build on, not just work around. Over the coming months I’ll explore what that means in practice: for chatbot development, business automation, digital agency work, and AI-powered content. I’ll write about risk classification, practical tooling, small-team compliance, and what it means to be “trustworthy” in a regulatory environment. If you’re on a similar path, building digital systems that touch real people, in a European context, you’ll likely face the same questions. This series is an attempt to think through them, in public, with a mix of strategy, practice, and curiosity. I’m not trying to win the EU game. I’m learning how to play it. On my own terms. ### Further reading - [The Hot Potato of Compliance](https://rob-hoeijmakers.ghost.io/the-hot-potato-of-compliance/) - [EU AI Act: first regulation on artificial intelligence | Topics | European Parliament](https://www.europarl.europa.eu/topics/en/article/20230601STO93804/eu-ai-act-first-regulation-on-artificial-intelligence) - [EU’s waffle on artificial intelligence law creates huge headache](https://www.politico.eu/article/how-the-eu-ai-rules-turned-into-massive-headache/) ### The Digital Markets Act (DMA) is Here and It's Changing How Apps Work URL: https://rob-hoeijmakers.ghost.io/eu-digital-markets-act/ Last updated: 2026-05-10T08:53:59.000Z If you’ve noticed that Google Maps links no longer open smoothly, or that Instagram is suddenly asking if Meta can use your data, you’re not alone. These aren’t bugs. They’re features of a new reality: the [**Digital Markets Act** (DMA)](https://digital-markets-act.ec.europa.eu/index%5Fen) has landed. This European law is reshaping how dominant tech platforms operate in the EU. And while it promises more fairness and freedom for users and developers, the shift also brings **confusion, friction, and a lot more choice screens**. Let’s unpack what the DMA is, what it’s changing — and how it might benefit or annoy you as a user. ## What is the Digital Markets Act (DMA)? The DMA is a new European law aimed at limiting the power of Big Tech. It targets **“gatekeepers”** — very large companies like Google, Apple, Meta, Amazon, Microsoft, and TikTok — who control key digital services that millions depend on daily. The law forces these companies to: - Stop favouring their own apps and services, - Open up to competitors, - And give users more control over defaults, data, and connections. [Digital Markets ActThe Digital Markets Act is the EU’s law to make the markets in the digital sector fairer and more contestable.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon-3.svg)Digital Markets Act (DMA)![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/ec-socialmedia-fallback-1.png)](https://digital-markets-act.ec.europa.eu/index%5Fen) ## Real-World Examples of DMA Changes Here’s how the DMA is already affecting everyday digital life in the EU: ### 1\. Google Maps links don’t open directly anymore When you tap a Maps result in Google Search, it no longer jumps straight into the Maps app. Instead, it may: - Open a browser preview, - Or ask which app to use. **Why?** Google isn’t allowed to automatically promote its own apps — it has to let you choose. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/06/2025-06-09_15-14-01.gif) Very annoying, you will need to go to maps manually. ### 2\. WhatsApp might soon talk to Signal or Telegram Messaging platforms like WhatsApp and Messenger will be required to support **cross-platform messaging**. This means you could: - Send a WhatsApp message to someone using another app — like Signal — and vice versa. **Why?** To break closed ecosystems and let users connect more freely. ### 3\. Apple is allowing sideloading (EU only) On iPhones and iPads in the EU, Apple now allows: - Installing apps from **alternative app stores**, - And sideloading from the web — something previously banned. **Why?** The DMA says users must be free to choose where they get their apps. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/06/iOS-Default-Apps.jpg) Sideloading means install alternative apps. Like WhatsApp as the default for calling. ### 4\. Meta must offer AI opt-outs Meta is rolling out **Meta AI** across Facebook, Instagram, Threads, and WhatsApp. But in the EU: - Users are shown **notifications and opt-out forms**, especially around how public content might be used to train the AI. **Why?** Under GDPR and DMA, users must be able to refuse the use of their personal data. [Should I Opt Out of Meta’s AI Training?Opting out sounds easy. But what if it costs visibility—or influence? I’m sharing the tension, not the answer. Curious what others are doing.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-120.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_2506-1-1.jpeg)](https://rob-hoeijmakers.ghost.io/should-i-opt-out-of-metas-ai-training/) ### 5\. YouTube and Gmail ask more questions - YouTube asks for consent before tracking. - Gmail may no longer be the default email handler. - Google’s “Sign in with Google” can’t be more prominent than other login methods. **Why?** Companies must stop making their services the “easy default” unless users actively choose them. ### 6\. Android users must pick a default search engine During setup, Android now shows a screen asking which search engine you want — Google, Bing, DuckDuckGo, etc. **Why?** Again, it’s about giving real choice and avoiding default bias. ## Why this matters — and why it’s annoying ### ✅ The benefits - **More control** over your data, apps, and default services. - **More competition**, which may lead to better alternatives. - **Less lock-in**, especially across platforms like Meta or Apple. ### ⚠️ The annoyances - **More popups**, consent screens, and app choices. - **Familiar things break** — like Maps links not opening where you expect. - **Extra friction** when doing simple tasks that used to be seamless. ## The bigger picture The DMA isn’t about making tech worse. It’s about making tech **more fair, transparent, and open** — especially when it’s controlled by companies too big to fail. The UX hit we’re all feeling now is part of a deeper redesign of the digital economy. Over time, we may get better tools, more trustworthy services, and platforms that respect both **user choice** and **developer opportunity**. Until then, yes — expect more buttons, fewer shortcuts, and a bit of a mess. But also: a lot more freedom underneath it. ### Further reading - [Apple could remove AirDrop from EU iPhones as battle heats up](https://9to5mac.com/2025/06/03/apple-could-remove-airdrop-from-eu-iphones-as-legal-battle-heats-up/) - [From Resistance to Strategy: What the EU’s Digital Acts Mean for the Work I Do](https://rob-hoeijmakers.ghost.io/eu-digital-acts/) - [From Shopping Cart to Server Rack: How Lidl Builds a European Cloud](https://rob-hoeijmakers.ghost.io/lidl-stackit-european-cloud/) ### Intentional Power: Europe Between Negawatts and Realpolitik URL: https://rob-hoeijmakers.ghost.io/intentional-power-europe-between-negawatts-and-realpolitik/ Last updated: 2026-05-14T11:54:13.000Z Artificial intelligence runs on energy. That’s something I hadn’t fully appreciated until recently. Every digital service, every model deployment, every GPU cluster puts pressure on the grid. And that pressure is rising fast—especially here in Europe. So this isn’t just an energy story. It’s a digital one. And it touches on sovereignty too. I’m starting to see how these things connect. And because I work in digital strategy and infrastructure, it feels worth sharing some thoughts—not as conclusions, but as part of a broader exploration. I’ve always liked the idea of *negawatts*. Slim systems. Thoughtful use. A resilient, considered approach to energy. The kind of thinking that values intention over indulgence. That mindset has shaped a lot of European energy policy—and it aligns with how I tend to approach digital work too. Efficiency, clarity, design with purpose. But recently I’ve been noticing a tension. Because we also need scale. Growth. And real power to run the tools we now rely on. ## From Less to More: The Shift in Energy Thinking Negawatts—using less energy instead of producing more—made a lot of sense when carbon reduction was the central challenge. It still does. But something has changed. Electric vehicles, heat pumps, and AI workloads are all increasing demand on a grid that hasn’t kept up. In many parts of Europe, it hasn’t grown much at all in the last decade. Meanwhile,[ Goldman Sachs expects European power demand to rise 10–15% by 2030](https://www.goldmansachs.com/insights/articles/data-centers-could-boost-european-power-demand-by-30-percent?utm%5Fsource=substack&utm%5Fmedium=email), mostly due to AI. That’s a shift I hadn’t really internalised. But it’s starting to feel very real. ## A World That Builds Faster Than We Debate Other regions are moving differently. The US-UAE AI deal, for example, gave Abu Dhabi access to half a million Nvidia chips and a 5-gigawatt energy base to power them—all within 100 milliseconds of Europe’s biggest cities. It made me wonder: what exactly are we building here? And how quickly can we build it? [Nvidia will supply advanced chips for Saudi Arabia’s Humain AI projectSaudi Arabia is making a massive push into AI infrastructure, working with Nvidia, AMD, and Amazon Web Services to equip its new state-owned company, Humain, with chips, software, and large-scale data centers.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/apple-touch-icon-180x180.png)THE DECODERMaximilian Schreiner![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/Humain-Invest-US.png)](https://the-decoder.com/nvidia-will-supply-advanced-chips-for-saudi-arabias-humain-ai-project/?utm%5Fsource=substack&utm%5Fmedium=email) ## Digital Sovereignty Needs Power I hear a lot about *digital sovereignty* in European discussions. But it’s clearer to me now that sovereignty doesn’t just mean policy. It means infrastructure. Cables, chips, campuses and electricity to run them. This is where the negawatt mindset starts to feel incomplete. Not wrong, just not quite enough on its own. We still need that focus on intention. But we also need the ability to scale and otherwise the ideas we care about don’t have anywhere to live. ## Strategic Materials: The Hidden Layer of Sovereignty There’s also the question of materials. A lot of the things we need, GPUs, EVs, wind turbines, depend on rare earths. And most of the world’s refining capacity is in China. Europe might shape policy, but we don’t mine the metals. That’s a dependency that’s hard to ignore if we’re serious about autonomy. [Rare Earth Elements![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/apple-touch-icon-144-precomposed-1.png)![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/FacebookIcon-1.png)](https://www.rareelementresources.com/rare-earth-elements) ## What If We Built with Intention and Capacity? This has led me to a question I keep coming back to: can we build systems that are both abundant and intentional? Not just more energy, but infrastructure that reflects how we want to live and work—supporting autonomy, resilience, and room to grow. Not everything needs to scale, but when it does, it should be for the right reasons. > **Industrial atrophy will follow.** High-energy processes (steel, chemicals, semiconductors) migrate to jurisdictions that guarantee 24/7 clean power at predictable prices, eroding the continent’s industrial tax base. Azeem Azar, Exponential View. ## This Is a Strategic Question Now Even if you’re mostly working in digital, this matters. Because we’re not just building tools—we’re building dependencies. And if we don’t create the conditions for those tools to run on our terms, someone else will. ## We Need a New Imagination I’m still thinking this through. But I’ve come to believe that energy, digital strategy, and sovereignty are now one conversation. Constraint has shaped Europe’s identity for decades. Maybe that doesn’t need to change. But perhaps it’s time we added something new alongside it: the ability to build at a scale that reflects what we actually want to protect. Not for excess. Just for agency. And maybe now is the time to start saying that out loud. ### Further reading - [🔮 Europe‘s power problem](https://www.exponentialview.co/p/europe-is-losing-unless) - [‘Sovereign AI’ is political branding. The reality is closer to digital colonialism](https://fortune.com/2025/06/09/ai-chips-geopolitics-tech-data-centers/) ### From Resistance to Strategy: What the EU’s Digital Acts Mean for the Work I Do URL: https://rob-hoeijmakers.ghost.io/eu-digital-acts/ Last updated: 2026-05-12T17:46:29.000Z ## The Internet I Grew Up With I’ve been working in the digital domain for over three decades. I remember when the internet felt like a boundless frontier—free, decentralised, full of promise. The tools were ours. The pace was ours. And the experiments? Limitless. But over time, that open web began to show its cracks: manipulation, polarisation, surveillance, and monopoly power. The last decade made the risks impossible to ignore. For a long time, I resisted regulation. It felt like a wet blanket over innovation. I saw the European Union’s growing body of laws as either too abstract or too heavy-handed. Were these acts meant to protect us, or just slow everyone down? That view has changed. ## A Turning Point What helped me shift wasn’t just reading the acts—it was engaging with them. Checking EU tenders, reading portals, exploring regulations from the inside out. The more I explored, the more tangible it became. Yes, the EU can feel bureaucratic and distant. But a friend said something that reframed everything: “The EU is countries choosing to cooperate.” That landed. It gave me a new way to see this vast regulatory effort. Not as a blocker, but as a framework. Not flawless, but strategic. Designed, ultimately, to create fairer and more sustainable digital ecosystems—for citizens, workers, startups, and those of us who build digital tools and strategies. [Elvis Eckardt Recruitment (@elviseckardtrecruitment) on ThreadsUS tech giants are now spending €151M/year lobbying the EU. Meta tops the list at \~€10M. Amazon, Microsoft, Apple, and Google are close behind. Regulation is now a strategic battleground, not just compliance.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Pcnemah90K8-3.png)Threads![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/358345322_1209296103094155_5315289913515751368_n.jpg)](https://www.threads.com/@elviseckardtrecruitment/post/DQ4OSSqD-BJ?xmt=AQF0SAOEpCVdjfOuiH9gxDbXuLaMGo5sAuWz8TXlqiVH0dbySJpJv65XQf4N58NmrJXKyiXE&slof=1) ## What the EU Is Actually Doing To global readers: Europe is no longer just a consumer of digital platforms—it’s trying to shape them. Through a cascade of interlocking regulations, the EU is now actively defining rules of the road for data, platforms, AI, identity and infrastructure. And it’s doing so with one foot in ethics and the other in competition strategy. Let’s break it down: ### Data & Infrastructure - **Data Governance Act (2022)**: Facilitates cross-sector data sharing via neutral intermediaries. - **Data Act (2023, in force 2025)**: Defines access rights to data generated by connected devices (IoT), especially between businesses. ### Platforms & Market Fairness - [**Digital Services Act (2024)**](https://rob-hoeijmakers.ghost.io/europes-dsa-and-the-new-contours-of-online-freedom/): Requires transparency and accountability from platforms, especially very large ones (VLOPs). - [**Digital Markets Act (2024)**](https://rob-hoeijmakers.ghost.io/eu-digital-markets-act/): Targets gatekeeper platforms (think Apple, Meta, Google), banning self-preferencing and mandating interoperability. ### Artificial Intelligence - [**AI Act (2024)**](https://rob-hoeijmakers.ghost.io/learning-to-work-with-the-eu-ai-act/): The world’s first horizontal AI law. Classifies AI systems by risk, bans certain uses (like mass surveillance), and mandates transparency. ### Identity & Trust - [**eIDAS 2.0 (2024)**](https://rob-hoeijmakers.ghost.io/eu-eid-eidas-eudi-wallet/): Introduces the EU Digital Identity Wallet. Think verified credentials for individuals and companies, usable across borders. ### Cybersecurity & Resilience - **NIS2 Directive (2024)** and **Cyber Resilience Act (2024)**: Extend security obligations to software providers and digital infrastructure. ## Why It Matters to My Work Working in the fields of AI, automation, and digital experience design, I began to notice how these laws touch real, everyday decisions. - **The** [**AI Act**](https://rob-hoeijmakers.ghost.io/learning-to-work-with-the-eu-ai-act/) makes me design for explainability. When building chatbots, I now consider how to communicate purpose, logic and limits to users. - **The Data Act** challenges assumptions about ownership in IoT and client-side integrations. Data access becomes part of the UX. - **The** [**DSA**](https://rob-hoeijmakers.ghost.io/europes-dsa-and-the-new-contours-of-online-freedom/) has already influenced how I think about transparency in WhatsApp bots and campaign automations. - **eIDAS 2.0** signals a shift toward verifiable identity workflows, especially in onboarding, consent, and user authentication. My work with [Schmuki](https://schmuki.nl), a Dutch digital and AI agency, has given me a close-up view of how these changes affect both strategic choices and technical design. The regulatory space isn’t background noise—it’s a real part of the architecture. ## Strategic Takeaways These aren’t rules carved in stone, but patterns I’ve started to notice—signals that shape how I approach digital strategy and product design in light of the EU's evolving regulatory framework. ### Transparency is no longer optional - From algorithmic explanations to clear data flows, being open is a value and a requirement. ### Openness will outperform lock-in - Platforms and services that embrace interoperability will have an edge, legally and commercially. ### Design for the future, not just the launch - Lifecycle thinking (security, updates, data rights) is now part of product strategy, not just IT. ## A Personal Note For years, I resisted the rulebook. I loved the chaos of newness, the joy of invention. But as the stakes grew, so did the need to understand where things were heading. The EU’s digital acts aren’t perfect, but they signal an intention to do things differently—to bring structure where there was fragmentation. It doesn’t always feel smooth, and sometimes it’s a lot to take in. But it’s part of the environment now, and working with it is proving more productive than pushing back. Everyone will find their own rhythm with these changes. I’m just sharing where I am with it. If anything here sounds familiar, or completely foreign, I’d be curious to hear how you're seeing it. This space is still shifting. ### Further reading - [‘Sovereign AI’ is political branding. The reality is closer to digital colonialism](https://fortune.com/2025/06/09/ai-chips-geopolitics-tech-data-centers/) - [Europe’s DSA and the New Contours of Online Freedom](https://rob-hoeijmakers.ghost.io/europes-dsa-and-the-new-contours-of-online-freedom/) ### Mapping Europe’s Digital Ruleset URL: https://rob-hoeijmakers.ghost.io/mapping-europes-digital-ruleset-notes-from-the-dutch-frontline/ Last updated: 2026-05-10T08:53:59.000Z **With a focus on the Netherlands** Over the past year, I’ve been tracing the outlines of Europe’s digital future, and not as an academic, but as a strategist building AI systems, advising clients, and navigating funding calls. What began as a tactical need (“how do we make this chatbot compliant?”) quickly turned into something more ambitious: a mapping of **Europe’s emerging digital infrastructure**. Not just the technologies, but the *rules*, the *funds*, the *agencies*, and the *intentions behind them*. This piece summarises what I’ve found so far. It’s focused on the Dutch ecosystem, but the structure, EU laws upstream, national instruments midstream, and local execution downstream, is broadly applicable across Europe. If you work in a startup, agency, research group, or fund, this may help clarify where to plug in. ## From Regulation to Reality: The Layers of the EU Stack Europe’s regulatory foundation is unusually ambitious — and uniquely coherent once you look at it as a **multi-layered digital constitution**. Here’s the shorthand I use when explaining it to clients: - [**GDPR**](https://rob-hoeijmakers.ghost.io/gdpr/) → who owns the data - [**eIDAS 2.0**](https://rob-hoeijmakers.ghost.io/eu-eid-eidas-eudi-wallet/) → who you are - [**EHDS**](https://rob-hoeijmakers.ghost.io/the-european-health-data-space/) → what keeps you alive - [**Digital Euro**](https://rob-hoeijmakers.ghost.io/digital-euro-cbdc/) → how you pay - [**DSA**](https://rob-hoeijmakers.ghost.io/europes-dsa-and-the-new-contours-of-online-freedom/)**/**[**DMA**](https://rob-hoeijmakers.ghost.io/eu-digital-markets-act/) → how platforms behave - [**AI Act**](https://rob-hoeijmakers.ghost.io/learning-to-work-with-the-eu-ai-act/) → how automation is governed - [**Chips Act / ESMC**](https://rob-hoeijmakers.ghost.io/esmc/) → what underpins it all Each piece targets a domain, but together they form a **rights-based, sovereignty-driven blueprint** for digital life. Europe may move slowly, but it legislates holistically. ## What Maps to What: Articles, DGs and Programmes Here’s how the topics I’ve been writing about connect to the EU’s institutional machinery. Think of this as a **working translation layer** between content, governance, and funding. | **Topic / Article** | **Responsible EU DG(s)** | **Relevant Programme / Policy Context** | | ------------------------------------------------ | ---------------------------- | ------------------------------------------------------ | | **eIDAS 2.0 / European Digital Identity Wallet** | DG CNECT | Digital Europe Programme (DIGITAL), eID Framework | | **EHDS / Health data systems** | DG SANTE, DG CNECT | EU4Health, European Health Union, European Data Spaces | | **Digital Euro / CBDC** | ECB, DG FISMA | Eurosystem Strategy, Retail Payments Strategy | | **ESMC / Semiconductor production in Dresden** | DG GROW, DG COMP (state aid) | European Chips Act (€43B), Industrial Strategy | | **GDPR / Data protection** | DG JUST, EDPB | Charter of Fundamental Rights, Digital Single Market | | **DSA / Platform accountability** | DG CNECT | Digital Services Act Package | | **DMA / Gatekeeper market regulation** | DG COMP, DG CNECT | Fair Digital Market Framework | | **AI Act / Risk-based regulation of AI systems** | DG CNECT, DG GROW | Coordinated Plan on AI, Digital Europe Programme | 💡 Note that the [Digital Euro](https://rob-hoeijmakers.ghost.io/digital-euro-cbdc/) is not an initiative driven by the European Commission (EC), but by the European Central Bank (ECB). The ECB leads the project as part of its mandate to maintain monetary stability in the euro area and to ensure the euro remains fit for the digital age. ## Programmes and Platforms: Where Strategy Becomes Money This ruleset is not theoretical. It’s backed by serious capital. The trick is navigating the overlapping **funding and implementation channels**: | **Instrument / Fund** | **Focus** | | -------------------------------------- | ----------------------------------------------------------- | | **Digital Europe Programme (DIGITAL)** | AI, cybersecurity, cloud, interoperability, skills | | **Horizon Europe** | R&D, foundational and applied science (incl. AI & robotics) | | **Chips Joint Undertaking (Chips JU)** | Semiconductor R&D and pilot lines | | **European Innovation Council (EIC)** | Deep tech scale-ups, SME acceleration | | **EU4Health** | Health innovation, data infrastructure (EHDS) | | **Connecting Europe Facility (CEF)** | Cross-border digital services and infrastructure | In the Netherlands, many of these flow through or are complemented by: - [**RVO** (Netherlands Enterprise Agency)](https://rvo.nl) - [**EDIHs** (European Digital Innovation Hubs)](https://european-digital-innovation-hubs.ec.europa.eu/edih-catalogue) - [**TTT-AI (Thematic Technology Transfer AI)**](https://ttt-ai.nl) - **ROMs** (Regionale Ontwikkelingsmaatschappijen), [like ROM InWest](https://rominwest.nl) ## TED, Tenders, and National Alignment Beyond grants, the **Tenders Electronic Daily (TED)** system is the EU's procurement portal — worth watching for market opportunities tied to public sector innovation. [TED ---- TED - EU Tenders, the Supplement to the Official JournalWelcome to the official portal of TED, Tenders Electronic Daily. TED contains all active notices published in the Supplement to the EU Official Journal.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/d3767476-8b80-a2c9-d9e8-245b087fde48)TED![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/flag-eu.svg)](https://ted.europa.eu/en/) [TenderNed, hét online marktplein voor aanbestedingenTenderNed is het aanbestedingssysteem van de Nederlandse overheid. U vindt hier alle opdrachten van de overheid op één plek.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon-22.ico)![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/logo-tenderned.svg)](https://www.tenderned.nl/cms/nl) In the Netherlands, TED activity often aligns with national strategies like: - [**NL AI Coalition**](https://nlaic.com) - [**Digital Government Agenda**](https://www.nldigitalgovernment.nl/digital-government-agenda/) - [**Health-RI**](https://www.health-ri.nl/en) - [**Smart Industry Fieldlabs**](https://smartindustry.nl/fieldlabs) RVO, in coordination with ministries and **Topsectors**, helps steer these into [**national innovation calls**](https://www.rvo.nl/onderwerpen/digital-europe-programme/digital-europe/calls-hoofdwerkprogramma-2025-2027), often co-funded by the EU. [Overzicht calls hoofdwerkprogramma Digital Europe ProgrammeOp deze pagina vindt u een overzicht van de onderwerpen uit het hoofdwerkprogramma van het Digital Europe Programme waarvoor u in de periode 2025-2027 financiering kunt aanvragen. Het betreft onderwerpen op het gebied van AI, data, cloud en digitale vaardigheden. U vindt hier ook informatie over de openings- en sluitingsdata, subsidiebudgetten en welke onderwerpen in aanmerking komen voor nationale cofinanciering. Op deze pagina vindt u een overzicht van de onderwerpen uit het hoofdwerkprogramma van het Digital Europe Programme waarvoor u in de periode 2025-2027 financiering kunt aanvragen. Het betreft onderwerpen op het gebied van AI, data, cloud en digitale vaardigheden. U vindt hier ook informatie over de openings- en sluitingsdata, subsidiebudgetten en welke onderwerpen in aanmerking komen voor nationale cofinanciering.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon-23.ico)RVO.nl![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/chevron.svg)](https://www.rvo.nl/onderwerpen/digital-europe-programme/digital-europe/calls-hoofdwerkprogramma-2025-2027) ## From Research to Market: The Triple Helix of Innovation One of the most effective structures I’ve seen is the **Triple Helix model**, long used in Dutch public-private innovation hubs. It weaves together: - **Knowledge institutes** (universities, TNO, NWO-funded labs) - **Commercial parties** (startups, corporates) - **Government support** (municipal, national, EU) The idea is not just to transfer knowledge, but to **co-develop and co-own technology** — particularly in AI, semiconductors, and life sciences. Programmes like **TTT-AI** formalise this model nationally; [**ESMC in Dresden**](https://rob-hoeijmakers.ghost.io/esmc/) applies a similar approach at the continental level. ## Why This Mapping Matters For digital professionals and founders, this landscape is complex — but not opaque. There’s structure to it: - **EU law sets the boundaries and protections** - **EU programmes fund cross-border R&D and deployment** - **National agencies translate them into local ecosystems** - **Startups and agencies create the applications — in AI, identity, health, or mobility** If you understand the logic of this stack, you can **navigate it strategically** and not just reactively. You can align your tech with regulation *before* it’s enforced. You can use procurement timelines as product roadmaps. You can position your startup in the flow of structural investment rather than chasing hype. 💡 This map focuses on institutions and funding flows — but entrepreneurs also need to navigate the influence layer: from [NLdigital](https://www.nldigital.nl) and [VNO-NCW](https://www.vno-ncw.nl) to [TechLeap](https://techleap.nl) and regional employer networks. These groups don’t just interpret rules, they help shape them and guide who gets early access to emerging opportunities. Add to that the role of ****venture capital** — not just as money, but as a signal. Funds like [Invest-NL](https://www.invest-nl.nl/nl), [Cottonwood](https://rob-hoeijmakers.ghost.io/venture-capital-deeptech-investing/), [Peak](https://peak.capital) and deeptech VCs such as Forward.One or 4impact often act as early validators of what's strategically aligned with EU ambitions. If Brussels writes the rules and The Hague translates them, these players help decide ****which bets get made**. ## Final Thought: Mapping as Method This is still work in progress. I’m not a policy specialist but I believe that **mapping is a practical strategy tool**, especially for founders, consultants, and investors working in Europe. Understanding the layers of rules, funds, and institutions helps you move with the system rather than against it. In future posts, I’ll explore how this applies specifically to AI, public sector adoption, and startup positioning. For now, I hope this map offers others a way into the terrain. [*Let me know* ](https://rob-hoeijmakers.ghost.io/about/)*if you’re working in this space or mapping it too.* ### Further reading - [My Venture into Venture Capital: What I Learned About DeepTech Investing](https://rob-hoeijmakers.ghost.io/venture-capital-deeptech-investing/) - [Shaping Europe’s digital future](https://digital-strategy.ec.europa.eu/en) - [Europe’s DSA and the New Contours of Online Freedom](https://rob-hoeijmakers.ghost.io/europes-dsa-and-the-new-contours-of-online-freedom/) - [From Shopping Cart to Server Rack: How Lidl Builds a European Cloud](https://rob-hoeijmakers.ghost.io/lidl-stackit-european-cloud/) ### Playing the EU Game: Learning the Digital Rules by Getting in the Game URL: https://rob-hoeijmakers.ghost.io/playing-the-eu-game/ Last updated: 2026-05-10T08:54:00.000Z I’ve never been good at learning new games just by reading the rules. Whether it’s a board game or a policy framework, I tend to get stuck in the margins, trying to understand every nuance before I make a move. But as with most things, the best way to learn is usually to start playing. This also turned out to be true for understanding the European Union's digital strategy. ## From Tenders to Terrain Earlier this year, I decided to find out how our agency, [Schmuki](https://schmuki.nl), might get access to Dutch and European tenders related to digital innovation. We work in AI, automation, and digital transformation—mostly in complex, public-facing domains like education, health, and government. It seemed obvious that Europe would have a programme for this. But where to start? The EU is vast and opaque, even for someone who follows digital regulation closely. I knew about [GDPR](https://rob-hoeijmakers.ghost.io/gdpr/), the [AI Act](https://rob-hoeijmakers.ghost.io/learning-to-work-with-the-eu-ai-act/), the [Digital Services Act](https://commission.europa.eu/strategy-and-policy/priorities-2019-2024/europe-fit-digital-age/digital-services-act%5Fen). But I wanted to go deeper: who’s actually behind these policies? How is money distributed? Where do you go if you're an SME wanting to contribute? ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/06/EU-Tenders.jpg) It was fiddly but I got access to the [EU portal for funding and tenders](https://ec.europa.eu/info/funding-tenders/opportunities/portal/screen/home) and found the DIGITAL program. ## Finding the Ministry of Digital So I started mapping the EU’s online presence. Slowly, a pattern emerged: the URLs and subdomains follow a logic. Once I realised that the EU works with "Directorate-Generals" (DGs) that act like ministries, it became easier. One of them—[**DG CONNECT**, or “CNECT](https://commission.europa.eu/about/departments-and-executive-agencies/communications-networks-content-and-technology%5Fen)” in short—is responsible for communications, networks, content, and technology. If the EU has a ministry for digital and AI, **this is it**. DG CONNECT runs programmes like [**DIGITAL Europe**](https://digital-strategy.ec.europa.eu/en/activities/digital-programme) and co-funds local ecosystems such as [**European Digital Innovation Hubs (EDIHs)**](https://digital-strategy.ec.europa.eu/en/policies/edihs). There’s also funding for [**AI Factories**](https://digital-strategy.ec.europa.eu/en/policies/ai-factories), datasets, cloud infrastructure, and retraining initiatives. Suddenly, I wasn’t just browsing jargon—I was seeing strategy. Digital capability as industrial policy. AI as public infrastructure. [Communications Networks, Content and TechnologyThe Directorate‑General for Communications Networks, Content and Technology develops and implements policies to make Europe fit for the digital age.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon-2.svg)European Commission![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/ec-socialmedia-fallback.png)](https://commission.europa.eu/about/departments-and-executive-agencies/communications-networks-content-and-technology%5Fen) ## The Game Comes Into Focus As I started to follow the links and drill into the programmes, a structure emerged: - DIGITAL Europe isn’t just policy—it’s a **€7.5 billion fund**. - EDIHs aren’t just consultants—they’re **local entry points for SMEs** to access the EU ecosystem. - Projects aren’t only cross-border—they are **co-designed with member states**, often aligned with national missions. Now, instead of scanning endless PDF files with glazed eyes, I began to **ask smarter questions**: Which Dutch organisations are [listed as EDIHs](https://european-digital-innovation-hubs.ec.europa.eu/edih-catalogue)? Who's received funding in the past two years? What are the open calls on the [Funding & Tenders Portal](https://ec.europa.eu/info/funding-tenders/opportunities/portal/screen/home)? How do local and EU strategies align in AI for education or health? The game board was suddenly visible. The rules started to make sense. [Digitaal oplaadpunt voor het mkb- EDIH Digital Hub NoordwestEDIH Digital Hub Noordwest inspireert en stimuleert mkb-bedrijven die met digitale innovaties verder willen komen.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon-21.ico)EDIH Digital Hub Noordwest![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/EDIH-Header-02.jpg)](https://edih-dhnw.nl) ## Rinse, Repeat — and Play There’s still a lot I don’t know. But I’ve realised that learning to work with the EU is **not about waiting until you understand everything**. It’s about playing *while* learning. Reading the terrain, talking to partners, mapping the namespace. Acting, then adjusting. That doesn’t mean I’m sold on the EU’s tempo or structure. The orderliness, the layered policies, the public dashboards—none of it feels remotely close to the speed or informality of Silicon Valley or Shenzhen. The rules here are visible, often admirable, but also heavy. And the gap between regulation and execution remains large. Still, we don’t play by the rules just because we must. There’s a sincere belief—at least in our work—that there’s value in a **more ethical, democratic way of shaping technology**. Even when it slows things down. Even when it constrains entrepreneurial freedom. This is the board we're given. So yes, we play by the rules. But we move fast, we experiment, and we build. That’s how you stay in the game—and how, maybe, you win. ### Further reading - [Understanding Digital Identity in the EU: eID, eIDAS, and the EUDI Wallet](https://rob-hoeijmakers.ghost.io/eu-eid-eidas-eudi-wallet/) - [Draghi report: How Europe’s Digital Lag threatens innovation](https://rob-hoeijmakers.ghost.io/europe-between-innovation-and-regulation/) - [Shaping Europe’s digital future](https://digital-strategy.ec.europa.eu/en) ### WhatsApp is Quietly Becoming a Platform — And That Changes Everything URL: https://rob-hoeijmakers.ghost.io/whatsapp-platform-shift-contact-identity-strategy/ Last updated: 2026-05-10T08:54:00.000Z It started with a contact I didn’t recognise. I tapped the number, as I often do, expecting the usual options. But this time, something was different. A new screen appeared: *“More ways to manage your contacts in WhatsApp.”* For the first time, I was asked: *Do you want to save this contact to your phone—or just to WhatsApp?* That small prompt stopped me in my tracks. Because what I was looking at wasn’t just a new UX detail. It was a strategic shift. WhatsApp is no longer just a chat app tied to your phone’s address book. It’s becoming a platform—one that stores identity, relationship context, and history independently from your device. This has deep implications for privacy, account separation, and how we design digital communication workflows. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/06/WhatsApp-Contacts-1.jpg) The message on managing contacts in WhatsApp and the new functions. ## Strategy in the Smallest Places Sometimes strategy reveals itself not through press releases, but in the interface. WhatsApp is rolling out a feature that lets you store contacts directly **in the cloud**, connected to your WhatsApp account rather than your phone’s contacts. And you can now manage these contacts across devices—including WhatsApp Web and the desktop app. It’s subtle. But it’s a massive move. No more forced syncing between WhatsApp and your system address book. No more blurring of lines between business and personal contacts. And crucially: **your contact list is now part of your digital identity**—portable, cloud-restorable, and device-independent. ## A Shift in Identity Infrastructure WhatsApp calls the underlying technology “IPLS”—Identity Proof Linked Storage. In practice, it means: - You can save a contact **exclusively in WhatsApp**. - You can choose whether or not to sync it to your phone’s contacts. - You can restore that list if you lose your device or change numbers. This is no longer just “contact storage.” This is identity-aware contact management. And when WhatsApp soon allows usernames instead of phone numbers (which it has hinted at), this will become even more profound. We’re moving toward **profile-based interaction**, not just phone-based messaging. ## What This Enables Strategically This change opens up new, smarter ways of using WhatsApp in digital strategy: - **Contact segmentation**: Keep personal contacts in your phone, business ones in WhatsApp. No more messy overlaps. - **Multi-role communication**: Use different WhatsApp environments (personal vs business) on the same device, each with their own contact layer. - **Privacy-first messaging**: Avoid exposing your phone number for all interactions—especially important in customer support, research, or activism. - **Team-based operations**: In WhatsApp Business, contact lists are now shared across devices and users. Think small-team CRM without extra tools. ## From App to Platform For years, WhatsApp’s power was its simplicity. But now, it’s evolving. What started as a mobile app tied to your [SIM card](https://rob-hoeijmakers.ghost.io/from-physical-sim-to-esim-on-iphone/) is now a cross-device, identity-aware, privacy-optional communications platform. Meta knows that future messaging tools aren’t just about chat—they’re about *relationship infrastructure*. This contact update is just one move in that direction. But like all platform shifts, it starts small. One new prompt. One toggle. One number you didn’t expect—until suddenly, you realise the logic behind your apps is changing. And with that, the strategies we design with them. ### Further reading - [WhatsApp Messenger and Business on a single iPhone](https://rob-hoeijmakers.ghost.io/whatsapp-messenger-and-business-app-on-the-same-iphone-with-a-virtual-number/) - [WhatsApp in 2026: From Phone Numbers to Names](https://rob-hoeijmakers.ghost.io/whatsapp-in-2026-from-phone-numbers-to-names/) - [Three ways to use WhatsApp via computer](https://rob-hoeijmakers.ghost.io/three-ways-to-whatsapp-via-computer/) ### ChatGPT Connectors: The Emergence of a New Interface Layer URL: https://rob-hoeijmakers.ghost.io/chatgpt-connectors/ Last updated: 2025-06-06T06:26:05.000Z The recent launch of **Connectors** in ChatGPT Business and Enterprise plans marks a pivotal shift in how we interact with large data stores. With a quiet rollout that deserves more attention, OpenAI is redefining what it means to “chat with your data”—this time, not just through uploads or memory, but via **direct integrations** into tools like Google Drive, OneDrive, SharePoint, Slack, Salesforce, Notion, and more. It’s early days, and the dust hasn’t settled. But already, the implications are striking. ## What Are ChatGPT Connectors? Connectors act as live data bridges between ChatGPT and third-party sources. If you’re used to tools like NotebookLM, or if you’ve worked with custom RAG (retrieval-augmented generation) setups via APIs or plugins, this feels like a native, out-of-the-box version of that concept—**MCP-like pipelines** for mainstream users. Think of connectors as permission-based extensions. Once enabled (on business-tier accounts), ChatGPT gains real-time query access to selected data silos—no need to upload, no need to preprocess. From a UX standpoint, it feels like talking to a hyperintelligent assistant who can, say, read all your Notion docs, access your latest project folders on Drive, or summarise your last five HubSpot deals. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/06/Google-Calendar-connector.jpg) Connector to Google Calendar. ## What’s Actually New Here? On the surface: seamless access. Under the hood: possibly [RAG](https://rob-hoeijmakers.ghost.io/exploring-retrieval-augmented-generation-rag/), layered access control, and a completely different mental model. ChatGPT moves from being a discrete, self-contained assistant to a kind of **context-aware gateway**—sensitive to documents, schedules, and histories it couldn’t see before. The shift is conceptual as well as technical. Where we once prompted ChatGPT with all we could fit in a single session (or document upload), we’re now stepping into a space where the **model pulls from dynamic, evolving knowledge bases**. ⚠️ Connectors in ChatGPT are available to users on Plus, Pro, Team, Edu, and Enterprise plans, but regional restrictions apply. For users in the EEA, Switzerland, and the UK, most connectors—including Dropbox, Box, OneDrive, and SharePoint—are ****not currently available** to those on the Plus and Pro plans, regardless of whether these connectors are listed as supported. Only Team, Edu, and Enterprise users in those regions have access. GitHub is the exception and is available to all users globally, including those in restricted regions. ## Opportunities—and Overwhelm The ability to perform deep, unstructured research across your cloud environments is incredible. But it's also disorienting. > “I always saw ChatGPT as a standalone tool. That’s what I liked about it, now suddenly, stuff flows in left and right—it seems. That’s the thing I need to get my head around.” This feeling is real, especially for users who appreciate clarity and modularity. With connectors, the scope of the assistant expands beyond what’s visible in a chat window. There’s power in that—but also opacity. Where’s the boundary? What’s cached, indexed, live? These questions aren’t yet answered clearly in OpenAI’s UX or documentation. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/06/ChatGPT-Connector-Google-Drive.jpg) ChatGPT Connector working with DeepResearch and Google Drive ## Use Cases Emerging The most immediate value comes from: - **Large-scale document querying** – Instead of uploading PDFs or using external tools like NotebookLM, you can now ask questions across your entire Drive or SharePoint. - **Sales & CRM automation** – Imagine surfacing customer insights in natural language from within your HubSpot or Salesforce accounts. - **Meeting & knowledge summarisation** – Using connectors to pull context from Notion, Slack, or Google Calendar to generate summaries, next steps, or briefs. In short: Connectors blur the line between “AI assistant” and “business operating system.” ## A New Interface Layer It’s worth stating: this isn’t just a new feature. It’s the emergence of a **new interface layer**—one that makes our cloud-based lives machine-readable, conversational, and dynamically retrievable. The implications are architectural. This could rival what Microsoft is doing with Copilot and Graph-based integrations. The key difference? ChatGPT connectors don’t tie you to a specific productivity stack. They’re intentionally broad and flexible. ## What Comes Next? OpenAI will likely expand connectors beyond the current business-only boundary. A natural evolution would include: - **Connector chaining** (e.g. “Fetch customer info from HubSpot, cross-check calendar availability, then draft an email”) - **Fine-grained data permissions and previews** - **User-managed caching or data snapshots for speed & control** For now, experimentation is key. Clip what you see. Document behaviours. And prepare to rethink what it means to query. I give a quick tour on ChatGPT and Connectors ### Google's AI for Individuals and SMBs: Clarity, Costs, and Strategic Choices URL: https://rob-hoeijmakers.ghost.io/googles-ai-for-individual-smb/ Last updated: 2026-05-10T08:54:00.000Z ## Introduction: Untangling the Options Google's AI offerings can appear deceptively simple, but the reality is far more complex. Over the past few months, I’ve examined how individuals and small businesses (SMBs) can access the best of Google’s AI tools—particularly Gemini Advanced. This article breaks down the two main paths: Google One (for individuals) and Google Workspace (for businesses), including how they compare in features, pricing, and practical value. [Google One vs Google Workspace: Where do you get real AI and Gemini Advanced?What do you get when you pay for AI at Google? An exploration of Gemini Advanced, NotebookLM and how to share it smartly via Google One.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-103.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_1822-768x576.jpeg)](https://rob-hoeijmakers.ghost.io/google-one-vs-google-workspace-for-gemini/) ## Google One AI Premium: Personal-Use Simplicity For individuals, Google offers the **Google One AI Premium** plan at €22/month. This includes: - Access to **Gemini Advanced**, Google's most powerful AI model - Usage inside tools like Gmail, Docs, and other Google apps - **Family sharing** until June 2025 (note: this may change) - Access to **NotebookLM**, a powerful research and synthesis tool This setup works well for personal productivity and experimentation. It also gives access to [NotebookLM](https://rob-hoeijmakers.ghost.io/from-pages-to-podcasts/), which may be the strongest proposition in Google's AI suite. However, it lacks the collaborative infrastructure needed in a business environment. There are also limited integration points with external AI services or enterprise tools. [Google One subscription for Google MeetTo use the premium features of Google Meet, like longer duration, I choose Google One as subscription instead of Google Workplace. Here is why.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-104.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_6369-2.jpeg)](https://rob-hoeijmakers.ghost.io/choosing-google-one-to-use-google-meet/) ## What Is Google Workspace? **Google Workspace** is Google's business productivity suite, formerly known as G Suite. It includes: - Gmail, Docs, Sheets, Calendar, Meet, and Drive - Admin controls, custom domains, and user management - AI features integrated into each core app (if using the right tier) - Most importantly: **shared Drives** for centralised team storage Workspaces are priced per seat, typically starting from €6 to €18/month depending on the plan. To get advanced AI features—including Gemini for Workspace apps like Gmail and Docs, and access to Gemini Advanced—you need to choose a plan or add-on that includes Gemini. Google currently offers this as an AI-enhanced Workspace tier, not as a separate "add-on" SKU. Pricing for these AI-enhanced tiers starts at around €13/month per seat (on annual commitment), in addition to the base Workspace cost. ## Why Workspace Might Offer Extra Value While Google One Premium focuses on individual users, **Workspace unlocks more advanced workflows**: - You get **shared Drives**, which integrate well with **ChatGPT Team** - You gain control over user permissions, file access, and collaboration - It may offer small branding and domain benefits for client-facing communications In our case, this integration with ChatGPT Team was the trigger to explore Workspace. It added practical value that One Premium alone could not. And yet—this value remains highly contextual. ## Reconsidering the Bundle: A Personal Perspective Despite recognising the functional benefits, I must admit: I strongly dislike bundles. Over the years, they have led to lock-in scenarios that reduced flexibility and slowed down decision-making. When my [business partner](https://schmuki.nl) challenged me to explain why we needed Workspace, I struggled to offer a compelling answer: - Gmail? Not essential. - Google Drive? Useful, but not unique. - Domain branding? A minor benefit. And the AI? Gemini, though fast-improving, still doesn’t feel as capable or responsive as ChatGPT. The only compelling argument for Google's AI right now might be **NotebookLM** — but you get that with Google One Premium too. ## Pricing Strategy: Why Is Google's AI So Cheap? Compared to ChatGPT Plus (€20) or ChatGPT Team (€25–€30+ per seat), Google's Workspace AI pricing is unusually low. Why? - Google treats AI as a **feature of an ecosystem**, not a product in itself - It’s using low pricing to **accelerate adoption**, particularly in the SMB market - Google can rely on its **infrastructure scale (TPUs, data centres)** to absorb AI costs better than smaller providers Essentially, Google isn’t trying to profit from AI seats directly—it wants you (and your team) in the Google ecosystem. ## Final Verdict: Context Matters For **individual users**, Google One Premium is the simpler choice and gives you access to Gemini Advanced and NotebookLM without any long-term commitment. For **small teams and SMBs**, Workspace with the AI add-on can be affordable and more capable—but only if you actively benefit from Drive integrations, team workflows, or domain branding. Yet if you're wary of vendor lock-in and already use ChatGPT heavily, Workspace may feel redundant. My own take? Use Workspace if the integrations are mission-critical. Otherwise, stay lean with Google One and remain flexible. In a follow-up post, I’ll explore how the [Google Drive + ChatGPT Team integration works in practice](https://rob-hoeijmakers.ghost.io/chatgpt-connectors/). That, too, was interesting, but only if you're all-in on connected workflows. *Note: Pricing reflects EU conditions as of mid-2025 and may vary by country or billing cycle.* ### Further reading - [The Workspace Advantage: Google’s Quiet Lead in the AI Shift](https://rob-hoeijmakers.ghost.io/google-gemini-workspace-future/) ### Beyond LLM’s: The Hidden Dangers of AI in Scientific Research URL: https://rob-hoeijmakers.ghost.io/ai-threat/ Last updated: 2026-05-10T08:54:00.000Z ## Realising My Misplaced Attention on the type of AI I genuinely believed I understood the risks associated with artificial intelligence. Like many others, my focus had been heavily on large language models, such as ChatGPT and Claude—tools I interact with daily and greatly value. I thought these sophisticated chatbots embodied AI's biggest threats, from misinformation to the potential displacement of countless jobs. But recently, something shifted dramatically in my perspective. ## An Eye-Opening Insight from Niall Ferguson My realisation came after reading an [enlightening interview with historian Niall Ferguson](https://www.noemamag.com/america-is-in-a-late-republic-stage-like-rome/), a respected expert known for connecting historical events with contemporary geopolitical strategies. Ferguson clearly explained something I'd missed: while language models capture headlines and public imagination, they're only a small part of AI’s broader landscape. The real threat is something quieter, highly specialised, and genuinely alarming. > **“Mankind is in the process of creating an intelligence it cannot fully understand, and whose capacities it may not be able to control.”** > — [*Henry Kissinger, Genesis*](https://www.hachettebookgroup.com/titles/henry-a-kissinger/genesis/9780316581295/)*: Artificial Intelligence, Hope, and the Human Spirit* ## The True, Hidden Power of Specialised AI According to Ferguson, the critical threat lies in AI systems explicitly engineered for research, strategic, and military uses: - AI designed to create highly potent viruses, potentially more lethal and infectious than any naturally occurring pathogen. - Autonomous weapon systems capable of identifying and engaging targets more rapidly and accurately than human operators. - AI-driven platforms that optimise nuclear strategy, geopolitical analysis, and real-time decision-making at a frighteningly precise level. > “It’s the power of the scientific AI that should worry us.” These aren't speculative scenarios or distant threats; they already exist. Backed by massive computational resources, extraordinary financial investments, and protected by stringent secrecy, these systems directly influence global power dynamics. What makes them especially dangerous is the Faustian dynamic: even if only states and multinationals have the resources to build them, the knowledge and tools they create can leak, mutate, or be repurposed—putting catastrophic capabilities within reach of much smaller actors. This is the contradiction Henry Kissinger foresaw. While nations fight to control compute, the real instability may come from diffusion: scientific AI that outpaces the institutions meant to contain it. ## Strategic AI and the Rise of Authoritarian Power Ferguson underscores the critical advantage authoritarian states, particularly China, gain through strategic AI deployment. Combining vast surveillance data with centralised political control and immense computing power, these nations create powerful tools for both domestic control and international influence. Democracies face a substantial challenge responding effectively to this concentrated power. ## Why This Matters for the Future of Geopolitics This insight profoundly reshaped my understanding of geopolitical competition. While public debate fixates on regulating chatbots or tackling AI-generated content in social media, the real and critical contest unfolds quietly, hidden from view. It's a competition around infrastructure, energy consumption, computational power, and strategic ambition. But it’s also a contest of restraint: the race to develop such tools is outpacing our capacity to imagine their consequences—or to govern their misuse. > I'M SHOCKED. > > While researching for my podcast, I realized China isn't just competing with the USA in AI. They're positioning to completely dominate us. > > Everyone thinks America leads in AI. But China is building the energy infrastructure to crush us. > > Here's what scares me: > > — Nathan Lands - Lore.com (@NathanLands) [May 28, 2025](https://twitter.com/NathanLands/status/1927696391209628072?ref%5Fsrc=twsrc%5Etfw) ## Shifting My Focus Ferguson's insights fundamentally challenged my earlier beliefs, prompting me to realign my attention from consumer-level AI to the specialised, strategically significant systems quietly shaping global power today. It’s a confronting shift—one that has left me both more curious and more uneasy. Over the coming weeks, I’ll be updating my reading list to reflect this new focus: not just books about LLMs and ethics, but deeper works on AI’s role in science, security, and strategy. If we want to understand where this is going, we need to look well beyond the chat window. --- ### Further reading - [America Is In A Late Republic Stage Like Rome | NOEMA](https://www.noemamag.com/america-is-in-a-late-republic-stage-like-rome/) - [Kissinger — NIALL FERGUSON](https://www.niallferguson.com/kissinger) - [Genesis](https://www.hachettebookgroup.com/titles/henry-a-kissinger/genesis/9780316581295/) ### Will I Manage Agents as I Would Employees? URL: https://rob-hoeijmakers.ghost.io/manage-agents-employees/ Last updated: 2025-05-15T15:00:34.000Z As businesses rapidly integrate digital agents, powered by AI, GPTs, and automated systems, leaders are grappling with new questions: Can these digital agents simply slot into our existing organisational structures, or do they require entirely new management approaches? ## A New Type of Workforce I'm not a Corporate Enterprise. I'm not leading thousands of employees. I'm an entrepreneur, a digital strategist, experimenting with [digital agents ](https://rob-hoeijmakers.ghost.io/agents-and-agency/)daily. And here's what I see clearly: digital agents are increasingly acting like team members rather than mere tools. For instance, I use GPT-based agents to write, analyse, summarise meetings, and organise tasks. They're reliable, available around the clock, and relentlessly consistent. But does that mean I manage them like employees? 💡 By the way, AI does three things with work and employees. It makes them/us redundant, it changes the work and it creates new work. I only see the latter happening moderately so there is still a lot of change (and unrest) waiting for us. To be continued. ## Managing Similarities: Clear Tasks, Clear Results Like human team members, digital agents need clear tasks and expectations. Ambiguity isn't good for AI, just as it isn't good for people. Defining clear objectives, KPIs, and tasks helps digital agents excel, much as it helps human employees thrive. ## Where Management Differs: Motivation and Learning Yet, there are critical differences. Digital agents don't need motivation or morale boosting. They don't seek growth opportunities or emotional support. However, they require careful and continuous training, often supervised, to ensure accuracy, fairness, and relevance. Unlike humans, agents don't naturally learn from experience. They must be explicitly trained, retrained, and guided. ## Organisational Shifts Consider [Moderna’s recent shift](https://www.wsj.com/articles/why-moderna-merged-its-tech-and-hr-departments-95318c2a), merging HR and Tech departments. This move signals a profound realignment around the combined value of humans and digital agents. [Why Moderna Merged Its Tech and HR Departments | Isabelle Bousquette | 33 commentsIn my latest for The Wall Street Journal, I delve into why vaccine-maker Moderna decided to merge its technology and HR departments. I've heard a lot of tech leaders say they're working more closely than ever with HR as they think through the implications that genAI will have on the workforce (a lot of companies have cited genAI efficiencies as a reason for job cuts, for example). But Moderna's decision to merge the two roles takes things one step further. Tracey Franklin, who joined Moderna as its HR chief and is now its chief people and digital technology officer, told me she is using the opportunity to redesign teams across the company based on what work is best done by people versus what can be automated with technology, including the tech it leverages from a partnership with OpenAI. Roles are being created, eliminated and reimagined as a result, she said. Real the full story here for exactly how Moderna is using OpenAI tech internally and what kind of impact it's having on the workforce there: https://lnkd.in/en7fsehw And then let me know what you think. Would it make sense to merge tech and HR at your organization? | 33 comments on LinkedIn![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/al2o9zrvru7aqj8e1x2rzsrca-2)LinkedInIsabelle Bousquette![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/1747049108662-2)](https://www.linkedin.com/posts/isabelle-bousquette%5Fwhy-moderna-merged-its-tech-and-hr-departments-activity-7327694207408553985-4SmI?utm%5Fmedium=ios%5Fapp&rcm=ACoAAAAGPccBqY4Rvlcj2zZ2PdvOk7tl5dQrkUs&utm%5Fsource=social%5Fshare%5Fsend&utm%5Fcampaign=copy%5Flink) Professor Venkat Venkatraman calls this the shift into the "agentic era"—[organisations designed around the combined strength of biological and digital agents](https://www.linkedin.com/posts/nvenkatraman%5Fwhy-moderna-merged-its-tech-and-hr-departments-activity-7327711474485796865-kWkC?utm%5Fsource=share&utm%5Fmedium=member%5Fios&rcm=ACoAAAAGPccBqY4Rvlcj2zZ2PdvOk7tl5dQrkUs). ## My Perspective From my position, managing agents feels less like traditional management and more like orchestration. It's about clearly defining roles, ensuring seamless integration between humans and AI, and continually refining processes. Digital agents won't replace management, they will redefine it. Leaders will become architects of hybrid teams, curators of collaborative ecosystems, and continuous optimisers of workflows. ## Looking Forward As more businesses adopt digital agents, these distinctions will become central. Managing digital agents effectively means understanding their unique needs and integrating them into an agile, adaptive organisational structure. Will we manage digital agents exactly like human employees? Probably not. But we'll certainly manage them with equal care, strategic clarity, and deliberate intent. After all, agents, may they be human or digital, work best when given clear purpose, clear tasks, and clear boundaries. And that's something every good manager already knows. ### Markdown in Google Docs: A Practical Step Forward URL: https://rob-hoeijmakers.ghost.io/markdown-google-docs/ Last updated: 2025-05-15T07:23:40.000Z In 2022, I [wrote about Markdown](https://rob-hoeijmakers.ghost.io/using-markdown-for-text/) and how it simplified my blogging life, turning writing from a formatting hassle into pure, clear communication. Since then, my appreciation for Markdown has grown even deeper—it's now my universal standard for all textual communication. ## Markdown: Simple Harmony for Your Words Imagine you're at a concert. Before the performance begins, the orchestra tunes itself—each musician listens carefully, adjusting their instruments to a single, clear note. Markdown is that note. It's simple, clean, and universal enough for everyone to "tune" their words to, ensuring clarity and harmony wherever those words travel. Markdown sits elegantly between richly formatted documents (like Word, Google Docs, presentations, or PDF documents) and plain text. It allows basic formatting—like **bold**, *italics*, [links](https://example.com), images, and headings—without complicated tools or menus. It’s as straightforward as jotting down notes but structured enough to communicate clearly and effectively. ## Why Markdown Matters More Than Ever Today, more platforms support Markdown, making it an even stronger bridge between different digital environments. Writing apps like [Ulysses](https://rob-hoeijmakers.ghost.io/blog-writing-workflow/) let you publish directly to platforms such as [Ghost](https://rob-hoeijmakers.ghost.io/ghost-blog/) or [WordPress](https://rob-hoeijmakers.ghost.io/how-about-wordpress/) seamlessly. Recently, Google's announcement that [**Google Docs**](https://workspaceupdates.googleblog.com/2024/07/import-and-export-markdown-in-google-docs.html?utm%5Fsource=twitter&utm%5Fmedium=unpaidsoc&utm%5Fcampaign=FY25-Q2-amer-googledocs%5Fsupportpage%5FProduct-Education&utm%5Fcontent=-&utm%5Fterm=-) **now fully supports importing and exporting Markdown** marks a significant milestone. You can effortlessly copy text into and out of Google Docs using Markdown, ensuring your content remains consistent and clear no matter where it goes next. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/05/Google-Docs-MarkDown.jpg) Copy as MarkDown and Paste from MarkDown in Google Docs Markdown makes particular sense when transferring richer formats—like presentations or PDF documents—to environments where markup and structure must be maintained, especially when publishing to the web. By converting to Markdown, essential structure is retained without the complexities and inconsistencies introduced by directly transferring richer formats. The beauty here isn't just simplicity; it's about control and portability. Your content isn't trapped in a particular format or tied to a specific app. You’re free to move ideas fluidly from notes to blogs, emails to documents, without losing their structure. ## Markdown as the Universal Note This flexibility turns Markdown into the digital equivalent of a tuning fork—a single, pure tone that lets everyone involved "tune up" their content. Whether you're collaborating with colleagues, exchanging ideas across different platforms, or even communicating with AI tools (like ChatGPT’s canvas, which fluently uses Markdown), your words resonate clearly and precisely. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/05/Markdown-in-ChatGPT.jpg) ChatGPT Canvas lets you share content as MarkDown. ### How I Use Markdown Today Today, my workflow primarily revolves around three tools: - **ChatGPT:** For initial writing support and content generation. - **Ulysses:** For refining content and publishing directly to the web. - **Google Docs:** For richer formatting, collaborative editing, and sharing content widely within teams or presentations. Markdown seamlessly bridges these tools, allowing me to move effortlessly between creation, refinement, and distribution. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/05/image-6.png) Ulysses is my writing and publishing tool and it is uses MarkDown. Texts can be published to Ghost and Wordpress. ### Choosing ChatGPT Team: From AI-Confidant to Trusted Colleague URL: https://rob-hoeijmakers.ghost.io/choosing-chatgpt-team-from-ai-confidant-to-trusted-colleague/ Last updated: 2026-05-10T08:54:00.000Z Recently, my [business partner and I](https://schmuki.nl) migrated from our individual ChatGPT Plus accounts to a consolidated ChatGPT Team workspace. Our motivations were straightforward: improved collaboration, streamlined billing, and—critically—enhanced data privacy assurances. The process, however, provided valuable insights into what such migrations entail, particularly regarding contractual clarity and data management. 💡 **Update of August 29, 2025: ChatGPT Team is now called* ***ChatGPT Business** *. Features and prices have not changed.* ## Why We Decided to Migrate Initially, the appeal of better cooperation and the simplicity of a single, annual billing cycle were major factors. Yet, privacy was perhaps the most significant driver. With personal ChatGPT Plus accounts, data privacy assurances are minimal, as user conversations could potentially be used to train OpenAI's models. Migrating to a business-oriented ChatGPT Team account provided a distinct advantage: clear contractual guarantees from OpenAI that our data would not be used for model training. Additionally, having a formal business contract through a dedicated corporate entity added confidence and professional legitimacy to our interactions. 💡 ****ChatGPT Plus is designed for individual use**, with the contract held personally and data potentially used to train models unless this is disabled in settings. It offers no admin controls or team-level data oversight, making it ideal for personal productivity and experimentation. ****ChatGPT Team, on the other hand, is intended for professional or organisational use**, with the contract held by the business. It includes team seat management, admin roles, and most importantly, data is **not* used to train models by default. It offers enhanced privacy, collaboration tools, and priority support—making it the appropriate choice when working with sensitive or proprietary information, or when compliance and team governance are required. ## Expectations vs. Reality We assumed our personal data, custom instructions, and chat histories from ChatGPT Plus would smoothly integrate into the Team workspace. Instead, we encountered a reality check: the Team environment started completely fresh, requiring manual migration of essential content. Another unexpected discovery was the billing structure: the personal workspace and the professional workspace remain distinct, each billed separately. While logical in hindsight, it initially felt counterintuitive and increased our anticipated expenses. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/05/image-1.png) You need to cancel the Plus account yourself, I advise to use a grace period. ## Practical Considerations in Migration The migration required manually transferring our most valuable memories and custom instructions, emphasizing the importance of dedicating sufficient time and planning for the transition. One practical recommendation for anyone considering such a migration is to implement a grace period where the personal and team accounts overlap. This dual subscription, while seemingly expensive, safeguards against losing essential data and allows time for a careful transition. Importantly, this overlap period mitigates a privacy risk: once your ChatGPT Plus subscription reverts to a free tier, there's a possibility that historical conversations might become available for AI training—something explicitly prevented by the business-oriented Team account. ## Psychological Shifts: From Intimate Companion to Professional Colleague An unexpected yet insightful outcome was the psychological transition. Over time, I'd inadvertently formed a more personal, even intimate connection with ChatGPT, perceiving it as a trusted confidant. Shifting to a clearly delineated business workspace helped recalibrate my perception: ChatGPT evolved into a reliable, professional colleague, distinct from personal engagement. This transition proved healthy, clarifying boundaries between personal and professional roles, reinforcing the value of separating private reflections from professional interactions, and ensuring both privacy and purpose were respected. ## Looking Ahead: Privacy as a Priority While ChatGPT Team addressed our immediate business needs, I still hold out hope for AI solutions offering deeper privacy guarantees—tools where I can comfortably share more personal reflections. Recognizing that OpenAI’s primary focus with its business offerings is collaborative productivity rather than personal intimacy helps manage expectations effectively. ## Key Lessons Learned From this migration experience, several key lessons emerged: - **Privacy Assurance:** Business-oriented ChatGPT Team accounts offer explicit privacy guarantees, critical for sensitive business interactions. - **Manual Migration:** Prepare to invest time and resources into manually transferring crucial data. - **Billing Clarity:** Understand and anticipate separate billing for personal and professional workspaces. - **Grace Period:** Consider overlapping subscriptions temporarily to protect your data integrity and privacy. - **Psychological Adjustment:** Be aware of emotional attachments and actively embrace the professional shift. Sharing these insights, I hope to guide others through their own digital migrations—emphasizing privacy, cooperation, and clarity. --- **Update**, two days after new ChatGPT subscription I got to see a wizard on the web version offering to transfer the chat history and GPT's: ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/05/image-3.png) Your ChatGPT Team workspace is ready. ### Further reading - [From Free to Paid: Choosing the Right AI Model (with a European Lens)](https://rob-hoeijmakers.ghost.io/ai-models-europe-free-vs-paid/) ### How OpenAI o3 Became a Geospy and Found My Photo’s Location URL: https://rob-hoeijmakers.ghost.io/openai-o3-geospy/ Last updated: 2026-05-10T08:54:01.000Z I wanted to test something I thought only humans were good at: being a visual detective. Think OSINT, think “geo spy” work—where you inspect a photo in detail and try to figure out where it was taken. You look at signage, prices, language, colours, and clues in the background. I gave this challenge to OpenAI’s new **o3 model**—an official release that represents a distinct step forward beyond GPT-4, available to all ChatGPT users, with advanced features reserved for Plus and higher-tier users. What followed surprised me. ## Watch the Video Here’s a short, narrated demonstration. I gave ChatGPT O3 a photo I took in winter—nothing spectacular, seemingly generic, and not instantly recognisable. Then I asked it to find out **where it was taken**. Here I show how it works and the process o3 follows to find the location of a photo. ## What Happened The Open AI o3 model took **5 minutes and 42 seconds** to analyse the image and narrate its internal thought process. Here's a breakdown of how it tackled the task: ### 1\. Initial Hints It spotted a chestnut vendor marked *Maroni*, associated with Switzerland or Italy. Prices were listed in **Swiss francs**, and the **Coop** logo pointed clearly toward Switzerland. ### 2\. Visual Forensics The model zoomed into tiny elements: - **A Swiss-looking licence plate** - The **windows of a bus or trolley** - A **restaurant sign**, possibly referencing “Orient” - A **vertical LED sign**, possibly from **Credit Suisse**, **BCV**, or **UBS** ### 3\. Online Agent Behaviour Then the agent decided to go online: It searched terms like *John Gut Maroni*, finding a vendor near **Biel Bahnhofplatz**. It cross-referenced this with: - **Google Images** - Historical **price lists** - **Car park signage** and LED display clues - The bilingual signage and Bernese number plates ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/05/image.png) The model o3 reads the price board, sees a name, assumes it is the owner and decides to do a web search. ### 4\. Conclusion It triangulated the evidence and **correctly identified** the location as **Bahnhofplatz in Biel/Bienne**, in the Swiss canton of Bern. ## Why This Matters This isn’t just about geolocation. The real takeaway is the **agentic behaviour**: - The model zooms, interprets, compares, and decides. - It **acts like an assistant or detective**, rather than a passive respondent. - It combines **multimodal input**, **web browsing**, and **reasoning over time**. This opens up exciting use cases—from autonomous research to visual diagnostics and interactive agents capable of complex tasks. ## Try It Yourself This kind of exploration is not just entertaining—it’s a preview of what’s becoming possible with language-image agents. If you want to experiment: - Use a **complex image** with a mix of visual cues. - Let the model **think out loud**. - Encourage it to use **external tools or web search**. It’s like watching a mind unfold. 💡 ****GPT-4 vs. o3: How a Simple Experiment Reveals the Difference** I made up a proverb and asked GPT-4 to explain it—GPT-4 confidently gave a detailed explanation, despite it being completely invented. Repeating this with OpenAI’s advanced o3 model, it actively researched online, realised the proverb didn’t exist, and clearly explained why. In short: GPT-4 fluently creates plausible narratives; o3 carefully verifies if they’re true. --- ### Further reading - [Lost in ChatGPT’s Models? A guide (March 2025)](https://rob-hoeijmakers.ghost.io/chatgpt-models-guide/) ### Should I Opt Out of Meta’s AI Training? URL: https://rob-hoeijmakers.ghost.io/should-i-opt-out-of-metas-ai-training/ Last updated: 2026-05-26T19:38:28.000Z ## I’m still not sure Meta is now asking permission to use our public data to train its AI models. That alone is worth pausing over. For a company with a long history of taking rather than asking, this shift feels less like a moral turn and more like a legal necessity—probably tied to the upcoming EU AI Act. But whatever the reason, the moment we’re in is real. They’re asking. You can opt out. And I’ve found myself circling around a single question: *Should I?* [Instagram Help Center![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/WjbGH9Owc-X.ico)](https://help.instagram.com/contact/233964459562201) ## My hesitation is human On the one hand, it seems like a small and obvious act of resistance. Why would I feed a system I don’t fully understand—especially one likely built for profit, not people? But on the other hand, I hesitate. Not for principled reasons, but because I’m afraid. What if opting out quietly hurts [my reach on Instagram](http://instagram.com/robhoeij/)? What if it excludes me from something emerging—something I don’t even know I’m missing? Will my content be deprioritised? Will my presence be weakened? These are selfish fears, maybe. But they’re not irrational ones. Because Meta doesn’t tell us what the consequences of opting out actually are. That opacity—intentional or not—turns a simple decision into a strange ethical gamble. ## What exactly are we giving? And still I wonder: what kind of contribution are we being asked to make? Is this just about the content I post—photos, captions, stories—or is it also about the network I inhabit? Are they training AI not just on what I *say*, but on who I *know*, what I *like*, how I *respond*? There’s **a finality to this** that makes it different from earlier data collection. Once your data has been used to train a model, it can’t be pulled back out. It’s baked into the neural fabric of the system. That makes this not just a technical or legal decision—it’s a cultural one. A form of surrender, or contribution, depending on how you look at it. 💡 ****Should I opt out? I wasn’t even sure what I’d be opting out** ***of** **.** Meta’s vague language didn’t clarify what “AI training” really meant. But I realised: once your data is absorbed, you don’t get to reach it again. So I opted out. ## The ethics aren’t abstract What’s frustrating is that I don’t know what Meta plans to do with the AI it trains on us. I’d like to think it might bring something good into the world—more accessible tools, more creative collaboration, better interfaces. But if I’m honest, I suspect **the gains will be theirs, and the risks will be ours**. And this matters, because social platforms aren’t neutral pipes. They’re cultural spaces. What we create, share, celebrate, and discover on them shapes the tone of our time. If those spaces become dominated by synthetic content—optimised for engagement, but void of origin—we lose something. Not everything. But something real. And if they’re using our authentic contributions to build that synthetic future, shouldn’t we at least know what that trade-off looks like? ## This is me thinking out loud I’m writing this because I don’t have a clear answer. I haven’t (yet) opted out. But I might. I’ve been moving between suspicion and curiosity, frustration and optimism. I’d rather not become cynical. I’d rather stay engaged. But I don’t want to drift into passive complicity either. So I’m asking: Have you opted out? Will you? Do you know more than I do? Can we still have this conversation—before the training is done and the inputs are locked forever? [Let’s talk](https://rob-hoeijmakers.ghost.io/about/). Not to panic. Not to point fingers. Just to stay human in the loop. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/05/image-2.png) In the mean time I opted out. ### My Venture into Venture Capital: What I Learned About DeepTech Investing URL: https://rob-hoeijmakers.ghost.io/venture-capital-deeptech-investing/ Last updated: 2026-04-24T17:21:47.000Z Earlier this year, I met [Patrick Claessen](https://www.linkedin.com/in/patrick-claessen/) from [Cottonwood](https://www.cottonwood.vc) at an event in Eindhoven. Our conversation sparked my curiosity about the real workings of venture capital, especially in the [DeepTech sector](https://rob-hoeijmakers.ghost.io/reflections-on-a-deep-tech-ai-event/). I invited him for a longer interview, aiming to learn more about how early-stage investing actually works. Patrick generously took the time to explain the world behind venture capital funds, hardware innovation, and the unique dynamics of startups that operate at the intersection of science and business. This post captures what I learned, with a personal lens: not a technical deep dive, but a practical journey into a system that shapes much of the technological future. ## The Start: A Personal Curiosity My background is rooted in digital strategy, AI, and content creation. I bridge business and technology every day, but venture capital always seemed a bit abstract to me. When I met Patrick, I realised this was an opportunity to make it concrete. I wanted to understand not only the big numbers and structures but also the human side: how do ideas become funded companies? What is the logic behind selecting just 2 or 3 investments out of 900 applications each year? [Reflections on a Deep Tech AI EventAttending the Deep Tech AI Event in Amsterdam was an inspiring experience. Here’s a reflection on the key takeaways and some of the exciting conversations that stood out during the event.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-95.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/https---cdn.evbuc.com-images-907573173-1755783377303-1-original.20241127-150707-1-2.jpeg)](https://rob-hoeijmakers.ghost.io/reflections-on-a-deep-tech-ai-event/) ## Key Insights from the Interview In our conversation, several themes stood out clearly. These insights reshaped the way I look at venture capital, innovation, and the broader startup ecosystem. ### 1\. DeepTech is Different Patrick explained that Cottonwood focuses specifically on "science-based hardware" ventures. Unlike pure software startups, hardware innovation demands bigger investments, longer development cycles, and carries much higher risks. But if successful, it often locks in markets through patents and infrastructure, making competition harder. ### 2\. Venture Capital is a Game of Focus and Patience Cottonwood typically comes in early and stays involved through several funding rounds, stepping back once the venture moves beyond their focus. They bring more than just capital—offering hands-on support with market access, strategy, and building the right network. While they don’t take on day-to-day roles, they back founders in building strong, self-sustaining teams through support with hiring, scaling, and organisational development. ### 3\. Europe versus the US: Different Capital Cultures A striking insight was the structural difference between Europe and the US. In America, venture capital is more mature, with generations of entrepreneurs reinvesting into new ventures. In Europe, the ecosystem still leans heavily on public funding and lacks the depth of private capital that fuels risk-taking in the US. ### 4\. Fragmentation is a European Weakness Patrick highlighted that [Europe](https://rob-hoeijmakers.ghost.io/tag/europe/)'s fragmented landscape often prevents startups from scaling. Regulations, smaller funds, and national borders make it harder for European companies to grow into global players. Coordination between funds and governments is improving, but there is a long way to go. ### 5\. Collaboration Over Competition Despite current trends towards protectionism and economic nationalism, Patrick strongly believes in cross-border collaboration. Cottonwood itself is a US-European fund, demonstrating that innovation thrives better when people work together rather than retreat into isolated silos. ## A Personal Reflection This conversation opened a new window for me. I now better understand why hardware startups are both thrilling and fragile, why venture capital is as much about partnerships as about money, and why [Europe's future](https://rob-hoeijmakers.ghost.io/tag/europe/) in technology depends heavily on how well we coordinate investments. I also appreciate more than ever the role individuals play—the founders, the funders, the mentors—each moving pieces of a larger, slower, and riskier puzzle than the fast-moving world of apps and SaaS startups. ## Final Note I am grateful to Patrick for sharing his time and expertise so openly. Our conversation was not an official interview for publication; it was a generous, personal exchange that offered valuable lessons for anyone curious about innovation, investment, and the often unseen engines of the future. I hope this post helps demystify a little bit of the world behind DeepTech investing and inspires others, like me, to keep asking questions and seeking to understand the forces shaping our digital world. [Deeptech VC: Cottonwood Technology Fund | Disruptive Capital for Disruptive Ideas![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon-32x32-1.png)Deeptech VC: Cottonwood Technology Fund![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/svg-3E)](https://www.cottonwood.vc) --- [An interview on the structural gaps holding Europe back in deeptechDeeptech isn’t SaaS. It’s often slow, risky, physical and still underfunded in Europe. What’s behind the gap with the US? A rare cross-Atlantic perspective.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-246.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/822E0B1B-0B3E-439B-8756-7321744F7843.jpeg)](https://rob-hoeijmakers.ghost.io/the-deeptech-dilemma-europe/) [Investing in Atoms: Alain le Loux on Building the Deeptech FutureDeeptech demands patience, guts, and hands-on conviction. This article profiles Alain le Loux and the rare venture model behind Europe’s hardware hopes.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-261.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/C8095225-ED9A-4896-9EC3-F900E3354459-1.jpeg)](https://rob-hoeijmakers.ghost.io/alain-le-loux-on-deeptech/) ### What We Mean When We Say "Flywheel" URL: https://rob-hoeijmakers.ghost.io/flywheel/ Last updated: 2026-05-10T08:54:01.000Z In the digital world — the AI space, the startup space, the tech-blog space — we throw around a lot of strange words. [*Benchmark*](https://rob-hoeijmakers.ghost.io/where-does-the-word-benchmark-come-from/)*.* [*Dashboard*](https://rob-hoeijmakers.ghost.io/why-dashboards/)*.* [*Blueprint*](https://rob-hoeijmakers.ghost.io/blueprints/). *Flywheel.* At some point, you just use them without thinking. Until you stop for a second and wonder: wait, what *is* a flywheel, actually? The flywheel is one of those metaphors that hides real, fascinating technology behind it. And once you get what a flywheel really does, the metaphor becomes even sharper — and a lot more useful. ## So, what *is* a flywheel? Literally, a flywheel is a heavy wheel that stores **kinetic energy** — the energy of motion. Once you get a flywheel spinning, it really, really wants to keep spinning. It resists changes. It carries momentum. Flywheels are everywhere once you start looking: - In old **steam engines** (to smooth out the force of the pistons) - In **pottery wheels** (so you can keep shaping clay without constantly kicking) - In **cassetterecorders** (to keep the tape running smoothly) - In modern **energy storage** systems (charging up a spinning mass instead of a battery) - And — dramatically — in **rollercoaster launches**, where a flywheel builds up massive speed and hurls the tiny cars onto the track in a split-second. The feeling of sudden acceleration when a rollercoaster launches? That’s flywheel energy, released all at once. It takes real effort to get a flywheel moving. But once it's spinning, it can do a lot with very little extra energy. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/04/IMG_2634.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/04/IMG_2635.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/04/IMG_2636.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/04/IMG_2637.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/04/storage-1.webp) Flywheels ## Why do we love the "flywheel" metaphor in business — especially startups? When people say "*we're building a flywheel*" or "*we're creating a flywheel effect*", they usually mean two things: ### 1\. Acceleration from stored energy Startups know this feeling intimately. You put in **endless work upfront**: building the product, finding early users, tuning the experience. It’s slow. It’s heavy. But if you survive that first painful stretch, the system starts to help you. **Early customers become advocates.** **New users drive more visibility.** **Energy starts feeding on itself.** Suddenly the "car" — your company — is speeding down the track. Not because you're pushing harder, but because the flywheel you built is releasing its power. ### 2\. Constant force and resilience A flywheel isn't just about speed. It's also about **stability**. It smooths out bumps. It keeps momentum going even when conditions aren't perfect. In a [startup](https://rob-hoeijmakers.ghost.io/reflections-on-a-deep-tech-ai-event/), that matters. A lot. You don’t want to be fighting friction at every turn. A working flywheel carries you over the slow patches without needing a crisis to re-motivate everything. In short: a flywheel turns fragile effort into lasting, building energy. ## The hidden beauty of the flywheel What I love about the flywheel — in engineering and in life — is how invisible the early stages are. At first, you push and push, and nothing seems to happen. You question whether you're even moving. But if you keep applying effort, at the right angles, in the right rhythms, **something shifts**. The wheel starts to help you. That’s not magic. That’s physics. But it feels like magic. And that's exactly why the flywheel metaphor feels so alive in startups, creative work, personal projects. The early grind matters more than you realise. Not because of what you get immediately, but because of **the momentum you store for later**. ## One more thing There’s a reason why real flywheels are heavy. If they were light, they couldn’t store any serious energy. The heaviness, the slow early grind, is part of the design. If your project, startup, or personal journey feels heavy at the start: Good. You’re building mass. You’re building momentum. Give it time. Keep pushing. And when the flywheel finally spins on its own, you’ll feel it. 0:00 /0:07 1× The humming and spinning --- ### Further reading - [Reflections on a Deep Tech AI Event](https://rob-hoeijmakers.ghost.io/reflections-on-a-deep-tech-ai-event/) - [Where does the word “benchmark” come from?](https://rob-hoeijmakers.ghost.io/where-does-the-word-benchmark-come-from/) - [Do dashboards protect against flying mud?](https://rob-hoeijmakers.ghost.io/why-dashboards/) - [From Blueprints to Bytes: How Original Designs Shaped Modern Technology](https://rob-hoeijmakers.ghost.io/blueprints/) ### Who Is the Author? Authenticity in the Age of AI URL: https://rob-hoeijmakers.ghost.io/who-is-the-author-authenticity-in-the-age-of-ai/ Last updated: 2026-05-10T08:54:01.000Z At a working session held at Google’s Amsterdam headquarters, organised by the Dutch Employers Association ([VNO-NCW West](https://www.vno-ncwwest.nl)) we, [Schmuki Digital & AI Agency](https://schmuki.nl), led a workshop on AI and ethics. The theme was broad: AI is now embedded in voice systems, search, chatbots, and decision-making processes. But within the group discussion, a more personal and immediate (ethical) issue surfaced: how to use AI, especially language models, without losing **authenticity**. Are you still saying what *you* mean? Is it in your tone, your words, your timing? Or has it quietly become part of a broader stream of AI content: **well-structured, grammatically sound, but ultimately indistinct?** Just another drop in what’s starting to look like a wave of synthetic, well-packaged, but empty communication. The word *authenticity* came up again and again. And it pointed to something important: when AI enters the creative process, we need to ask not only whether something works, but whether it’s still *ours*. ## Finding direction in the language When I start to loose direction I study the language, the etymology of the words. The word [*author*](https://www.etymonline.com/search?q=author) comes from the Latin *augere*, meaning “to grow” or “to originate.” An author isn’t just someone who writes something down, they’re the one who makes it possible, let it **grow into the world**. [*Authentic*](https://www.etymonline.com/search?q=authentic) is related. It comes from the Greek *authentes*: “one who does things with their own hands.” In other words, someone directly involved. Not copied or simulated, the real source. > Maybe the most human thing we can do, in the age of AI, is to take responsibility for what we say—and to mean it. That’s where trust begins. In a time when machines can now generate fluent text in seconds, this original meaning becomes more relevant, not less. The difference between content that’s *crafted* and content that’s *produced* is not just philosophical—it’s practical. It gives [*authority*](https://www.etymonline.com/word/authority). ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/04/IMG_2580.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/04/IMG_2583.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/04/IMG_2570.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/04/IMG_2564.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/04/Rob-Hoeijmakers-at-Google.jpeg) Some authentic impressions of the day. ## Use AI to execute and for vision The real shift isn’t just in how fast content is made. It’s in where the *vision* comes from. More and more, businesses let AI systems suggest not only the words, but the topics, the tone, and even the timing. That’s efficient—but it also moves the centre of gravity away from the human, the maker, the one who knows why something matters. AI can structure, rewrite, improve, or speed things up. It’s a tool. A sharp one. But it shouldn’t be asked to lead. The vision—the decision of what’s worth saying, and why—still needs to come from someone. Preferably someone who’s lived it. Seen the inside of a problem. Thought it through. That’s where value is created. Without that, we end up with content that (maybe) performs, but doesn’t connect. That follows trends, but doesn’t say anything new. That fills feeds, but leaves no mark. > As AI automates execution, it hands back the keys of the world to those that can ask the best questions - the creatives. [pic.twitter.com/c96AnETcmK](https://t.co/c96AnETcmK) > > — David Galbraith (@daveg) [April 24, 2025](https://twitter.com/daveg/status/1915439144500633860?ref%5Fsrc=twsrc%5Etfw) ## Embodied Knowledge is what we look for In a world of generated sameness, one thing stands out: lived-through knowledge. This doesn’t need to be dramatic or expert-level. It just needs to be *real*. [Embodied knowledge is slower](https://rob-hoeijmakers.ghost.io/not-doing-is-also-an-act/). It takes time. But it builds trust. It creates relationships. And increasingly, it’s what people are searching for, because it’s one of the few signals left that something is actually worth paying attention to. This isn’t a call to return to hand-coded websites or handwritten letters. It’s a practical stance: choose to be the author. Use AI to support your voice, not replace it. Shape your message before you let the system polish it. It’s not about being perfect. It’s about being responsible for what you create. Owning the idea. Showing experience. Committing to it. And that applies whether you’re a solo creator, a growing business, or a team working inside a large organisation. The tools have changed. But the core remains: who is the source? ### Further reading - [AI Signals The Death Of The Author | NOEMA](https://www.noemamag.com/ai-signals-the-death-of-the-author) - [Not Doing Is Also an Act](https://rob-hoeijmakers.ghost.io/not-doing-is-also-an-act/) ### Not Doing Is Also an Act URL: https://rob-hoeijmakers.ghost.io/not-doing-is-also-an-act/ Last updated: 2026-05-10T08:54:02.000Z **AI is fast. But intentional growth still takes time.** ## A Fundamental Shift AI has changed the way I work—drastically and fundamentally. As a knowledge worker, I’ve always moved between learning, structuring, and creating. But lately, the ground beneath that process has shifted. Tools don’t just assist—they generate, suggest, complete. You’re no longer chasing insight. Insight is chasing you. And not just in what you read or write, but in how you speak. Dictation, transcription, voice-to-note—these new frictionless input modes have *rewired my rituals*. Ideas now enter faster than ever. I talk to capture thoughts. I draft while walking. I work without sitting down. ## From Scarcity to Abundance One of the biggest shifts wasn’t technical. It was psychological. I had to move from a *psychology of scarcity*—where ideas were rare and needed to be hunted—to a *psychology of abundance*. Now the problem isn’t how to get input, it’s how to handle the overflow. AI tools give you everything: outlines, metaphors, copy, code. And with frictionless input, you can produce at the speed of thought. That’s powerful—but also disorienting. Because in this new reality, **just because something can be done doesn’t mean it should be**. ## Pacing, Not Chasing The temptation is to produce more, learn more, publish more. But if you keep feeding the machine without feeding yourself, you’ll burn out on your own, and borrowed, brilliance. You’ll lose the time to reflect, absorb, and actually become the person behind your output. That’s when I realised: *not doing is also an act*. It’s not hesitation. It’s authorship. It’s not avoidance. It’s alignment. In Taoist thought, this idea is known as *wu wei*: action through inaction. It’s not laziness. It’s precision. You’re not reacting to every signal—you’re choosing your own. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/05/Gpz_AdIWAAAIh-8.jpeg) Probably the same can be said for conversations with ChatGPT. ## Integration Over Acceleration AI invites you to speed up. Frictionless input makes everything feel urgent. But growth still requires digestion. Learning still takes pacing. Reflection still matters. **And so does memory**. Without *declarative memory*—the ability to recall, relate, and re-apply what you've learned—reflection becomes passive. What I’ve found is that intentional recall (like through spaced repetition systems such as [Anki](https://rob-hoeijmakers.ghost.io/human-memory-in-a-digital-world/)) helps me *embody* knowledge. It’s how I test whether something has really stuck, or whether I just passed through it too fast to care. In a world of instant knowledge, **real understanding still takes time**. Slowing down doesn’t mean you’re falling behind. It means you’re actually here. --- ### Further reading - [Anki, Human Memory in a Digital World](https://rob-hoeijmakers.ghost.io/human-memory-in-a-digital-world/) ### Search in the Age of AI URL: https://rob-hoeijmakers.ghost.io/search-in-the-age-of-ai/ Last updated: 2026-05-10T08:54:02.000Z *From puzzles to conversations*. *How search became fluent*. Search used to feel like solving a puzzle. You had to guess the right keywords, strip away the natural rhythm of your question, and hope the system knew what you meant. Most of the time, it didn’t. Today, something different is happening. Search is becoming fluid. Conversational. Even responsive. The keyword box is giving way to an interface that listens, responds, remembers. What once felt like a static lookup is now turning into a dialogue. This change is more than a surface-level improvement. It’s a shift in how we think, how we find, and how we connect ideas. Behind the scenes, the architecture has evolved—from keyword indexes to semantic embeddings, from matching results to generating answers. But on the surface, it’s changing how we use search to orient ourselves in the world. In this article, I’ll walk through how search has evolved. Not just technically, but experientially. I’ll share what this shift means for how we ask, how we retrieve, and how we make sense of things. ## 1\. The Keyword Era: When Search Was a Puzzle The early web taught us to speak in code. Instead of asking *“What are the risks of RSV for newborns?”* you learned to type *“RSV newborn risk factors”. A*nd even then, it helped if you knew how Google thought. This style of search was efficient, but narrow. It required precision. The interface didn’t support uncertainty, follow-up questions, or clarification. Search was a one-shot action. You either hit the target, or you rewrote the query. And behind the curtain? It was a beautiful but limited machine. Tokenisation, inverted indexes, relevance scores, link-based ranking. All built for speed, not for nuance. For years, that’s what search meant: find the right words, match them well, and rank the results. But human language isn’t made of keywords. It’s made of questions, intentions, detours, and follow-up thoughts. ## 2\. Enter AI: Search Becomes a Conversation The arrival of large language models didn’t just change how we *interact* with information. It changed how systems *retrieve* it. With AI, you can ask a half-formed question. You can clarify it later. You can stay in the flow. The model doesn’t just return links, it interprets what you mean, fetches relevant content, and composes an answer in real time. This shift is profound. Under the hood, a process called [retrieval-augmented generation](https://sebastianraschka.com/blog/2023/rag.html) (RAG) quietly rewires everything: - The model retrieves chunks of relevant information from a database - It blends them with your query - It generates a contextual response This doesn’t replace search, it *wraps around it*. Suddenly, the search engine isn’t just a finder. It’s a participant. Instead of clicking through results, you’re building understanding together. [Project: AI Chatbot for RSV Awareness through Web and WhatsAppDiscover how I partnered with ReSViNET to build an AI-powered chatbot that provides global RSV education, available via web and WhatsApp.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-76.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_8521-1-1-1.jpeg)](https://rob-hoeijmakers.ghost.io/awareness-via-chatbot-collaboration-with-resvinet/) ## 3\. Search Expands: From Text to Everything The old model of search worked best in one mode: text in, text out. But human understanding isn’t limited to words. We process images, sound, video, structure. And now, search is catching up. AI has cracked open the door to **multimodal search**, the ability to look *inside* other forms of content. Not just text, but timelines. Not just titles, but tone. - Tools like [Whisper](https://rob-hoeijmakers.ghost.io/openais-whisper-speech-to-text-technology/) turn speech into searchable transcripts. - Apps like [Descript](https://rob-hoeijmakers.ghost.io/descript/) let you scrub through video as if it were text. - [LLMs](https://rob-hoeijmakers.ghost.io/a-comparative-analysis-of-top-language-models/) like GPT-4 can “watch” images, interpret screenshots, even extract meaning from diagrams or interfaces. - [Zeta Alpha](https://rob-hoeijmakers.ghost.io/zeta-alpha/) brings deep, semantic search to internal research data—designed for teams working with complex knowledge in academic or scientific contexts. - Glean focuses on intelligent enterprise search, helping employees find the right document, person, or answer across internal company systems. - [Gemini](https://rob-hoeijmakers.ghost.io/google-gemini-video-ai/) adds powerful multimodal capabilities, making it possible to search within images, diagrams, and even across video content. [Speech recognition, Speech-to-Text, Dictation, and Transcription: What’s the Difference?Just like picking the right writing tool, choosing the right speech technology makes all the difference—here’s how to decide.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-77.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_4550-3.jpeg)](https://rob-hoeijmakers.ghost.io/asr-speech-to-text-dictation-and-transcription-whats-the-difference/) Underneath this magic is a technical breakthrough: **embeddings**. These are mathematical representations of meaning. Instead of indexing words, systems now index *relationships*, between concepts, sentences, sounds, even moods. This means search doesn’t just look for matches. It looks for *meaningful proximity*. What *feels* related, what *sounds* aligned, what *rhymes* with your intention, even if it doesn’t use the same words. This is why AI-powered search feels intuitive. It moves like thought. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/04/Easter-Eggs.jpg) Visual search for Easter Eggs. ## 4\. Infrastructure Shift: From Indexing to Understanding Behind the scenes, the entire infrastructure of search has changed. The old stack, tokens, inverted indexes, PageRank, still exists. But around it, new layers have formed: - [**Vector databases**](https://rob-hoeijmakers.ghost.io/scaling-rag-applications-with-weaviate/) store semantic embeddings instead of just text - [**RAG pipelines**](https://rob-hoeijmakers.ghost.io/exploring-retrieval-augmented-generation-rag/) orchestrate retrieval, re-ranking, and response generation - [**Agent frameworks**](https://rob-hoeijmakers.ghost.io/agents-and-agency/) allow systems to run multiple queries in parallel, adapt them, test answers, and respond in natural language These layers are still under rapid construction, but they point to something clear: we’re not just building better finders. We’re building *interpreters*. Systems that can reason across time, across modalities, across contexts. This makes search an active process, not just a lookup, but a loop. Each answer refines the next question. If that sounds familiar, it’s because it’s how humans learn. [The End of Google Search (as we know it)Google didn’t warn me. It just erased my blog. What looked like a bug turned out to be a glimpse into the future of search—and it’s not built for us anymore.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-81.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_2069-2.jpeg)](https://rob-hoeijmakers.ghost.io/the-end-of-google-search-as-we-know-it/) ### 4.5 Adding Structure: The Role of the Knowledge Graph Probabilistic models like GPT are powerful, but they don't “know” things in the traditional sense. They work by recognising patterns and predicting likely continuations, not by retrieving structured facts. That’s where [**knowledge graphs**](https://rob-hoeijmakers.ghost.io/semantic-web/) come in. A knowledge graph represents entities, like people, places, concepts, and their relationships in a structured, deterministic way. It’s a way of encoding meaning with precision. You could think of it as a map of interlinked truths. When combined with language models, knowledge graphs do two important things: - **Grounding**: They reduce hallucination by letting the system check facts and draw from verified sources. - **Enrichment**: They provide explicit relationships that help the AI understand context and nuance it might otherwise miss. This hybrid approach, combining probabilistic reasoning with structured logic, is becoming the new norm. It lets AI systems not only sound fluent, but also stay anchored in facts. And it brings us one step closer to search systems that don’t just guess well, but also *understand* what they're working with. [Semantic web and the knowledge graphThe semantic web is about describing the web of connections between ideas and objects.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-78.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_9740.jpeg)](https://rob-hoeijmakers.ghost.io/semantic-web/) ## 5\. From Results to Real Understanding Search used to be about *answers*. But increasingly, it's about *orientation*. What’s going on here? What’s relevant? What matters right now? AI systems are getting better not at *knowing everything*, but at *knowing what to show you next*. They retrieve what you didn’t know to ask. They connect pieces. They compress the irrelevant and surface what’s useful. They don’t replace your thinking, they scaffold it. This changes how we work, learn, research, explore. It even changes how we *write*. This article, for example, has been shaped by conversations I’ve had with AI, not just to find information, but to build understanding. [Blog writing workflowA easy to use workflow for writing, translating and publishing blog articles. On Mac and iPhone. I will use paper, Ulysses, LanguageTool, Deepl and Ghost.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-79.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/Image-08-03-2022--19-43-1-1.jpeg)](https://rob-hoeijmakers.ghost.io/blog-writing-workflow/) ## 6\. Context Is the New Query One of the biggest shifts in search, often overlooked, is the rise of **contextuality**. We no longer just ask isolated questions. We build conversations, trajectories, threads of thought. In traditional search, every query was a reset. The system had no idea what you just asked. But with AI, memory matters. Each prompt can build on the last. Each session becomes a layer of context. Some systems, like [ChatGPT with memory](https://rob-hoeijmakers.ghost.io/chatgpts-memory/) turned on, even remember your past sessions, your tone, your preferences. This changes everything. It means we can think more fluidly, ask in fragments, revise our line of inquiry without starting over. The system becomes less of a tool, more of a thought partner. And technically, it’s made possible by embedding layers, session-aware architectures (think identity), and increasingly sophisticated memory models. In practical terms, it’s the difference between searching for a fact, and *working something out* over time. As users, we’re beginning to expect this. We want systems that remember what we’ve said, how we said it, and what we meant. Not because it’s convenient, but because it reflects how thinking really works: iteratively, relationally, in motion. In that sense, **context has become the query**. [Identity unmaskedExplore ‘identity’ in the digital era, its dual role in admin and online worlds, and the convergence enabled by facial recognition tech.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-82.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_5603-1.jpeg)](https://rob-hoeijmakers.ghost.io/identity-unmasked/) ## From Search to Sensemaking What we’re witnessing isn’t the end of search. It’s its expansion. Search is no longer just a way to find things. It’s how we think through them. How we enter a topic, stay with it, revise our questions, and follow the thread. From puzzles to conversations. From keywords to context. From input/output to flow. The tools are evolving quickly. The infrastructure is still taking shape. But the experience is already here: **search that listens, that learns, that adapts**. And it’s just the beginning. Happy Easter, keep on finding answers! --- **Related posts**: --- ### Further reading - [Site Search explained](https://rob-hoeijmakers.ghost.io/website-search/) - [ChatGPT and the art of asking questions](https://rob-hoeijmakers.ghost.io/the-art-of-asking-questions/) ### Locked Out by X URL: https://rob-hoeijmakers.ghost.io/locked-out-by-x/ Last updated: 2026-05-26T19:30:24.000Z On a regular Monday morning, just after posting a fairly innocent message on X, I was suddenly locked out of my account. No warning, no flagged content—just a message saying “We found some unusual activity on your account.” I was asked to prove I was human before I could continue. This is what happened next. ## A Human Verification Challenge That Almost Got Me The system prompted me to “Pass a challenge.” That was all the instruction I got. No indication of how long it would take or what kind of challenge it would be. The challenge itself was strange. It used a rotating orbital interface and asked me to tap on matching icons. The icons weren’t just abstract—they were difficult to identify, constantly moving, and visually inconsistent. It was unclear when the test started or ended, and I failed the first time. 0:00 /0:11 1× The X challenge to proof I am not a bot. I had to do five of these. Moving into the indicated orbits. When I finally passed it on the second attempt, I expected to be let back in. Instead, I landed on a white screen with no way forward. No confirmation, no dashboard—just two links: *Help* (which sent me to a generic support page), and *Log out*. At that point, I genuinely thought I might be permanently locked out. Who do you call? There’s no support desk. No human fallback. You're on your own, staring at a dead end. So I logged out. Then I had to log back in—and since I use two-factor authentication, that added yet another layer to an already fragile moment. Only after all that was I back in. But it didn’t end there. ## Then Came the Label After proving I was human, I regained access. But not full trust. I received another message saying my account had been “labelled” for possible platform manipulation—terms like spam, artificial amplification, and disruptive behaviour were mentioned. None of which applied. It was vague and unsettling. I was informed that my reach might be limited, my posts excluded from replies, trends, and recommendations. I hit the “Request review” button, half-expecting it to disappear into a void. Surprisingly, a few minutes later, I got a message: the label had been removed. No further explanation. No clarification on what triggered the label in the first place. But at least the account was back to normal. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/04/X-Locked-Out.jpg) The locked out screen in X and the labeled account. ## What Does This Say About Platform Identity? I write a lot about identity and social platforms, so this hit a nerve. These platforms are increasingly shaping how we show up in public space—and also how we get excluded from it. When your account is locked or labelled, you’re instantly treated as suspicious. The burden of proof shifts to you, and the process for restoring trust is unclear at best. What if I hadn’t passed the challenge? What if I was visually impaired, using assistive tech, or didn’t speak English fluently? The test was abstract, awkward, and certainly not inclusive. This is more than a UX issue. It’s about power and access in digital environments. [Locked Out of Instagram or Threads? Avoid and appeal a ban.Banned from Instagram or Threads? Here’s how to avoid getting locked out and the steps to take for a swift and successful account recovery.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-67.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_4133.jpeg)](https://rob-hoeijmakers.ghost.io/locked-out-banned-instagram-threads/) ## Why I’m Sharing This I got back in. Many won’t. And I think it’s important to talk about that. If you're building a professional identity, a public presence, or even just trying to participate, these kinds of opaque interventions can feel destabilising. Worse, they can push people out of the conversation entirely—without explanation, recourse, or fairness. Social media platforms like X need to do better at explaining their enforcement actions. Transparency isn’t optional when your platform is where people work, speak, organise, and connect. I’m sharing this so others can see how this works under the hood—and maybe be a little more prepared than I was. --- ### Further reading - [Paid subscription on X](https://rob-hoeijmakers.ghost.io/paid-subscription-on-x/) ### What Zeta Alpha Gets Right About AI (That Others Don’t) URL: https://rob-hoeijmakers.ghost.io/zeta-alpha/ Last updated: 2026-05-10T08:54:03.000Z ## It started at the Amsterdam AI Salon That’s where I met Arjen de Hoop. We had [a great initial exchange](https://rob-hoeijmakers.ghost.io/the-end-of-google-search-as-we-know-it/)—one of those rare conversations where someone doesn’t pitch the usual AI buzzwords, but actually listens, and talks back with clarity. We agreed to follow up. A week later, we met again in a quiet Google Meet session. Arjen had just returned from the Hannover Messe, where [Zeta Alpha](https://www.zeta-alpha.com) had drawn attention—not just on the floor, but in [*Frankfurter Allgemeine Zeitung*](https://www.faz.net/pro/digitalwirtschaft/transformation/die-fehlende-ki-strategie-der-industrie-wird-zum-problem-110390344.html). The article described how Germany’s manufacturing industry is struggling with scattered knowledge, siloed systems, and a growing need for focused, secure, and truly helpful AI tools. Tools not built for clicks or scale, but for engineers, researchers, and analysts. People who need results they can trust. [Die fehlende KI-Strategie der Industrie wird zum ProblemDas Potential der generativen KI ist auch in deutschen Industrieunternehmen angekommen. Doch bei der Umsetzung hapert es, zeigt eine Studie des VDMA. Nur 7 Prozent haben bisher eine KI-Strategie.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon-1.svg)Frankfurter Allgemeine ZeitungNina Müller![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/auf-der-hannover-messe-2025.jpg)](https://www.faz.net/pro/digitalwirtschaft/transformation/die-fehlende-ki-strategie-der-industrie-wird-zum-problem-110390344.html) ## What I saw in the demo Zeta Alpha isn’t building just another AI chatbot or dashboard. It’s building something more foundational: a system that helps companies structure, surface, and actually use the knowledge they already have—but can’t find. Through neural search and multimodal AI, Zeta Alpha lets people ask complex questions in natural language and get reliable, source-backed answers—whether the data lives in PDFs, PowerPoint slides, Excel sheets, or diagrams. What stood out to me was the *precision*. It doesn’t throw in fuzzy language or guesses. It retrieves what matters. And because the models are trained with internal data—and often fine-tuned with synthetic examples and a human-in-the-loop process—it gets smarter over time without compromising security. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/04/zeta-alpha-presentation.jpeg) Conversation with Arjen de Hoop from Zeta-Alpha. ## Why this matters now European companies are waking up to the limits of generic, cloud-based AI services. Tools like ChatGPT or Bard are brilliant for brainstorming, but not for navigating sensitive or specialised internal knowledge. That’s where [Zeta Alpha ](https://www.youtube.com/watch?v=FxFYxNmV834)steps in. It’s built with an Enterprise mindset: private, auditable, deployable on-prem or in the private cloud. Your data stays yours. Your intellectual property doesn’t get absorbed into a general model. And your teams can build real capability, not just automation. [The dangerous myth: building RAG yourselfArtificial intelligence is revolutionising the business world. Retrieval Augmented Generation (RAG) is at the centre of this development…![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/10fd5c419ac61637245384e7099e131627900034828f4f386bdaa47a74eae156-1)MediumGerhard Kürner![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/1-i-jmErXn-K-VhS2RaMdi6g.jpeg)](https://gerhardkuerner.medium.com/the-dangerous-myth-building-rag-yourself-16011e3e8fbc) RAG, buy or build? ## Agents that act, not just chat Once the search layer is strong, Zeta Alpha introduces AI agents. These agents don’t pretend to be all-knowing—they operate within your company’s context, on your data, with your rules. Think: summarising customer feedback across regions, generating regulatory reports, extracting trends from lab results. Quietly powerful, and grounded in real needs. 💡 ****AI agents** offer a range of practical applications, from virtual assistants for product documentation and automated maintenance schedule extraction to onboarding support and legal drafting based on internal data. A particularly effective implementation is the **Deep Research Agent*: a multi-stage research framework powered by AI agents that can analyse complex datasets, synthesise information from multiple sources, and generate structured reports of approximately four pages. ****Typical use cases** include trend mapping, competitor analysis, policy evaluation, and internal knowledge consolidation—enabling faster, more scalable decision-making across teams. ## What stayed with me Zeta Alpha isn’t trying to be everything to everyone. It’s built for organisations that operate in complexity—industries like manufacturing, pharma, chemistry, consulting, and R&D. These companies don’t need hype. They need precision, context, and trust. That’s exactly what Zeta Alpha offers. In a time when digital independence and strategic clarity are more critical than ever, that kind of focus isn’t just useful. It’s essential. --- ### Further reading - [Platform | Zeta Alpha](https://www.zeta-alpha.com) ### Scaling RAG Applications with Weaviate. URL: https://rob-hoeijmakers.ghost.io/scaling-rag-applications-with-weaviate/ Last updated: 2026-05-10T08:54:03.000Z In my ongoing exploration of [Retrieval-Augmented Generation (RAG)](https://rob-hoeijmakers.ghost.io/exploring-retrieval-augmented-generation-rag/), I recently took a deep dive into Weaviate — a Dutch company making waves with its open-source vector database. If you're working on AI agents, knowledge bots, or anything that relies on semantically rich information retrieval, understanding how a vector database like Weaviate fits into the architecture is not optional. It's essential. ## Why Weaviate caught my attention Several of my recent projects, [especially in healthcare and multilingual knowledge sharing](https://rob-hoeijmakers.ghost.io/awareness-via-chatbot-collaboration-with-resvinet/), have reached the limits of what can be done with off-the-shelf tools. I needed something that could scale. A friend with a deep tech background pointed me to [Weaviate](https://weaviate.io)—a vector-first database designed for AI-native applications. After attending two demos and speaking directly with one of their experts, it started to make sense. ## What makes a vector database different? Traditional databases store and retrieve exact matches: strings, numbers, Booleans. Vector databases operate in an entirely different paradigm. They store the "meaning" of content by translating text, images, and other data into vectors — long lists of numbers that live in a multi-dimensional space. When a user sends a query, it too gets turned into a vector. The database then finds the closest match *by meaning*, not by keywords. This approach powers fast, flexible semantic search and is ideal for RAG setups, where you combine a vector search with a large language model to generate rich, context-aware answers. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/04/home-4.jpg) ## Live demos that brought it to life The demo I attended was hosted by [JP Hwang](https://www.linkedin.com/in/jphwang/), and it stood out immediately. He managed to make the subject matter both light and insightful — no small feat when dealing with something as abstract as vector embeddings and semantic search. His hands-on approach made the technology tangible in minutes. One of the demos used a dataset of 1,000 movies, showcasing how Weaviate ingested title, overview, year, and popularity fields, vectorised them using models like Cohere or OpenAI, and enabled search queries like: - "Movies about history" - "Green action hero" - or even questions in Dutch or Korean, against an English dataset What impressed me most was how fast and flexible it was. You can: - filter like in SQL (e.g. release year between 2000 and 2010) - perform fuzzy or semantic searches ("history" finds Back to the Future) - rank results based on distance in vector space - combine this with generative AI to summarise or extract meaning ## Scaling up: the real differentiator Here's where vector databases shine. Traditional brute-force search across millions of objects breaks down fast. Vector databases like Weaviate use approximate nearest neighbour (ANN) search, which keeps retrieval fast even as your dataset grows exponentially. This makes it feasible to build enterprise-grade search and assistant experiences across billions of records. For example, if you're building a chatbot that answers questions from PDF manuals, medical journals, or user feedback, you can: - convert your content into chunks - embed them into vector space - store them in Weaviate - query them live via API or LLM All without retraining your model every time you add new content. ## Open-source roots, enterprise readiness Weaviate is **open source**, so you can run it locally or on your own infrastructure. For hosted services, they offer a usage-based cloud model starting at around $25/month, with the option to scale to enterprise support. [Vector Database Pricing | WeaviateCompare pricing options for our different levels of vector database services and solutions.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon-19.ico)Weaviate![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/default-2.jpg)](https://weaviate.io/pricing) ## Final thoughts If you're serious about building AI systems that need high-quality, flexible information retrieval at scale, vector databases aren't just nice-to-have. They're foundational. Weaviate, with its thoughtful integrations, support for multiple modalities and languages, and strong developer experience, is one of the most compelling options I've seen. I'll be keeping an eye on their roadmap, and possibly doing a follow-up on how it performs in production settings. \--- Interested in this space? [Let me know](https://rob-hoeijmakers.ghost.io/about/) if you want to swap notes or brainstorm use cases. --- ### Further reading - [The AI-native database developers love | Weaviate](https://weaviate.io) ### Signals, Systems, and Speech URL: https://rob-hoeijmakers.ghost.io/signals-systems-and-speech-rethinking-meaning-in-the-age-of-ai/ Last updated: 2026-05-26T19:35:05.000Z ***Rethinking Meaning in the Age of AI*** ### Part 2 of a conversation with Larry Swanson In part one, we explored [the world of podcasting](https://rob-hoeijmakers.ghost.io/unpacking-the-world-of-podcasting/). In this second conversation, the format opens up: we dive into knowledge infrastructure, AI's impact on identity and work, and the emerging architectures of meaning in a digitised world. Larry had just landed a new role as Community Growth Manager at a knowledge graph and AI company. The way he got there says a lot about how modern careers are evolving. ## Listening as a Career Move Larry joined the Never Search Alone programme, a methodical approach to career reflection and outreach. He talked to thirty people. Not recruiters, but peers. The goal: **identify resonance**. What energises you? What patterns emerge? Through these conversations, Larry uncovered a clear fit: he lights up when he talks about common infrastructure, semantics, community. Within weeks, he had a job offer. His new role blends community building, AI evangelism, and developer outreach. He describes it as a cross between developer relations, content strategy, and live representation. As he put it, he's giving a human face to a deeply technical field. And it suits him. [Never Search AloneTake charge of your career and find a job you love. Join a free Job Search Council![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/63bc36985b4b46aea0b3266c_favicon.png)Never Search Alone logo![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/63cb0d958353842b3f4bef9c_social-share.jpg)](https://www.phyl.org) ## Probabilistic Meets Deterministic Our conversation moved quickly from job titles to the deeper shifts behind them. Larry works at the intersection of [**knowledge graphs**](https://rob-hoeijmakers.ghost.io/semantic-web/) and **LLMs**. He sees how neuro-symbolic loops combine structured, ontology-driven data with pattern-seeking models. These hybrid systems are already powering enterprise-level search and reasoning. We explored how deterministic knowledge systems (like Prolog, OWL, or RDF) give shape to the otherwise fluid outputs of LLMs. I shared my experience with companies building [RAG](https://rob-hoeijmakers.ghost.io/exploring-retrieval-augmented-generation-rag/) pipelines ([Weaviate](https://rob-hoeijmakers.ghost.io/scaling-rag-applications-with-weaviate/), [Zeta Alpha](https://rob-hoeijmakers.ghost.io/zeta-alpha/)) that layer domain-specific knowledge over general-purpose AI. Without this structure, large language models hallucinate. With it, they inform. 💡 What are ****neurosymbolic loops**? Neural systems like [LLMs](https://rob-hoeijmakers.ghost.io/web-traffic-and-the-rise-of-llms/) are flexible and generative, but prone to hallucination. Symbolic systems like knowledge graphs are structured and logical, but rigid. A neurosymbolic loop combines both: LLMs generate or interpret content, symbolic systems validate and refine it, and the cycle continues. This interplay grounds creativity in fact and brings structure to fluidity. [Rules and Guesses: The Twin Engines Behind Smarter AIAI isn’t just smart, it’s a duet. One part rigid and logical, the other fluid and generative. Together, they form the neuro-symbolic systems behind real insight.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-153.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_4826.jpeg)](https://rob-hoeijmakers.ghost.io/rules-and-guesses/) ## Speaking is the New Writing One theme that surfaced was the return to orality. I’ve dictated thousands of prompts, and I see voice as not just an interface, but a form of data generation. [Whisper-level](https://rob-hoeijmakers.ghost.io/openais-whisper-speech-to-text-technology/) ASR combined with LLMs is not just [transcription](https://rob-hoeijmakers.ghost.io/transcribe-audio-on-macos-and-ios/)—it's transformation. That shift isn’t just technical. It’s cognitive. Spoken prompts invite a different kind of thinking. There's improvisation, rhythm, a tempo of thought. And that, in turn, changes the kind of data we generate. It changes how AI listens. [Speech recognition, Speech-to-Text, Dictation, and Transcription: What’s the Difference?Just like picking the right writing tool, choosing the right speech technology makes all the difference—here’s how to decide.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-56.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_4550-2.jpeg)](https://rob-hoeijmakers.ghost.io/asr-speech-to-text-dictation-and-transcription-whats-the-difference/) 💡 Transformer-based ASR models like Whisper process audio in parallel, allowing for accurate live captioning and translation. More background in [The Silent Revolution: How AI is Transforming Speech Recognition](https://rob-hoeijmakers.ghost.io/speech-recognition/). ## Websites for Machines We also talked about [SEO](https://rob-hoeijmakers.ghost.io/my-basis-for-seo/). I’ve worked in SEO for over a decade, but now I see its endgame. Optimising for Google is no longer about humans reading pages. It's about machines parsing meaning. Structured data, schema.org, entity graphs — these are the new foundations. Human visitors are rare. The web, increasingly, is for robots. I recently came across a sharp distinction: **Google is optimised for *keyword intent*; LLMs for *conversation intent*.** That nails the transition. Traditional SEO was about matching search terms. Today's systems respond to how people think aloud. They interpret narrative intent, not just lexical signals. That doesn’t mean content is obsolete. It means we publish differently. We publish as signals, as identities, as connected nodes. And ironically, the human part becomes more important: voice, conviction, presence. [The End of Google Search (as we know it)Google didn’t warn me. It just erased my blog. What looked like a bug turned out to be a glimpse into the future of search—and it’s not built for us anymore.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-57.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_2069.jpeg)](https://rob-hoeijmakers.ghost.io/the-end-of-google-search-as-we-know-it/) ## Where Are We Going? Larry and I share concern about the social impacts of these shifts. LLMs are not neutral tools. Their integration into education, healthcare, and identity systems raises deep questions. We agreed: it’s not the tech that changes society. It’s the way we relate to it. For Larry, that means bringing clarity to complex systems. For me, it means articulating purpose in a time of overload. We both believe in infrastructure with meaning. Larry has also been exploring these questions more deeply in the *TruthCollapse* series with Noz Urbina, where they discuss the blurred lines between fact, fiction, and generated content. It’s a fitting name: the collapse isn’t just technical, but epistemological. What we believe—and why—is being reshaped by systems we barely understand. Their conversations add another layer to this dialogue: one of responsibility, narrative, and trust. [Truth Collapse by Noz UrbinaInvestigating truth collapse - the societal, psychological, cultural, and systemic metacrisis of algorithmic life - and how to make it better.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/truthcollapse-_icon-2.png)Truth CollapseNoz Urbina![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/WHITE.png)](https://truthcollapse.com) *This was part two of our ongoing dialogue. More will follow once Larry is settled in to his new job (and life).* --- ### Further reading - [Unpacking the World of Podcasting: An Interview with Larry Swanson (Part 1)](https://rob-hoeijmakers.ghost.io/unpacking-the-world-of-podcasting/) ### LinkedIn Verification: When Identity Becomes a Profile URL: https://rob-hoeijmakers.ghost.io/linkedin-verification-when-identity-becomes-a-profile/ Last updated: 2026-05-13T15:17:08.000Z Until recently, I didn’t think much about LinkedIn verification. You show you are who you say you are — simple enough. Maybe a handy badge, a bit of extra trust. But lately, a few things I read and experienced made me look at it differently. Here’s what struck me: verification on LinkedIn isn’t just a check. It becomes your profile. And that’s something else entirely. Maybe a handy badge, a bit of extra trust — [the process itself is straightforward](https://rob-hoeijmakers.ghost.io/how-to-get-verified-on-linkedin/). But lately, a few things I read and experienced made me look at it differently. ## What does verification actually mean? In the classic sense, verification is temporary: a quick check to confirm your identity. You show your passport, enter a code, and move on. It’s a moment — not a statement. On LinkedIn, though, verification is baked into your public identity. Your name, as it appears on your ID, becomes the name on your profile. No room for additions. No brand names. No nuance. What starts as a confirmation becomes a fixed label — your digital business card. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/04/Rob-Hoeijmakers-Schmuki-Linkedin.jpeg) I verified [myself](https://www.linkedin.com/in/rob-hoeijmakers/) and I did not see a difference, maybe the badge gives a bit of authenticity? ## It doesn’t work equally for everyone I only began to notice this when I started paying attention to how people present themselves. Women with hyphenated or combined last names. Freelancers with a trade name or brand. Creatives who use a pseudonym or add a descriptor to their name. People with cultural naming structures that don’t match Western conventions. That’s where friction starts. LinkedIn doesn’t allow company names in your surname field. Or personal additions like “| AI Consultant” next to your name. At the same time, the official rules *do* allow for shortened first names or nicknames. So you can be “Rob” even if your ID says “Robertus Johannes”. It’s not that strict — but the system still *suggests* that your real name should match your official name, and that suggestion tends to steer people toward conformity. The space for interpretation exists — but it’s not openly encouraged. And that can make it harder to claim a professional identity that doesn’t fit neatly into standard forms. ## So why verify in the first place? LinkedIn suggests verification is for your benefit: more trust, more visibility, fewer bots. But in practice, the clearest effect is this: **your profile becomes locked**. And if your identity is more layered, dynamic, or context-specific, that lock may not serve you. For Microsoft, the value is clear — more verified profiles mean better data, better targeting, and more confidence for recruiters and advertisers. For users, it’s worth asking: what are you giving up in exchange for that badge? ## A gentle note I’m not making a grand statement here. Just a side note. A shift in perspective. I’ve always used my legal name online. But I’m starting to see that there’s often space between who you officially are and how you want to be seen. And allowing that space — even just a little — makes room for a more practical, more recognisable, more human way of being present online. ### Further reading - [How to get verified on LinkedIn](https://rob-hoeijmakers.ghost.io/how-to-get-verified-on-linkedin/) - [Account, profile and identity](https://rob-hoeijmakers.ghost.io/accounts-profiles-identity/) ### The End of Google Search (as we know it) URL: https://rob-hoeijmakers.ghost.io/the-end-of-google-search-as-we-know-it/ Last updated: 2026-05-10T08:54:04.000Z ## The Nightmare It felt like the biggest nightmare for someone who considers Google the main gateway to his audience: my blog, where I write extensively about digital life and the professional world, all but disappeared from the index. Hundreds of pages, suddenly gone. No warnings, no manual actions—just silence. At first I thought it was a technical error, maybe something related to the [domain move](https://rob-hoeijmakers.ghost.io/rethinking-social-media-for-2025/) I did in December. But the deeper I looked, the clearer the picture became. It wasn't personal. It was systemic. ## The Cul-De-Sac of SEO Blogging The kind of blogging I’ve done for years—SEO-driven, informational, high-frequency—has quietly reached a dead end. A cul-de-sac. And while it’s still painful to say, I now understand why. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/04/Google-Search-.jpg) Around early February my blog got axed. I shifted my attention to other sources.. ## The Moment of Realisation What really helped me see it clearly was a conversation I had with [Arjen de Hoop](https://www.linkedin.com/in/arjen-de-hoop-076029163/) ([Zeta-Alpha](https://rob-hoeijmakers.ghost.io/zeta-alpha/)) working on [Retrieval-Augmented Generation (RAG)](https://rob-hoeijmakers.ghost.io/exploring-retrieval-augmented-generation-rag/) and agent-based browsing systems. He explained how [large language models (LLMs)](https://rob-hoeijmakers.ghost.io/a-comparative-analysis-of-top-language-models/) like ChatGPT or Gemini don’t just retrieve documents—they use search engines to **navigate the web like an agent**. They’re not reading blogs to be inspired or to find reflections. They’re searching for **factual, structured, actionable information**—what marketing calls the **bottom of the funnel**. That was an eye-opener. Because what lives at the **top of the funnel**—where people are still orienting themselves, learning, reflecting, or comparing—is exactly where blogging has traditionally been strongest. It’s where I’ve lived for years as a writer: helping people make sense of the digital world. Suddenly, I understood. Google is shifting too. Their AI-driven Search Generative Experience doesn’t need orientation content the same way anymore. It needs raw material to shape into an answer—**machine-readable**, **factual**, **transactional**. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/04/MMS-105.jpg) Here I get the information from Arjen (right). A moment of realisation. At the [Amsterdam AI Salon](https://aisalon.amsterdam). ## The Shift in the Landscape At first, I took the de-indexing personally. I thought it must have been a technical issue—and to some extent, I still suspect part of it might be. But once I started digging, I found I wasn’t alone. This has been happening for months. Entire swathes of top-of-the-funnel content are being de-indexed. And while I don’t write spam and I don’t churn out thin content, I do write from a human-first, exploratory angle—exactly what the system now devalues. The landscape has shifted. Google is evolving toward a hybrid AI/search model. And what it needs now is not just web pages—it needs **clean, structured, narrative-ready data** to feed its models. It wants to **assemble the story**, not link to it. It wants to answer the user directly. ## So Where Does That Leave Me? So, I’m not taking it personally anymore. I’m adjusting. This blog will evolve. The idea that I can write something, wait, and let Google deliver the right audience to me—that idea is gone. From now on, I will share what I learn, directly. Through newsletters, through social media, through communities. Google is no longer my primary amplifier; people are. That’s a big shift, but it’s also liberating. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/04/image.png) The process of being deindexed. ## This Is the First Step Something else clicked recently, and it’s worth adding to this picture. Many people have started to notice—and complain—that Google Search isn’t what it used to be. That it’s harder to find thoughtful, human answers. That search is starting to feel empty, even when it's full of results. And I think this is the missing puzzle piece. Search is no longer primarily for humans. It’s becoming an **index for machines**. An index for large language models to **build natural-language responses**, not to direct traffic to human-written pages. That doesn’t just change how content is surfaced—it changes the *reason* content exists in the first place. It means the search engine’s true user may no longer be you. It’s the AI system that’s building answers on your behalf. That’s the shift we’re in. And we need to reckon with it. This article is the start of that transition. If you’re a digital professional feeling the same shift, let’s talk about it. We’re not just watching the landscape change—we’re part of the generation that will redefine it. ### Further reading - [Web Traffic and the Rise of LLMs](https://rob-hoeijmakers.ghost.io/web-traffic-and-the-rise-of-llms/) ### Descript - From Visual Storytelling to Storytelling with Visuals URL: https://rob-hoeijmakers.ghost.io/descript/ Last updated: 2026-05-17T13:38:40.000Z ### My First Experiences with Descript I'm genuinely excited about [Descript](https://get.descript.com/2m457a55pcg2). Why? Because it flips the entire logic of video editing. Traditional tools like iMovie or Premiere are built around visual storytelling: image comes first, speech follows. Descript turns that upside down—letting you start with what you say, and *then* shape how it looks. It enables *storytelling with visuals*, not visual storytelling. And that change is huge. ## AI as the bridge: from soundwaves to text The magic lies in how Descript uses AI. The moment you record something—an interview, a voiceover, a live session—it transcribes the audio into text. Instantly. That transcription becomes your timeline. You’re no longer scrubbing through clips or guessing where something happened. You edit your video by editing text. That’s the real power: AI turns speech into editable material. Not just for convenience, but to unlock a completely different workflow. Here I am at work: ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/03/2025-03-30_16-12-10.jpg) The interface, with the video I made below. Here is the result: ## From struggling with iMovie to freedom through text I’ve spent years wrestling with tools like iMovie. Everything revolves around image—timing, layers, visual expression—and I simply don’t speak that language. I don’t know the lingo, and I don’t always understand the expressive codes. And although I enjoy explaining things and doing interviews, anything over a minute became difficult to produce in one go. What I really wanted was to work with text *and* video—on equal footing. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/03/iMovie-versus-Descript.jpg) From a visual timeline in iMovie to a textual timeline in Descript. Two different worlds. ## Editing interviews like editing text Descript makes that possible. Record a video interview—via Zoom, Teams, [Meet](https://rob-hoeijmakers.ghost.io/choosing-google-one-to-use-google-meet/) or otherwise—and you instantly have a transcript. From there, you can cut, rearrange, and refine just like editing a document. You can break things into scenes, add titles, remove filler words or silences, or even use AI to generate missing bits in your own voice. Yes, even dubbing and translation are possible—with your own voice, in another language. ## Image as a second language I create my title cards in [Canva](https://rob-hoeijmakers.ghost.io/canva) and import them into Descript. That feels natural. I still work from the story, from the words. Only after that comes the layout—picture-in-picture, subtitles, background footage, or stock visuals from Descript’s own library. What’s crucial: I don’t need to *learn how to think in visuals*. I can stay rooted in what I know—speaking, writing, reading—and still create something that feels rich and polished. ## Direction starts with speech What I’ve noticed already is that I think differently while recording. Because I know I’ll be editing the *text* later, I start making intuitive decisions as I speak. I feel more like a director—even while simply doing an interview. The editing process starts with the words, not the footage. And that’s powerful. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/03/Descript-start-page.jpg) Welcome page in the Descript app. With translation and dubbing options. ## One caveat: foreign languages One limitation I’ve run into is when working in a language other than English—in my case, Dutch. Transcription isn’t always accurate, and dubbing sometimes misses the mark. It works, but don’t expect the same polish or precision you’d get with English. Worth keeping in mind if you’re planning multilingual work. ## Final thoughts Descript is more than a video editor. It’s a rethink of the creative process. It empowers people like me—who work through language, not images—to become visual creators without needing to *become* visual creators. And AI is what makes that possible, by turning soundwaves into text, and text into a timeline. I still have a lot to learn, but one thing is clear: I’ve found a new language for storytelling. And it starts with words. Will keep you updated of the progress. --- ### Further reading - [Descript: Edit Videos & Podcasts Like a Doc | AI Video Editor](https://get.descript.com/2m457a55pcg2) ### Fibre Feels Faster — And It’s Not About Speed URL: https://rob-hoeijmakers.ghost.io/fibre-feels-faster-and-its-not-about-speed/ Last updated: 2026-05-10T08:54:04.000Z I recently[ switched from coaxial cable internet to full fibre](https://rob-hoeijmakers.ghost.io/switch-coax-fibre-internet/). Like many, I had always focused on raw speed — chasing the highest Mbps, measuring download tests, upgrading whenever a new, faster package became available. But making the jump to fibre changed my perspective entirely. It wasn't the gigabit download speeds that impressed me. It was how **responsive** everything became. Web pages snapped open. Video calls felt more natural. Cloud apps no longer lagged. Surprisingly, the magic had little to do with maximum speed. The difference was all about **latency, symmetry, and stability**. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/03/IMG_9385.jpeg) Fibre popping up. ## The Illusion of Top Speed We’re conditioned to think that faster internet means faster everything. But in practice, most daily activities use a fraction of your available bandwidth: - Browsing, messaging, video calls, email — they all work fine on 50-100 Mbps. - 4K streaming? About 15-25 Mbps per stream. - Video calls? Typically under 5 Mbps per stream. More bandwidth helps if you're downloading huge game files, doing large cloud backups, or have multiple heavy users in the same household. But most of the time, **you don’t need more speed**. What makes the connection *feel* fast is something else: low latency. ## Latency: The Acceleration, Not the Top Speed Think of it like driving a car. The top speed might be 250 km/h, but how often do you drive that fast? What you feel every day is **acceleration** — the immediate response when you hit the gas. Latency is your internet's acceleration. It measures the time it takes for a tiny data packet to reach its destination and come back. On coax (DOCSIS), my latency hovered around **20-30 milliseconds** on a good day, with spikes during busy hours. On fibre? **3 to 5 milliseconds**, stable. That difference is enormous, especially in how it *feels* when working, browsing, or on video calls. Websites open instantly. Cloud apps stop hesitating. It’s not speed — it’s torque. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/03/Fibre-low-latency.jpg) Latency on fibre is minimal. ## Symmetrical Upload: The Hidden Power Another quiet revolution: fibre gives you **the same speed up and down**. Most cable plans throttle upload — 500 Mbps down but 40 Mbps up is common. Symmetrical speeds matter more than we think: - Video calls and screen shares run smoother. - Cloud backups and uploads finish faster. - AI tools, video processing, and modern workflows increasingly rely on solid upload speeds. With fibre, I stopped thinking about uploads. They just *go*. ## Stability: No More Sharing the Neighbourhood Cable internet is a shared medium. Your neighbours' peak usage can slow you down. Fibre changes that. It’s a dedicated line from your home to the provider's network. No peak slowdowns. No fighting for bandwidth. Just consistent performance, all day. ![Bus versus star network. Your own line.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/03/Bus-Network.jpg) Bus versus star network. Your own line, and not shared with neighbours. ## WiFi: The Last Bottleneck Even with fibre and top speeds coming into your home, **WiFi often becomes the limiting factor** for your devices. Walls, distance, interference, and device limitations all reduce your actual speeds over wireless. That’s why I made sure to **wire the most important devices** directly: computers, NAS, and anything critical for work or large data transfers. It’s the only way to fully experience the advantages fibre offers. ## Global Perspective: How Fortunate We Are Here’s what surprised me the most. After digging into global internet data, I realised how rare my new connection is: - **Global fixed broadband average**: \~120 Mbps download / 50 Mbps upload - **Mobile internet average**: \~50 Mbps down / 12 Mbps up - Latency in many parts of the world: **30 to 80 milliseconds** Fibre is growing, but full symmetrical gigabit connections with low latency are still mostly limited to places like South Korea, Japan, Scandinavia, and select European cities (like Amsterdam). In global terms, this isn’t normal — it’s a quiet privilege. ## What I Learned Switching to fibre wasn’t about hitting new speed records. It was about feeling the internet respond the way it should. Smooth. Immediate. Reliable. I stopped chasing bandwidth numbers. What matters more is this invisible layer of quality: low latency, stable uploads, and not sharing a congested line. That’s what makes fibre truly fast. And the best part? It’s not about speed anymore. --- ### Further reading - [Switching to Optical: My Journey to Fibre Internet](https://rob-hoeijmakers.ghost.io/switch-coax-fibre-internet/) 💡 There are two fundamentally different types of fixed internet networks in use today: cable networks and fibre-optic networks. Cable networks typically use a system called ****Hybrid Fibre Coax (HFC)**, where fibre runs to a neighbourhood cabinet, but the final connection to the home is made over coaxial cable. These networks were originally built for cable television and are often closed systems, meaning only the cable provider can offer services on them. While download speeds can be high, upload speeds are usually limited, and capacity is shared with others in the area — which can affect performance during busy periods. ****Fibre-optic networks (FTTH — Fibre to the Home)** deliver fibre connections all the way into the home. They provide symmetrical speeds — fast uploads and downloads — with low latency and stable performance, even under heavy use. Fibre networks are often designed as open systems, allowing multiple service providers to operate on the same infrastructure. This makes fibre a more future-proof option, ready to handle growing digital demands. ### Switching to Optical: My Journey to Fibre Internet URL: https://rob-hoeijmakers.ghost.io/switch-coax-fibre-internet/ Last updated: 2026-04-25T13:49:11.000Z For years, I stuck with [coaxial internet](https://rob-hoeijmakers.ghost.io/fibre-feels-faster-and-its-not-about-speed/). It worked, but prices steadily climbed while speeds plateaued. Fibre optic internet had been promised for years, but recently it became a real, physical option — not just in the street, but ready to be pulled into the home. What followed was a surprisingly satisfying mix of hands-on work, infrastructure excitement, and the thrill of breaking free from a monopoly. ## The Big Day: Fibre Enters the House The network operator in my area is **Open Dutch Fiber**, the company responsible for laying and owning the fibre infrastructure. Yesterday, their crew showed up — four men working with surprising force — drilling, digging, and pulling the cable under the garden and terrace. It was rough work for something as delicate as a glass fibre line, but by the end of the day, the connection reached inside my home, terminating at a small optical socket on the wall. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/03/IMG_1766.jpeg) The green cable is the coax. The glass fibre needs to be installed at this point. From there, I took over. With a self-install package from **Odido**, my chosen service provider, I completed the installation myself. Fibre connects to a **media converter**, which turns light signals into a standard LAN connection. One Ethernet cable later, my router was online — the switch complete. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/03/IMG_1768.jpeg) This is the glass fibre end point. From here it can be connected to a media converter and router. From that point, I adjusted my Wi-Fi settings. Thanks to the Mac ecosystem, sharing Wi-Fi details was seamless, while my Macs remained connected via LAN. I fine-tuned the setup, configuring the wireless printer (WPS), setting up router-based rules for incoming [VPN connections](https://rob-hoeijmakers.ghost.io/virtual-private-network-vpn/), and assigning a fixed IP to the NAS. Surprisingly, the most time-consuming part was manually entering account credentials into the [Kobo e-readers](https://rob-hoeijmakers.ghost.io/kobo-vs-kindle/). ## The Immediate Win: Speed, Cost, and Competition Right away, the difference was clear. For years, I paid **€75 monthly** for around **600 Mbps** on coax. With fibre, I now pay **€25** for comparable speed — thanks to a promotional rate, but that’s beside the point. The real shift is infrastructure-driven competition. Open Dutch Fiber operates the network, but you can **choose your provider** — Odido, or one of several others offering services over the same line. If I ever want to switch, it’s just a matter of changing providers — no need to lay new cables. And competition is heating up. Dutch telecom giant **KPN** is rolling out its own fibre network in the same area. Yes, the streets will be opened up again — noisy and messy — but the result? True network-level competition. In time, residents will have the option to switch **not just providers**, but **networks** altogether. That’s a significant shift after decades of being locked into a single coaxial line. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/03/IMG_1788.jpeg) The media converter. Optical wire, the yellow one, comes in, LAN-cable goes out to the router. ## Fixed Wireless Access (FWA): 5G as a Serious Alternative Another interesting development is the rise of **Fixed Wireless Access (FWA)** — a broadband connection over 5G. Providers like **Odido** now offer FWA as a genuine alternative or backup to fibre and cable. A 5G antenna connects your home to the mobile network, delivering broadband speeds without a physical cable. While fibre remains the most robust choice for speed and stability, FWA adds flexibility. It works especially well when fibre to the antenna is available or for homes where digging is complicated. As networks improve, FWA strengthens competition in what was once a coax-dominated market. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/03/fwa.jpeg) Odido FWA modem ## Keeping It Simple: Unbundled and Free What made this switch even easier was staying **unbundled**. No TV, no special TV channels, no traditional landline — just internet. It meant I wasn’t tied down by complex packages or contracts. My current provider offers some bundling, but I keep it light, treating internet as what it has become: a utility. The freedom to switch providers — or even networks — whenever better service or pricing comes along is how connectivity should work. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/03/Unbundled.jpg) Unbundled setup. Internet only. ## Final Thoughts: Real Infrastructure Change Making the switch felt good — exciting, a little daunting, but deeply satisfying. More than faster speeds or lower costs, this is about the shift towards real digital infrastructure — where choice is built in, and you’re no longer stuck with a single provider or network. It’s a reminder that progress often happens quietly, beneath our streets and gardens, until one day you realise — the speed of light now runs straight into your home. [Fibre Feels Faster — And It’s Not About SpeedI used to chase top speeds. Fibre showed me what really matters: low latency, equal uploads, and no peak slowdowns. That’s when the internet feels fast.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-46.png)Rob Hoeijmakers (robhoeij)Rob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_4704-1.jpeg)](https://rob-hoeijmakers.ghost.io/fibre-feels-faster-and-its-not-about-speed/) ### How ZEISS and ASML Enable the Modern Chip Industry URL: https://rob-hoeijmakers.ghost.io/zeiss-and-asml/ Last updated: 2026-05-10T08:54:04.000Z ASML's dominance in the global chip industry is a prime example of how European technology and collaboration can shape the world market. Less visible, but equally fascinating, is the role of Germany's [ZEISS](https://zeiss.com) in this story. Together, they form the silent force behind the modern digital world. ## European Roots: ASM and Philips ASML was founded in the early 1980s as a joint venture between [**ASM International**](https://www.asm.com) and [**Philips**](https://philips.com). ASM is a Dutch pioneer in thin film technology for the semiconductor industry, founded by Arthur del Prado. Philips contributed capital, technology, and manufacturing experience. Both companies recognised that the future lay in **lithography**, etching patterns onto silicon using light, the core of chip production. This collaboration laid the foundation for what would later become ASML: an independent player that would completely reshape the lithography world. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/03/FHD-LqRX0AYqej4.jpeg) Humble roots of ASML in Eindhoven ## ZEISS: A Divided Giant from Germany Parallel to this development, ZEISS experienced its own unique European history. After World War II, the company was split: - **Jena** ended up in East Germany (GDR) - **Oberkochen** was rebuilt in West Germany (FRG) Both branches developed separately for decades, with the fall of the Berlin Wall marking the turning point towards reunification. That complex history remains visible, even in the **ZEISS logo**, where the elliptical curve at the bottom refers to a lens and the company's optical heritage. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/03/GmWiEv8WoAAF4YK-2.jpeg) Zeiss's history in logo's ## The Power of Light: Lithography and EUV ASML's technology revolves around **projecting light patterns onto silicon**. The latest generation of machines uses **extreme ultraviolet light (EUV)** with a wavelength of 13.5 nanometres. This allows chipmakers to print transistors at an atomic scale. However, EUV only works with optics of unprecedented precision. [**Carl Zeiss SMT GmbH**](https://www.zeiss.com/semiconductor-manufacturing-technology/home.html) is the world's sole supplier of the required mirrors and lenses, so precise that even a molecular deviation would be fatal. ## Strategic Interdependence The dependency is mutual and absolute. ASML: - invested **€1.5 billion** in Carl Zeiss SMT - acquired a **24.9% stake in 2016** - has since received **billions in dividends** Carl Zeiss SMT's revenue grew from **€1.2 billion (2016)** to **€4.1 billion (2024)**. ASML openly calls Zeiss SMT the **critical link** in its supply chain: without Zeiss, Veldhoven grinds to a halt. [ASML Obtains Regulatory Approvals for Strengthened Partnership With Zeiss![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon-10.ico)ASML](https://www.asml.com/en/news/press-releases/2017/asml-obtains-regulatory-approvals-for-strengthened-partnership-with-zeiss) ## Technology, Geopolitics, and Economic Power Each EUV machine costs **over $150 million** and is only supplied to allied nations. China is kept out. **Seventeen years of research** and **multi-billion investments** led to this dominant position. This is not an American or Chinese success story but a **European technological triumph**, born of collaboration, reconstruction, and precision. [The $150 Million Machine Keeping Moore’s Law AliveASML’s next-generation extreme ultraviolet lithography machines achieve previously unattainable levels of precision, which means chips can keep shrinking for years to come.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon-11.ico)WIREDWill Knight![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/Business-ASML---Employees-assembling-an-EUV-system--ASML-.jpg)](https://www.wired.com/story/asml-extreme-ultraviolet-lithography-chips-moores-law/) ## Conclusion: European Strength in a Global Industry From ASM and Philips to ASML, from Jena and Oberkochen to a reunified ZEISS: these two companies together form the backbone of the global semiconductor industry. Their interconnection demonstrates what Europe is capable of when technological forces are united. The **light of Europe**, literally captured in lenses and mirrors, sets the global standard for the digital future. ### Further reading - [Strengthening Europe’s Semiconductor Backbone: The Role of ESMC](https://rob-hoeijmakers.ghost.io/esmc/) ### BOAT: A New Category by Gartner and the Future of Enterprise Automation URL: https://rob-hoeijmakers.ghost.io/boat-enterprise-automation/ Last updated: 2025-04-12T15:11:00.000Z This week, I attended a conference aboard the **SS Rotterdam**, hosted by **Incentro**, where we discussed a fascinating new concept introduced by **Gartner**: **Business Orchestration and Automation Technologies (BOAT)**. Standing on the deck, looking over the skyline of Rotterdam, it became clear to me why this new category makes so much sense. ## Why BOAT? For years, enterprise automation has been fragmented. **Robotic Process Automation (RPA), Business Process Automation (BPA), Low-Code Platforms, and iPaaS** (Integration Platform as a Service) have been evolving in parallel. Yet, their overlap has grown significantly, and organizations increasingly need a **unified framework** to orchestrate these technologies efficiently. [Gartner's BOAT category](https://www.gartner.com/en/webinar/621105/1385624) consolidates these fragmented disciplines into a single, overarching approach, allowing businesses to streamline automation across departments and systems. The fundamental idea is to create an integrated **automation fabric** that can intelligently coordinate tasks, workflows, and decisions at scale. ![Automation markets are overlapping](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/03/Figure_1_Automation_Markets_are_Overlapping.jpeg) Automation markets are overlapping ## Incentro’s Position in the BOAT Ecosystem [Incentro](https://www.incentro.com/en-NL), a digital transformation specialist, is well-positioned in this emerging landscape. Their deep expertise in cloud-native development, automation solutions, and enterprise integration aligns naturally with the principles of BOAT. This event reinforced their role as a key player in helping businesses adopt **end-to-end process automation** that goes beyond isolated tools and into fully orchestrated digital ecosystems. ![Field CTO Floris Weegink and conference host Vivianne Bendermacher](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/03/IMG_1606.jpeg) Field CTO Floris Weegink and conference host Vivianne Bendermacher ## Companies Shaping the BOAT Landscape Alongside Incentro, several other companies are actively shaping the BOAT category, each contributing to different aspects of automation and orchestration: - [**Camunda**](https://camunda.com) – Focuses on process orchestration, enabling businesses to design, automate, and optimize workflows efficiently. - [**UIPath**](https://www.uipath.com/) – A leader in Robotic Process Automation (RPA), specializing in automating repetitive tasks with AI-powered bots. - [**Mendix**](https://www.mendix.com/) – A low-code development platform that accelerates application creation and integrates with automation workflows. - [**Snowflake**](https://www.snowflake.com/en/emea/) – A cloud-based data platform that enables data sharing, integration, and automation across enterprise systems. - [**Workato**](https://www.workato.com/) – A no-code automation platform that connects apps, automates workflows, and drives business process integration. - [**Glean**](https://www.glean.com) – An AI-powered enterprise search tool that enhances knowledge management and improves workflow automation. These companies highlight the diversity of technologies converging under the BOAT umbrella, from **workflow automation** to **AI-driven search and decision-making**. ![Business Orchestration and Automation Technologies (BOAT)](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/03/gartner-boat-graphic.jpeg) Business Orchestration and Automation Technologies (BOAT) ## Glean: The Intersection of AI, Search, and BOAT Another topic that caught my attention at the event was [**Glean**](https://www.glean.com/), an AI-powered enterprise search platform. Designed to **seamlessly index company knowledge across multiple platforms**, Glean enables employees to quickly find relevant information across internal systems, much like how **Google searches the web**. At first, I wondered: **How does Glean fit into BOAT?** The connection became clear when considering how **knowledge accessibility fuels automation**. A BOAT framework that integrates Glean can **eliminate bottlenecks caused by information silos**, making automation more effective. Imagine an AI-driven automation system that can instantly retrieve the right document, policy, or customer history—without human intervention—before executing a process. That’s where Glean and BOAT intersect. [Work AI for all - AI platform for agents, assistant, searchGlean is the Work AI platform connected to your enterprise’s data. Find, create, and automate anything. Explore what Work AI can do for you!![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/67189bf127c626679a308b22_32x32-20Glean-20Favicon.png)Databricks![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/66df3ec7d95ad78e65888721_Website-20preview-20card.webp)](https://www.glean.com) ## The Broader Impact of AI and LLMs on Enterprise Automation Generative AI and **Large Language Models (LLMs)** are fundamentally reshaping how businesses think about automation. The ability of LLMs to **understand context, interpret natural language, and generate insights in real-time** is accelerating the adoption of **AI-driven automation strategies**. When combined with **BOAT**, [LLMs](https://rob-hoeijmakers.ghost.io/a-comparative-analysis-of-top-language-models/) have the potential to take automation beyond simple workflows into true **cognitive orchestration**. This means: - AI agents proactively assisting employees by retrieving relevant data before they ask. - Automated decision-making based on contextual insights rather than predefined rules. - Intelligent process automation that adapts dynamically to business needs. ## Final Thoughts The **BOAT framework** marks a shift towards **fully orchestrated, intelligent automation**. As AI, search, and process automation converge, the future of enterprise IT will be defined by how well companies integrate these technologies into their workflows. The conference aboard the SS Rotterdam was a fitting setting to discuss this topic—a reminder that just like navigating a ship, **business automation needs a clear strategy, coordination, and the right technology to stay on course**. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/03/IMG_1582.jpeg) SS Rotterdam ### Speech recognition, Speech-to-Text, Dictation, and Transcription: What’s the Difference? URL: https://rob-hoeijmakers.ghost.io/asr-speech-to-text-dictation-and-transcription-whats-the-difference/ Last updated: 2026-05-26T19:34:48.000Z ## Choosing the Right Tool for the Job A few decades ago, choosing how to write meant deciding between a pencil, a pen, or a typewriter—each suited to a different task. Today, with speech technology, we face similar choices: When should we use **voice commands, live captions, dictation, or transcription?** Understanding these fundamental technologies ensures we pick the right tool for the job, whether it's issuing hands-free commands, getting accurate captions, or turning spoken words into structured text. Let’s break down the differences. ## Automatic Speech Recognition (ASR) vs. Speech-to-Text: The Foundation Speech recognition technology is everywhere, but how does it really work? To understand the difference between ASR and Speech-to-Text, we need to start with the basics—how machines process human speech and turn it into something useful. ### **ASR: The Technology That Listens** Automatic Speech Recognition (ASR) is the **brain behind speech technology**. It listens to audio, recognizes words, and converts them into text. ASR relies on: - **Acoustic models** to understand sound patterns - **Language models** to predict word sequences - **Machine learning** to improve accuracy over time Think of ASR as the "hearing" part of a voice assistant—it's listening, but it doesn’t always understand perfectly. ### **Speech-to-Text (StT): Turning ASR into Usable Text** Speech-to-Text (StT) takes ASR’s raw output and makes it **usable for humans** by adding: - Punctuation and capitalization - Better formatting - Improved readability **ASR is the engine, Speech-to-Text is the final product.** #### **Key Differences** | Feature | ASR (Automatic Speech Recognition) | Speech-to-Text (StT) | | ------------------ | ------------------------------------------------ | ---------------------------------------- | | **What it does** | Converts speech into raw text. | Produces human-readable text. | | **Use cases** | Voice assistants, search queries, live captions. | Transcription, dictation, meeting notes. | | **Output quality** | May contain errors and missing punctuation. | More structured, readable text. | ## Dictation vs. Transcription: Not the Same Thing Transcribing speech and dictating text might seem similar, but they serve different purposes. Dictation is **intentional speech-to-text**, where the speaker controls the words and pauses for clarity. Transcription, on the other hand, captures speech as it naturally happens, often including multiple speakers and requiring post-processing. ### **Dictation: Talking to Your Device on Purpose** [Dictation](https://rob-hoeijmakers.ghost.io/from-typing-to-talking/) is when you **speak deliberately** to produce written text. You might: - Dictate a message on your phone. - Speak out an email. - Use voice typing to write a report. Dictation is **structured speech-to-text**—you control the words and often **pause for clarity**. [From Typing to TalkingFrom speech synthesis to dictation, I explore how evolving AI tools make communication swifter and more natural. Dive into my journey through spoken and written digital modes.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-33.png)Rob Hoeijmakers (robhoeij)Rob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_8495-2.jpeg)](https://rob-hoeijmakers.ghost.io/from-typing-to-talking/) ### **Transcription: Capturing Natural Speech** Transcription is more like a **fly on the wall**—it captures spoken words as they happen. It’s used for: - [**Meetings and interviews**](https://rob-hoeijmakers.ghost.io/fireflies-ai-transcriptions/) (where multiple speakers talk naturally) - **Courtroom and medical documentation** - [**Podcasts**](https://rob-hoeijmakers.ghost.io/ai-is-transforming-podcast-production/) **and videos** Transcription often **requires cleanup**, such as speaker identification and punctuation. [Transcribing Audio on a Mac and iOS: My Workflow and ToolsNeed seamless transcription on Mac & iOS? Apple’s built-in tools are limited—here’s how I use Transcribe for accurate, multilingual, cloud-based audio transcription.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-35.png)Rob Hoeijmakers (robhoeij)Rob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_0860.jpeg)](https://rob-hoeijmakers.ghost.io/transcribe-audio-on-macos-and-ios/) #### **Key Differences** | Feature | Dictation | Transcription | | -------------------------- | ------------------------------------------------ | --------------------------------------------------- | | **How speech is recorded** | Speaker controls and dictates. | Natural speech is captured as is. | | **Editing needed?** | Usually minimal. | Often requires corrections. | | **Typical users** | Professionals writing reports, emails, or notes. | Journalists, researchers, and legal/medical fields. | ## Live Captions: A Special Case of Transcription Live captions are **real-time transcription**, but because they’re generated instantly, they prioritize **speed over accuracy**. | Feature | Live Captions | Automatic Transcription | | ------------- | ----------------------------------- | -------------------------------------- | | **Speed** | Instant. | Processed after recording. | | **Accuracy** | Lower, due to real-time processing. | Higher, since errors can be corrected. | | **Use cases** | Accessibility, live events. | Meeting transcripts, official records. | ## ASR Without Text: What About Voice Commands? Did you know **not all ASR generates visible text?** Many ASR-based systems **never show you what they transcribe** because they’re built to trigger actions instead. ### **Examples of ASR Without Text Output:** - **Voice Assistants:** "Turn off the lights" → ASR processes → Lights switch off. - **Voice Search:** "Best coffee shop near me" → ASR converts speech into a search query. - **Navigation:** "Take me to Central Station" → ASR processes the command → GPS system responds. Here, ASR isn’t about producing readable text—it’s about recognizing intent. [Beyond keyboards and screensReconnecting with Bram led to a bold experiment: writing this entirely by voice. Here’s why the future of work may no longer need keyboards or screens.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-34.png)Rob Hoeijmakers (robhoeij)Rob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_8674-1.jpeg)](https://rob-hoeijmakers.ghost.io/beyond-keyboards-and-screens/) ## Wrapping It Up: A Simple Hierarchy ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/03/Automatic-Speech-Recognition--ASR-.jpg) ASR (Automatic Speech Recognition) ### **Final Takeaways** Choosing the right speech tool means understanding when to use ASR, Speech-to-Text, dictation, or transcription: - **ASR is the core technology** powering speech recognition. - **Speech-to-Text makes ASR output human-readable.** - **Dictation is controlled, while transcription captures free speech.** - **Live captions are fast but less accurate.** - **Not all ASR produces readable text—voice commands process speech without displaying it.** Understanding these distinctions helps you **choose the right tool for the job**, whether it’s voice-controlled automation, live captions, or structured documentation. --- ### Further reading - [The Silent Revolution: How AI is Transforming Speech Recognition](https://rob-hoeijmakers.ghost.io/speech-recognition/) ### The Silent Revolution: How AI is Transforming Speech Recognition URL: https://rob-hoeijmakers.ghost.io/speech-recognition/ Last updated: 2026-05-10T08:54:05.000Z Imagine if every conversation you had was transcribed perfectly—no matter the accent, speed, or background noise. Imagine that your phone, your car, and even your email assistant could understand spoken language with near-human accuracy. This isn’t just a futuristic dream—it’s happening now, thanks to a breakthrough in AI called **self-attention**. For decades, speech recognition technology struggled with accuracy, speed, and adaptability. But something changed in recent years, making it *dramatically* better. This shift is so profound that it’s reshaping how we interact with technology—affecting everything from voice assistants to real-time captions, automated customer service, and AI-powered content creation. So, what exactly happened? A breakthrough in **self-attention** gave speech recognition an *engine upgrade*, boosting accuracy and efficiency like never before. ## The Problem: Why Speech Recognition Was Hard Understanding speech is surprisingly difficult for machines. Some of the biggest challenges include: - **Speed Variability** – People speak at different speeds, making it difficult for models to keep up. - **Contextual Meaning** – The meaning of words depends on surrounding words (e.g., "I read a book" vs. "I will read a book"). - **Noisy Environments** – Background sounds make speech harder to transcribe correctly. - **Similar-Sounding Words** – "their" vs. "there" or "to" vs. "two" can be confusing for machines. Older speech recognition models tried to solve these issues by processing speech *step by step*, but they struggled with long sentences, accents, and background noise. Then, something changed. ## The Breakthrough: How Self-Attention Changed Everything The **self-attention mechanism**, the core of Transformer models like OpenAI’s [**Whisper**](https://rob-hoeijmakers.ghost.io/openais-whisper-speech-to-text-technology/) and Meta’s **wav2vec 2.0**, introduced a completely new way to handle speech. Instead of processing speech **bit by bit**, like traditional AI models, **self-attention allows AI to analyze an entire audio sequence at once**. Think of it like this: - **Old AI models** tried to read a book one word at a time without knowing what came next. - **Self-attention** lets the AI see the whole sentence at once and understand what makes the most sense. This new approach solves many of the past challenges: ✅ **Context Awareness** – The AI can "see ahead" in the sentence and understand meaning better. ✅ **Faster Processing** – Since it doesn’t have to go word-by-word, speech is recognized in real-time. ✅ **Higher Accuracy** – The AI can handle different accents and noisy environments more effectively. ## Behind the Scenes: How ASR Really Works Modern speech recognition isn’t just one AI model doing everything—it’s actually a combination of different AI systems working together: ### Acoustic Model (Listening to Sounds) - The AI listens to raw audio and converts it into a sequence of phonemes (the smallest units of sound in speech). - This part of the system is often powered by deep learning models like **wav2vec 2.0**, which have been trained on massive amounts of audio. ### Language Model (Understanding Meaning & Fixing Errors) - The first transcription is usually rough—it might miss words or have small errors. - A separate **language model** (like GPT-based systems) corrects errors, adds missing words, and ensures proper grammar and punctuation. - This is where **AI-powered auto-punctuation** comes in, making transcripts more readable. ### Final Formatting & Contextual Adjustments - Depending on the application (subtitles, meeting notes, voice assistants), additional tweaks are made to improve clarity and structure. By using **different** [**AI models**](https://rob-hoeijmakers.ghost.io/a-comparative-analysis-of-top-language-models/) **at different stages**, modern ASR systems can **combine speed, accuracy, and fluency**, making speech-to-text more natural than ever. So, speech recognition **isn’t a perfect science—it’s closer to an art**, shaped by **context, probability, and educated guessing**. AI has become far better at this “guesswork” than older deterministic models, which is why the leap in performance feels so dramatic. ## How Speech Recognition Transformed My Digital Habits The improvements in speech recognition have drastically changed how I interact with technology. I now: - Use [**dictation on my Mac**](https://rob-hoeijmakers.ghost.io/asr-speech-to-text-dictation-and-transcription-whats-the-difference/) more than ever, replacing much of my typing. - **Speak to ChatGPT** more often than I type, making interactions faster and more natural. - Have made **voice memos an integral part of my routines**, capturing ideas and thoughts instantly. - **Use** [**transcription**](https://rob-hoeijmakers.ghost.io/transcribe-audio-on-macos-and-ios/) **by default for online meetings**, ensuring I never miss key details. This shift has not only changed my habits but has **significantly improved my productivity and efficiency**. With AI handling speech so seamlessly, I can focus more on thinking and communicating rather than typing and transcribing. ## The Future: Localized and Private Speech Recognition As AI-driven ASR continues to improve, we are likely to see a shift towards **more local processing on personal devices**, making speech recognition: - **More private** – Instead of sending voice data to the cloud, ASR will work locally, reducing privacy concerns. - **Faster and more efficient** – Local processing means real-time transcription with less latency. - **More personalized** – With integration at the [**OS level**](https://rob-hoeijmakers.ghost.io/os-level-translations/), ASR will adapt to **user-specific dictionaries**, handling names, brands, and industry-specific jargon more accurately. This shift will **enhance usability, security, and accuracy**, making speech recognition an even more powerful tool in our digital lives. --- ### Further reading - [Whisper by OpenAI: Unlocking Speech Recognition for Transcription and Beyond](https://rob-hoeijmakers.ghost.io/openais-whisper-speech-to-text-technology/) - [Transcribing Speech on a Mac and iOS: My Workflow and Tools](https://rob-hoeijmakers.ghost.io/transcribe-audio-on-macos-and-ios/) ### Lost in ChatGPT's Models? A guide (March 2025) URL: https://rob-hoeijmakers.ghost.io/chatgpt-models-guide/ Last updated: 2026-04-27T19:22:58.000Z I'm a power user of **ChatGPT Plus** — someone deeply immersed in OpenAI's ecosystem, regularly exploring new features. Yet recently, I found myself completely lost in the labyrinth of models, variations, and capabilities available. GPT-4o, GPT-4.5, o3-mini… I realised clarity was needed. Here’s my simplified take, intended to help you quickly choose the right model, without getting lost like I did. ## Why Are There So Many Models? OpenAI has expanded rapidly, introducing specialised models to handle diverse user needs: - **General Tasks (GPT-4o):** Multimodal (text, image, audio), great for everyday use. - **Creative & Structured Tasks (GPT-4.5):** Optimised to minimise hallucinations and enhance creativity. - **Advanced Reasoning (o3-mini & o3-mini-high):** Specialised for complex logic, math, and coding. Yet, the abundance of variants has made selecting the right one surprisingly challenging—even for experienced users like myself. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/03/ChatGPT-Versions-March-2025.jpg) ChatGPT Versions - March 2025 ## Which Model for Which Task? Here's a quick, intuitive guide to help you choose wisely: ### 1\. General, Everyday Use: GPT-4o Think of GPT-4o as your all-round assistant—capable, versatile, and quick: - **Multimodal**: Excellent for interpreting images, generating text, and understanding audio. - **Default choice** for most questions, quick brainstorming, or general productivity. ### 2\. Creative Work & Structured Writing: GPT-4.5 If your work demands creativity and precision, GPT-4.5 is your specialist: - Ideal for marketing copy, storytelling, brand communications. - Minimises hallucinations, ensuring trustworthy output for professional use. ### 3\. Logic & Reasoning: o3-mini (and o3-mini-high) For logic-heavy tasks, coding, and step-by-step problem-solving: - **o3-mini**: Fast reasoning, perfect for moderately complex questions. - **o3-mini-high**: Stronger reasoning, best suited for intricate logical tasks and code-related queries. ## Quick Selection Overview: | Scenario | Recommended Model | | --------------------------- | ----------------- | | Everyday tasks | GPT-4o | | Creative, structured tasks | GPT-4.5 | | Logic, coding, reasoning | o3-mini-high | | Faster, simpler logic tasks | o3-mini | ## My Recommendation for Most Users: - **Default:** Always start with GPT-4o. - **Creative projects:** Jump straight to GPT-4.5. - **Reasoning:** o3-mini is your companion. ## Personal Shortcuts (How to Quickly Remember): - **"o" means omni:** GPT-4o does everything. - **"4.5" means refined:** GPT-4.5 refines creativity and accuracy. - **"o3-mini" means reasoning simplified:** Think logic, but fast. - **"o3-mini-high" means deeper logic:** When complexity demands more careful reasoning. ### Rediscovering Paste: A Productivity Boost for Mac and iOS Users URL: https://rob-hoeijmakers.ghost.io/paste-clipboard-with-history/ Last updated: 2026-05-10T08:54:05.000Z Sometimes, you stumble upon an old tool you once relied on, only to forget about it—until it suddenly reappears. That happened to me recently with *Paste*, a clipboard manager for macOS and iOS. I had used it years ago, then moved on. But now, I’m reminded of how incredibly useful it can be. Let me explain. ## A Smarter Clipboard Paste isn’t just a basic copy-and-paste tool—it’s a clipboard history manager that transforms the way you handle text and images. Typically, when you copy something, it replaces the previous item in your clipboard. With Paste, that limitation disappears. It stores a history of everything you’ve copied, letting you access and reuse past clips effortlessly. For anyone active on multiple social media platforms, working across different apps, or frequently translating snippets of text, having a clipboard history is a game-changer. It allows you to keep track of recent pastes over days or even weeks, making it easy to retrieve anything you've copied without having to redo the process. I demo the Paste app. ## The iOS Advantage: A Clipboard Keyboard One of Paste’s standout features on iOS is its dedicated keyboard. Instead of switching between apps to find something you copied earlier, you can access your clipboard history directly from the keyboard. This is especially handy when posting to social media, responding to messages, or sharing snippets of text or images. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/03/Paste-App---Keyboard.jpg) Paste Keyboard in Apple Mail and iMessage. ## Addressing Security Concerns The main reason I stopped using Paste initially was a concern over privacy. I used to copy and paste passwords, and I worried that storing them in the cloud was a security risk. Back then, I didn’t fully understand how Apple’s private cloud worked, and the idea of my passwords being accessible somewhere felt unsafe. Revisiting the app, I’ve learned that Paste has built-in security measures. It allows you to exclude sensitive content—like passwords—from being stored in the clipboard history. This means you can enjoy the productivity boost without worrying about your data security. ## A Productivity Boost Worth Revisiting Paste has once again earned its place in my workflow. It’s one of those tools that you don’t realize you need until you experience the convenience it offers. Whether you're researching, managing social media, or simply juggling multiple tasks, having a clipboard history makes everything smoother. If you haven’t tried Paste yet—or, like me, had forgotten about it—it might be worth giving it another look. You might just find it to be an essential part of your digital toolkit. --- ### Further reading - [Paste – Clipboard Manager for Mac, iPhone, and iPad](https://pasteapp.io) ### Fireflies.ai: Meeting Transcriptions & Automation URL: https://rob-hoeijmakers.ghost.io/fireflies-ai-transcriptions/ Last updated: 2026-05-10T08:54:05.000Z Meetings are an inevitable part of any professional's life. Some are insightful, some are necessary, and some just… happen. But one of the biggest challenges has always been capturing key takeaways without spending hours taking notes. [**Fireflies.ai**](https://app.fireflies.ai/login?referralCode=01K45WGDZV18D7M546M82NM6XX) has completely changed the way I handle meetings, voice data, and automation—but it's not without its flaws. ## A Silent Assistant in Every Meeting One of the things I love most about Fireflies is its seamless integration with my [**Google Calendar**](https://rob-hoeijmakers.ghost.io/digital-diary-and-calendars/). It doesn’t matter if I’m in **Teams,** [**Google Meet**](https://rob-hoeijmakers.ghost.io/choosing-google-one-to-use-google-meet/)**, or Zoom**, whether I initiated the call or someone else did—Fireflies just shows up. It joins as a virtual participant, records the meeting, and transcribes everything. No extra effort required. This single feature alone has drastically changed my workflow. I no longer need to manually start recordings, take detailed notes, or worry about missing something crucial. The AI does it all, ensuring I can focus on the conversation itself. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/03/Ask-Fred.jpg) Ask Fred is the AI Search that can help you find conversations, actions etc. He can also be used in real time. ## From Meetings to Actionable Data Once a meeting is recorded, all transcriptions are stored in Fireflies' **cloud environment**. From there, I can: - **Download transcripts** and process them in tools like **ChatGPT** to generate summaries or extract insights. - **Share key moments** directly with colleagues, instead of sending long notes. - **Use Fireflies’ built-in AI** to analyze sentiment, track decisions, or identify key speakers. What makes Fireflies stand out is its **speaker recognition and voice captioning technology**. It accurately identifies speakers, making transcripts easier to navigate—especially for multi-participant discussions. The ability to search through spoken words like you would with text is invaluable. [Review of transcription software: Tactiq versus FirefliesJoin me on a journey as I test and contrast Tactiq and Fireflies.ai, seeking the best transcription tool for video meetings in a web strategist’s life.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-30.png)Rob Hoeijmakers (robhoeij)Rob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_8046-1-1.jpeg)](https://rob-hoeijmakers.ghost.io/review-of-transcription-software-tactiq-vs-firefliesai/) ## Where Fireflies Falls Short As much as I love what Fireflies brings to the table, there are some frustrating limitations. Fireflies requires you to set a language beforehand, which is inconvenient when meetings switch between English and Dutch, leading to transcription inaccuracies. Given modern speech-to-text advancements, automatic language detection should be standard. Additionally, its AI summarization and sentiment analysis don’t support multilingual meetings, limiting their usefulness. Many AI-powered services handle multiple languages, and Fireflies should offer at least basic multilingual AI processing. Finally, a persistent annoyance is that Fireflies often requires reauthorizing Google access after logging out, which adds unnecessary friction and could be streamlined. ## Automation: Taking It to the Next Level Despite these limitations, Fireflies remains a powerful tool—especially when combined with [**automation platforms like Make.com**](https://rob-hoeijmakers.ghost.io/make-dot-com-automation/). With a bit of setup, I can: - **Send transcripts to Google Cloud or Notion** for archiving and retrieval. - **Trigger workflows based on meeting content**, like automatically sending action points to a task manager. - **Create AI-powered summaries or reports**, feeding transcripts into LLMs like ChatGPT for structured outputs. The combination of **voice data, AI, and automation** turns what used to be passive information into **actionable knowledge**. ## Why Fireflies Became Essential for Me I regularly conduct interviews with companies and organizations about products, workflows, and business processes. Transcription plays a crucial role in my productivity, allowing me to stay focused during conversations without worrying about note-taking. By working directly from the transcript, I can streamline the writing process and turn discussions into structured articles with greater accuracy and depth. In the past, I often struggled with fragmented notes, forgotten details, and the tedious process of documenting meetings afterward. Fireflies changed that by providing instant access to searchable, shareable, and automatable voice data. While it’s not flawless, the benefits far outweigh the occasional drawbacks. For anyone juggling multiple meetings per week—or even per day—this tool is a game-changer. It offloads note-taking, turning conversations into valuable, structured data that can be easily referenced and repurposed. That said, there’s still room for improvement, particularly in multilingual support, login stability, sharing settings, and customization options. I’d love to hear from others using Fireflies—how do you maximise its potential? --- ### Further reading - [Fireflies.ai - #1 AI Teammate](https://app.fireflies.ai/login?referralCode=01K45WGDZV18D7M546M82NM6XX) ### Signal vs. WhatsApp URL: https://rob-hoeijmakers.ghost.io/signal-vs-whatsapp/ Last updated: 2026-04-25T13:15:58.000Z I recently delved into Signal, the well-known WhatsApp competitor, and initially thought it was mainly a moral choice—something people use to signal their commitment to privacy. However, after deeper research, I realized that Signal is not just about virtue signaling; it's technically superior in several ways and it really benefits your privacy. ## Signal vs. WhatsApp: Encryption & Metadata Collection Both Signal and WhatsApp use the Signal Protocol for end-to-end encryption (E2EE), meaning messages and calls are protected from interception. However, privacy goes beyond encryption—metadata (who you talk to, when, and how often) is just as valuable. Signal collects almost no metadata. It doesn’t log who you communicate with or when. Even the last time you used the app is only stored in a vague way ("within the last week"). WhatsApp (Meta) collects a lot of metadata, including phone numbers, contact interactions, and device information. Even though messages are encrypted, metadata can be used for profiling and tracking. This difference exists because WhatsApp runs on Meta’s servers, where data collection is part of their business model. Signal is run by a non-profit focused purely on privacy. ## Features and Limitations While Signal excels in privacy and security, it lacks certain features that make WhatsApp appealing, especially for businesses and social interactions: ### Business and API Support - **WhatsApp Business API**: WhatsApp allows businesses to programmatically engage with customers, making it an essential tool for marketing and customer support. - **Signal has no business API**, meaning companies cannot scale communication via automation. ### Groups and Communities - **WhatsApp offers both groups and communities**, which allow large-scale discussions and interactions. - **Signal supports groups but lacks communities**, making it less ideal for larger organizations or structured group discussions. - **Polls are missing** in Signal, a small but noticeable feature that adds to WhatsApp’s engagement tools. ## Audio and Video Call Quality: Signal vs. WhatsApp When it comes to call quality, WhatsApp generally outperforms Signal, particularly in variable network conditions. **Audio Calls**: WhatsApp uses variable bitrate (VBR), which dynamically adjusts call quality based on available bandwidth. This results in clearer and more consistent audio, even in weaker connections. Signal, on the other hand, relies on constant bitrate (CBR) for privacy reasons, which can lead to choppier audio when bandwidth is limited. **Video Calls**: WhatsApp’s video calls tend to be smoother and more stable, thanks to its adaptive technology. Signal, while secure, can struggle with stuttering or lag, even on strong devices. For users who rely heavily on voice and video calls, WhatsApp currently has a clear edge in quality. [Signal rivals WhatsAppPeople often suggest using Signal over WhatsApp for privacy, but what makes Signal more private, and how does it compare to WhatsApp’s features?![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-28.png)Rob Hoeijmakers (robhoeij)Rob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_8242.webp)](https://rob-hoeijmakers.ghost.io/signal-messenger-emphasises-privacy/) I checked Signal before but if there is no network there is no use. The network is growing. ## My personal approach How I personally use Signal is to switch existing and long standing personal conversations and smaller groups to Signal when possible. Use it as a personal channel so to speak and give it more visibility and priority in my notifications. As letters. And use WhatsApp as an exchange for open notes. I stopped backing the messages up and delete them on a regular basis. I recently[ did a poll](https://rob-hoeijmakers.ghost.io/is-whatsapp-social-media/) and to my surprise 50% of the people see WhatsApp as communication and the other 50% as social media. I viewed it as communication but now also treat it as social media. That approach works really well for me. It is a minor change to my routines but stops giving away useless (but priceless) information to a company. ### Connecting a Kobo Reader to an iPad via USB-C: Easier Than Expected URL: https://rob-hoeijmakers.ghost.io/connect-kobo-reader-to-an-ipad-via-usb-c/ Last updated: 2026-05-10T08:54:06.000Z When I set out to connect my **Kobo eReader** to an **Apple device with** [**USB-C**](https://rob-hoeijmakers.ghost.io/usb-c/), I anticipated a few complications, especially regarding power requirements. Online discussions and past experiences suggested that many accessories, when connected to an iPad or iPhone, demand additional power to function correctly. ## The Setup: USB-C and Power Concerns To prepare for potential power issues, I purchased: 1. **A USB-C to Micro USB cable** (to connect the Kobo eReader) 2. **A USB-C splitter** (to provide additional power if needed) The idea was straightforward: if my iPad or iPhone complained about the Kobo drawing too much power, the splitter would allow me to inject external power into the connection. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/03/D6571AFF-499C-46D4-AA0D-72E94D5CD157_1_105_c.jpeg) [The cable with USB-C and USB Micro](https://www.cablematters.com/pc-714-123-braided-usb-c-to-micro-usb-20-cable.aspx) ## The Surprise: No Additional Hardware Needed I first tested the [**USB-C to Micro USB cable**](https://www.cablematters.com/pc-714-123-braided-usb-c-to-micro-usb-20-cable.aspx) directly on my **iPad**. Surprisingly, it worked **immediately**—no external power source was required. The eReader mounted successfully, allowing file transfers without any issues. This was unexpected, as Apple devices often display a **"requires too much power"** error when connecting accessories. The Kobo eReader, however, seems to draw just enough power to remain within the acceptable limits. I recorded a **short video** showing how seamless the connection was: Demo of attaching and opening the drive on iPad ## Next Steps: Testing on an iPhone I haven’t yet tested this setup on a **USB-C iPhone**, as I don’t have one at hand. However, I plan to try it out with someone else’s device to see if the results are the same. If it works just as effortlessly, it would confirm that **Kobo eReaders are plug-and-play with Apple’s USB-C ecosystem**—a great convenience for users who want to transfer books without hassle. ## Key Takeaway If you’re looking to connect a **Kobo Reader to an iPad via USB-C**, **all you need is the right USB-C to Micro USB cable**. No extra hardware, no workarounds, just plug in and go. Once I complete the iPhone test, I’ll update this post with the final verdict. But for now, it’s safe to say that **this setup works surprisingly well right out of the box.** ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/03/956E1E77-349C-46B7-85D7-E188B79D5ADA_1_105_c.jpeg) You see the KOBOeReader as a drive here. ### Further reading - [Connecting My Kobo Directly to an iPhone 16 Pro](https://rob-hoeijmakers.ghost.io/connecting-my-kobo-to-iphone/) - [USB-C: A Universal Plug with Many Faces](https://rob-hoeijmakers.ghost.io/usb-c/) - [Braided USB-C to Micro USB 2.0 Cable](https://www.cablematters.com/pc-714-123-braided-usb-c-to-micro-usb-20-cable.aspx) ### Why Automation with Make.com Made Me So Happy URL: https://rob-hoeijmakers.ghost.io/make-dot-com-automation/ Last updated: 2026-05-12T07:23:41.000Z Lately, I’ve been experimenting with automation using [Make.com](https://www.make.com), and it has been a game-changer. Compared to the expensive and often rigid Zapier, Make.com feels more like a **creative workspace**—visual, flexible, and far more engaging. I’ve used [**Apple Shortcuts and Apple Automator**](https://rob-hoeijmakers.ghost.io/llm-driven-automation-and-anthropics-computer-use/) extensively, and while they’re great for local and device-based automation, **Make.com is perfect for cloud-based workflows**. It allows you to connect a vast range of online services, from **CMS platforms like** [**WordPress**](https://rob-hoeijmakers.ghost.io/how-about-wordpress/) **and** [**Ghost**](https://rob-hoeijmakers.ghost.io/ghost-blog/) to translation services, email marketing tools, and beyond. ## From 40 Minutes to a Few Clicks One of my recent projects involved **automating content syncing between Ghost and WordPress**. I used to do this manually—copying and pasting text, handling images separately, and relying on Chrome extensions to speed things up. It took me **40 minutes per article**. Now, Make.com handles it in seconds. I designed the workflow with the **manual process in mind**, identifying repetitive steps and automating everything possible. There are still minor tweaks I handle manually—like assigning categories and tags—but those are quick and simple. ## Key Insight: Start with Images, Not Text One thing I learned while building this automation is that **writing an article manually starts with text, but programmatic workflows must start with images**. Why? Because **the text needs image URLs**. In Make.com, this means fetching and processing images first, then inserting them into the text before publishing. A small but essential shift in mindset when working with automation. ## AI + Make.com = Next-Level Workflows Another exciting element is the ability to **integrate ChatGPT via an API**. This opens up a whole new world of possibilities, allowing me to **prototype intelligent, agentic workflows**—not just moving data around, but processing, refining, and even generating content dynamically. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/03/image.png) OpenAI's ChatGPT as part of an automation in Make.com [Make.com](https://www.make.com/en/register?pc=robhoeij) isn’t just a tool—it’s a **playground for automation**, bridging cloud apps, AI, and efficient workflows. ### The Digital Euro: My First Exploration URL: https://rob-hoeijmakers.ghost.io/digital-euro-cbdc/ Last updated: 2026-05-10T08:54:06.000Z ## Introduction: A Currency Shift with Controversy The digital euro is stirring debate. Unlike cash, which offers anonymity, this new form of money is fully traceable. While this could reduce illicit activities, it also hands significant control to the state. This article serves as a **first exploration** of this phenomenon. I aim to outline the key aspects, examine different perspectives, and start forming a well-rounded understanding. Rather than taking a firm stance, this piece is about gathering information and inviting discussion. I’m still forming my own opinion, and I’d love to hear yours. ## What Is the Digital Euro? As Europe moves towards a more digitized financial system, the digital euro emerges as a key development. But what exactly is it, and how does it differ from other forms of digital money? - It’s a **central bank digital currency (CBDC)** issued by the European Central Bank (ECB). - Unlike cryptocurrencies, it’s **not decentralized**—all transactions are logged by the central authority. - It differs from regular digital money (bank accounts, Apple Pay, Google Pay), which is managed by private banks. Essentially, it’s **public money in a digital form**, replacing cash but with new rules and oversight. Unlike traditional digital payments, it allows direct transactions without the need for private banks, giving the central bank full control over issuance and transfers. ## The Case for the Digital Euro Proponents argue that: - 👉 **Stability:** A state-backed currency, unlike crypto, which is volatile. - 👉 **Security:** Enhances protection against fraud and counterfeiting. - 👉 **Financial Inclusion:** Provides a cash alternative in an increasingly digital world. ## The Case Against the Digital Euro Critics worry about: - 💡 **Loss of Privacy:** Transactions can be monitored, limiting personal freedoms. - 💡 **Government Control:** In extreme cases, funds could be frozen or restricted. - 💡 **The Transition from Cash:** Even if optional at first, physical cash could be phased out over time. History shows that policies can shift, and what starts as an optional feature might later become mandatory. ## Where Is the EU Headed? The European Central Bank is in the exploration phase, but other countries have already moved forward. [Central Bank Digital Currency TrackerA CBDC is virtual money created by a central bank. As cryptocurrencies and stablecoins become popular, central banks provide alternatives![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon-300x300.png)Atlantic Council![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/trackerthumbnail.png)](https://www.atlanticcouncil.org/cbdctracker/) China, for example, has launched the **digital yuan** with a clear government-controlled structure. The EU model may differ, but it’s worth tracking global trends. [CBDC: what, why and howA central bank digital currency (CBDC) is a virtual banknote as it were. But what exactly is it? And what role does DNB play? Read more![](https://static.ghost.org/v5.0.0/images/link-icon.svg)Logo DNB![](https://www.dnb.nl/media/obslqutz/dnb-logo-desktop.svg)](https://www.dnb.nl/en/innovations-in-payments-and-banking/digital-euro-what-why-and-how/) ## Final Thoughts: An Open Debate This article is only the **beginning of an exploration** into the digital euro. There are many angles to consider, and this is just a first step in understanding the implications of this shift in money as we know it. Beyond the digital euro, I am also examining the broader European digital future, including [the identity wallet and the eID initiative](https://rob-hoeijmakers.ghost.io/digital-identity-id-empowerment-or-state-control/) by the [European Union](https://rob-hoeijmakers.ghost.io/tag/europe/). These developments together shape a significant transformation in how digital identity and financial systems interact. ### Further reading - [Wero: Europe’s new payment brand](https://rob-hoeijmakers.ghost.io/wero-europes-new-payment-brand/) ### Identity Wallet: A Tool for Empowerment or a Mechanism of Control? URL: https://rob-hoeijmakers.ghost.io/digital-identity-id-empowerment-or-state-control/ Last updated: 2025-04-10T15:13:10.000Z ## Introduction: Piecing Together the Puzzle Digital identity is more than just technology—it’s about **trust, power, and control**. As I started researching digital identification, I realized how much trust plays a role in how we perceive it. At a **national level**, systems like DigiD or BankID are familiar. We understand them. But when the discussion moves to an **EU-wide digital ID (EUDI)**, something changes. The EU feels distant, bureaucratic, and harder to influence. Why do we accept national ID systems more easily? What makes an EU-wide digital ID different? And how do identification systems shift from being **a tool for convenience to a mechanism of control**? I don’t have all the answers, but let’s lay down some puzzle pieces together. ⚠️ This is me exploring the topic of digital identity and identification in the EU. It is a far reaching topic and a complex puzzle, there might be mistakes or errors, just let me know, and I will correct. ## 1\. Identity vs. Identification: A Crucial Distinction One of the first things I needed to untangle was the difference between **identity** and **identification**: - **Identity** is who you are—your name, age, profession, beliefs. It’s something personal, shaped by culture, experience, and context. - **Identification** is proving who you are. It requires an external system—a passport, ID card, login credentials—to validate your identity. 📌 **Key question:** Who controls identification? Governments? Private corporations? Or should individuals have more say in how their identity is verified? 💡 ****The Hidden Load of Identity in Centralized Systems** Many people think of identity as just a name, address, or date of birth. But in centralized systems, identity is far more than that—it becomes an extensive digital profile that carries significant privacy implications. A centralized system doesn’t just store who you are but also how you behave: ✔️ ****Personal details** – Name, age, nationality, and official documents ✔️ ****Financial footprint** – Bank accounts, transactions, tax records ✔️ ****Communication patterns** – Call logs, messages, email metadata ✔️ ****Health data** – Medical history, prescriptions, fitness tracking ✔️ ****Social connections** – Contacts, friendships, work relationships ✔️ ****Digital behaviour** – Search history, purchases, app usage, location The more a system knows, the more ****control** it has over verification, tracking, and access. Identity becomes ****not just a key to services, but a tool for surveillance, profiling, and even manipulation**. ## 2\. The Trust Factor: Why National Digital IDs Work (For Now) Most European countries already have **national digital identity systems**, and they work fairly well. Examples include: - **Netherlands:** [DigiD](https://www.digid.nl/en) (used for taxes, healthcare, and official documents) - **Belgium:** [eID & Itsme](https://eid.belgium.be/en) (widely accepted in banking and public services) - **Sweden:** [BankID](https://www.bankid.com/en/) (a private-sector-driven ID used for everything from banking to e-commerce) - **Estonia:** [e-Residency & ID-Card](https://www.e-resident.gov.ee) (an advanced system allowing digital signatures, online voting, and business registration) I use DigiD myself, and while I don’t love it, I **trust it more than a system run by the EU**. Why? Because: - I **know who’s in charge**—my national government, not a distant bureaucracy. - There’s some level of **accountability**—even if flawed, my vote impacts national policies. - It’s still **mostly voluntary**, though increasingly necessary for services. 📌 **Key point:** Familiarity breeds trust. We trust national systems **because they are closer to us**. [Understanding Digital Identity in the EU: eID, eIDAS, and the EUDI WalletThis article serves as a reference point for further discussions on digital identity, authentication, and legitimisation in the EU. Future articles will explore how businesses, organisations, and individuals can leverage these systems for more secure and streamlined online interactions.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-22.png)Rob Hoeijmakers (robhoeij)Rob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_0771-1-1.jpeg)](https://rob-hoeijmakers.ghost.io/eu-eid-eidas-eudi-wallet/) ## 3\. The EU-Wide Digital ID: A Different Beast? The[ EU’s **eIDAS 2.0** framework](https://digital-strategy.ec.europa.eu/en/policies/eidas-regulation) aims to create a [**European Digital Identity Wallet**](https://commission.europa.eu/strategy-and-policy/priorities-2019-2024/europe-fit-digital-age/european-digital-identity%5Fen), allowing citizens to store credentials (passport, driver’s license, medical records) and use them across borders. While this sounds convenient, the **trust gap** between national governments and the EU raises concerns: - **The EU feels distant** – Policies are made at the **Commission level**, where citizens have little direct influence. - **Data centralization** – A pan-European system could create a **single point of control** over identity verification. - **Function creep** – What starts as a **convenience** could quickly become an **obligation**. 📌 **Key question:** Will an EU-wide ID system be an **empowering tool** or a **mandatory requirement**? 🇪🇺 The ****digital euro** and the ****European Digital Identity (EUDI)** are converging as part of Europe’s broader digital transformation, raising critical questions about ****privacy, financial autonomy, and state control**. While a [central bank digital currency (CBDC)](https://rob-hoeijmakers.ghost.io/digital-euro-cbdc/) promises ****secure, cash-like transactions**, its integration with digital identity wallets could streamline authentication and payments—yet also enable ****transaction monitoring and financial restrictions**. This intersection of ****money and identity** is a pivotal shift that demands scrutiny. In a future piece, I’ll explore whether these systems truly ****empower citizens or tighten state oversight**, and what safeguards are needed to protect privacy and autonomy in this evolving landscape. ## 4\. From Voluntary to Mandatory: The Gradual Shift We’ve seen this before. During **COVID**, digital passes went from **optional travel tools** to requirements for **accessing everyday life**. Could the same thing happen with an EU Digital ID? A possible trajectory: - First, it’s **only for government services**. - Then, it’s required for **banking, e-commerce, and social media**. - Eventually, there’s **no alternative**—you must use it for identification everywhere. 📌 **Key concern:** The issue isn’t whether digital ID is useful. It’s **who decides when and how it is used**. > WHY DECENTRALIZED BLOCKCHAIN IS THE FUTURE OF DIGITAL IDENTITY > > The UK is accelerating plans for a digital identity system, with PM Keir Starmer's government introducing: > > \- A new Office for Digital Identities. > > \- Digital ID cards for veterans and digital driver’s licenses by… [pic.twitter.com/4s1FSKNtBj](https://t.co/4s1FSKNtBj) > > — Mario Nawfal (@MarioNawfal) [January 5, 2025](https://twitter.com/MarioNawfal/status/1875859433365860424?ref%5Fsrc=twsrc%5Etfw) ## 5\. What Would a Trustworthy Digital ID Look Like? For a European digital ID to gain trust, it must: - **Give users control** – Individuals must decide who accesses their data. - **Avoid centralization** – No single government or corporation should control identity verification. - **Guarantee alternatives** – No mandatory use, no loss of choice. - **Be transparent** – With clear oversight and accountability. 📌 **Final thought:** Digital ID should be **a tool, not a requirement**. Trust must be **earned, not imposed**. ## Conclusion: The Puzzle Remains Unfinished This topic is too big to fit neatly into one article. There are **technical, political, and ethical layers** to unravel. I set out to understand the trust gap between national and EU identification systems, and I’m left with even more questions. What do you think? Is an EU-wide digital ID a **necessary step for the future**, or a **dangerous move toward centralized control**? The discussion is far from over. ### Transcribing Speech on a Mac and iOS: My Workflow and Tools URL: https://rob-hoeijmakers.ghost.io/transcribe-audio-on-macos-and-ios/ Last updated: 2026-05-26T19:35:47.000Z If you’re looking for an efficient way to transcribe speech on a Mac and iOS, you might have explored Apple’s built-in options. While Apple Intelligence offers basic transcription in Notes and other apps, its language support is still limited. For a more reliable, multilingual solution, I use **Transcribe**. ## My Transcription Workflow Transcribing audio across devices should be simple, efficient, and accurate—here’s how I make it work seamlessly. 1. **Recording Audio** – I typically capture voice notes using Apple’s **Voice Memos** app or another recording tool. 2. **Sending to Transcribe** – For fast and accurate transcription across multiple languages, I send the audio to **Transcribe** (available on macOS and iOS). It has excellent audio perception and a cloud-based function, making it highly reliable. 3. **Review and Edit** – Once transcribed, I review and make minor edits directly within **Transcribe**, ensuring clarity and accuracy. 0:00 /1:02 1× Workflow demo of transcribing audio: VoiceMemo > Transcribe > ChatGPT ## Why Use Transcribe? Transcribe is a dedicated transcription tool with the following features: - **Multi-Language Support** – Accurately transcribes speech in various languages. - **Cloud-Based Processing** – Ensures fast and efficient transcription with minimal device load. - **Seamless Integration Across Devices** – Works effortlessly between Mac and iOS. - **Intelligent Formatting** – Auto-punctuation and speaker recognition enhance readability. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/02/Transcribe-video.jpg) One more cool feature, Transcribe also transcribes video and let's you export as SRT for subtitles. ## Transcribe Pricing Overview Transcribe offers a simple pricing model that makes it accessible for regular users and professionals. You can explore their plans [here](https://www.transcribe.wreally.com/). For those who regularly transcribe audio, **Transcribe** provides a seamless and efficient experience across Mac and iOS. The multi-device support ensures convenience for users who switch between desktop and mobile. [TranscribeTranscribe App and Online Editor. Your personal assistant for note taking and transcribing. Our voice transcription service saves you time and helps you focus on what’s important.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/apple-touch-icon-3.png)TranscribeRoutes Software SRL![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/messengers_placeholder_1200_630.png)](https://transcribe.com) What’s your preferred transcription setup? Let me know in the comments! ### Why Can't We Transcribe Phone Calls in the EU? URL: https://rob-hoeijmakers.ghost.io/transcribe-phone-calls/ Last updated: 2026-05-10T12:45:59.000Z ## A Simple Wish: Transcribing Calls Apple is rolling out call transcription in iOS 18, but not in the [EU](https://rob-hoeijmakers.ghost.io/tag/europe/). The US gets it, but here, we're stuck in the past. The same applies to WhatsApp calls and voice messages—no transcription available. This isn't just a minor inconvenience. Transcription is an essential tool for productivity, allowing people to revisit conversations, search for key points, and work more efficiently. I often prefer calling over video calls—it saves bandwidth and is less distracting—but without transcription, phone calls feel like an outdated communication method. ## The Privacy Roadblock The EU’s strict privacy regulations, including [GDPR](https://rob-hoeijmakers.ghost.io/gdpr/) and the ePrivacy Directive, complicate call transcription. The requirement for explicit consent makes implementation tricky, though video conferencing apps like Zoom and Microsoft Teams have figured it out. If transcribing video meetings is allowed, why not phone calls? [Fireflies: Meeting Transcriptions & AutomationFireflies.ai changed how I handle meetings, capturing every conversation with ease. Its AI-driven transcription is powerful, but some features need fine-tuning.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-42.png)Rob Hoeijmakers (robhoeij)Rob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_1385-3.jpeg)](https://rob-hoeijmakers.ghost.io/fireflies-ai-transcriptions/) ## Apple and Android: A Fragmented Landscape Apple emphasizes privacy but still offers call transcription in other regions. Why not here? Google, meanwhile, has cracked down on call recording apps, making it harder to find a reliable, [GDPR-compliant](https://rob-hoeijmakers.ghost.io/gdpr/) transcription tool for Android. WhatsApp, the dominant calling app in the EU, also doesn’t offer transcription. While iOS transcribes voice messages, it’s unclear if that feature will roll out in the EU. The result? We’re left manually taking notes during calls—hardly an efficient solution. ## Workarounds Aren’t Enough Sure, you can put your phone on speaker and record with another device, but that’s clunky and unreliable. Manual note-taking is time-consuming and prone to errors. Meanwhile, [LLMs have made transcription more powerful than ever](https://rob-hoeijmakers.ghost.io/a-comparative-analysis-of-top-language-models/), with the ability to generate summaries and extract action points—if only we had access to them. 💡 When I am in the office I might sometimes record a call with Apple Voice Memos by putting the phone on speaker. Then I use [Transcribe](https://rob-hoeijmakers.ghost.io/transcribe-audio-on-macos-and-ios/) and [ChatGPT](https://rob-hoeijmakers.ghost.io/try-out-chatgpt-yourself/) for analysis. ## A Call for Change Tech companies like Apple, Google, and Meta need to prioritize GDPR-compliant transcription. Solutions should: - Focus on transcription rather than recording. - Use LLMs for summarization and action item extraction. - Implement simple, clear consent mechanisms. - Be seamlessly integrated into phones and communication apps. The EU has an opportunity to balance privacy with innovation. We need to push for better transcription options—whether for mobile calls, VoIP calls, or voice messages. ⚠️ ****WIP:** I am exploring Dutch (EU) VoIP services that offer ****call recording and transcription** while ensuring full compliance with ****privacy regulations (GDPR, ePrivacy)** and ****telecom laws**. While my article suggests that Apple's restrictions on call transcription in the EU stem from regulatory limitations, it’s possible that Apple itself is imposing these barriers, rather than legal constraints alone. By looking into VoIP alternatives, I aim to find a solution that works within both ****legal and technical** boundaries. Will keep you posted. ### Apple’s New ‘Invites’ App: A First Look URL: https://rob-hoeijmakers.ghost.io/apples-new-invites-app-a-first-look/ Last updated: 2026-04-25T13:16:35.000Z So, Apple just dropped a new app called *Apple Invites*—which, funnily enough, was rumoured to be called *Confetti* before launch. Naturally, as someone who has always been *interested* (read: notoriously bad) at [calendaring and event planning](https://rob-hoeijmakers.ghost.io/digital-diary-and-calendars/), I had to check it out. The premise is simple: it’s an iOS app with cloud support that lets you set up and send event invites. But from the overall feel of the app, it’s clearly aimed at the *personal* space rather than professional or business use. Now, if you’ve followed discussions around calendars and event planning, you’ll know this is still an *open* market—fragmented by ecosystems. Google, Apple, Meta, Microsoft… they all have their own systems, and if you live outside Apple’s walled garden, it’s not clear how well this will work for you. Since I’m already inside the ecosystem, I can’t tell how seamless (or frustrating) it will be for others. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/02/image.png) Invites App Icon ## First Impressions? Honestly—neutral. There’s nothing groundbreaking here. It’s a nice, polished version of a regular calendar invite, with some extra flair: ✅ You can add an image ✅ There’s playlist integration ✅ The map function is fine ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/02/Apple-Invites.jpg) Apple Invites ## What’s Missing? I was really hoping this would replace those messy *"let's-find-a-date-that-works-for-everyone"* services (*Doodle*, etc.). But nope—it’s still based on *me* setting up an event and sending invites, rather than a more collaborative scheduling tool. So if you were hoping for Apple to fix the eternal group scheduling nightmare, this isn't it. [Introducing Apple Invites, a new app that brings people togetherApple today introduced Apple Invites, a new app for iPhone that allows users to create custom invitations to gather friends and family.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon-5.ico)Apple Newsroom![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/Apple-Invites-upcoming-lp.jpg.og-1.jpg)](https://www.apple.com/newsroom/2025/02/introducing-apple-invites-a-new-app-that-brings-people-together/) ## How Does It Compare to WhatsApp's Voting Module? WhatsApp already offers an invite system within group chats that includes a built-in voting feature, discussion thread, and the ability to seamlessly move from decision-making to an invite. This makes the process much more interactive, allowing participants to collectively decide on a time and location before finalising the event. In contrast, *Apple Invites* feels more like a one-way street—the organiser sets the time and details, and others simply accept or decline. If you're looking for a truly collaborative planning tool, [WhatsApp](https://rob-hoeijmakers.ghost.io/whatsapps-global-reach/) (or external tools like [Doodle](https://doodle.com/en/)) might still be the better option. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/02/Whatsapp-voting.png) WhatsApp voting I’m still curious where Apple takes this, though. Maybe tighter iMessage and Calendar integration? A real-time scheduling feature? Something social? For now, it’s *fine*—but nothing that’s making me rethink how I plan events. If you’ve tried it, I’d love to hear your thoughts. Is this a useful addition or just another app that’ll collect dust in a folder? Drop a comment! ### Understanding Digital Identity in the EU: eID, eIDAS, and the EUDI Wallet URL: https://rob-hoeijmakers.ghost.io/eu-eid-eidas-eudi-wallet/ Last updated: 2026-05-10T08:54:06.000Z Digital identity is becoming increasingly important in both personal and professional contexts. As the European Union strengthens its framework for **secure digital identification**, three key terms often come up: **eID**, **eIDAS**, and the **EUDI Wallet**. This article serves as a foundational reference for understanding their roles, relationships, and impact on digital identity, authentication, and legitimisation. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2026/04/eu_identity_diagram.png) The EU's digital identity system has three layers: a legal framework (eIDAS), a national identity key (eID), and the wallet that becomes yours once activated (EUDI). ## 1\. eID (Electronic Identity) **eID** stands for **Electronic Identity**, referring to a digital means of identifying oneself online. It is a **national** electronic identification system issued by an EU member state and used to authenticate individuals and businesses. ### Key Characteristics: Each EU country has its own **eID system**, such as: - **DigiD** (Netherlands) - **Itsme** (Belgium) - **eID** (Germany) - Used for **secure login** to government, banking, healthcare, and private sector services. - Some national eID systems are **notified under eIDAS**, meaning they are recognised across the EU for cross-border authentication. ### Role in Digital Identity: - Enables individuals to **prove who they are online**. - Reduces the need for multiple accounts and passwords. - Forms the basis for more advanced digital identity solutions. ## 2\. eIDAS (Electronic Identification, Authentication and Trust Services) **eIDAS** is the **EU regulation (Regulation 910/2014)** that provides a **legal framework** for digital identity and trust services across the European Union. ### Key Characteristics: - Ensures **cross-border recognition** of electronic identities. - Defines standards for **electronic signatures, digital seals, timestamps, and website authentication**. - Regulates **qualified trust service providers (QTSPs)** that issue digital certificates. - Enables the legal use of digital signatures for contracts, government documents, and other transactions. ### Role in Digital Identity: - Establishes the **legal foundation** for secure online transactions. - Enables **businesses and individuals** to use their national eID for services across the EU. - Creates **trust and security** in digital interactions. [eIDAS RegulationThe eIDAS regulation facilitates secure cross-border transactions by establishing a framework for digital identity and authentication. It aims to create confidence in electronic interactions and promote seamless digital services in the EU.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon.svg)Shaping Europe’s digital future![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/logo-ec--mute.svg)](https://digital-strategy.ec.europa.eu/en/policies/eidas-regulation) ## European Digital Identity Wallet (EUDI) The [EUDI](https://ec.europa.eu/digital-building-blocks/sites/display/EUDIGITALIDENTITYWALLET/EU+Digital+Identity+Wallet+Home) is the EU's framework for digital identity, and the [EUDI Wallet](https://ec.europa.eu/digital-building-blocks/sites/display/EUDIGITALIDENTITYWALLET/EU+Digital+Identity+Wallet+Home) is one of the key tools that implements this vision (introduced under **eIDAS 2.0**). ### Key Characteristics: - Functions as a **digital identity wallet** for EU citizens and businesses. - Allows users to store and share **verified credentials** (e.g., eID, driver’s license, medical records, diplomas). - Based on **self-sovereign identity (SSI)** principles, giving users **control over their data**. - Can be used for **both governmental and private sector services**. ### Role in Digital Identity: - Aims to create a **unified digital identity system** across the EU. - Provides a **secure and interoperable** way to verify identity online and offline. - Reduces reliance on third-party authentication services (e.g., social logins from Google or Facebook). [EU Digital Identity Wallet Home - EU Digital Identity Wallet -![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon-6.ico)EU Digital Identity Wallet -Constantin-Catalin MIRICA![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/illu-the-wallet-sec-2.svg)](https://ec.europa.eu/digital-building-blocks/sites/display/EUDIGITALIDENTITYWALLET/EU+Digital+Identity+Wallet+Home) ## The Connection Between eID, eIDAS, and EUDI Wallet 1. **eID** is a **national** digital identity used for authentication. 2. **eIDAS** provides the **legal framework** ensuring interoperability and trust for eID and digital transactions across Europe. 3. **EUDI Wallet** is an **EU-wide solution** under **eIDAS 2.0**, designed to integrate eID and other credentials into a **secure, digital-first system**. Together, these three components aim to create a **cohesive and secure digital identity ecosystem** in Europe, with the goal of enabling individuals and businesses to verify their identities efficiently and safely in an increasingly digital world. --- **Update (April 2026)** This article was referenced by Mariana Meneses in *The Quantum Record*: “[Nations adopt digital IDs for citizens while critics highlight privacy issues](https://thequantumrecord.com/philosophy-of-technology/nations-adopt-digital-ids-for-citizens-while-critics-highlight-privacy-issues/)”. The article places digital identity in a broader political and infrastructural context, highlighting both its role in access to services and the tensions around privacy and control. ### Further reading - [Passports, Phones, and the Future of Identity](https://rob-hoeijmakers.ghost.io/passports-phones-and-the-future-of-identity/) - [Identity Wallet: A Tool for Empowerment or a Mechanism of Control?](https://rob-hoeijmakers.ghost.io/digital-identity-id-empowerment-or-state-control/) ### OS-Level Translation: A Multilingual Productivity Hack URL: https://rob-hoeijmakers.ghost.io/os-level-translation-a-multilingual-productivity-hack/ Last updated: 2026-05-10T08:54:07.000Z Apple’s built-in translation tools in iOS and macOS make switching between languages effortless. Integrated across apps, they streamline communication, enhance privacy, and reduce reliance on third-party services. ## Effortless Translation Across Apple Devices Apple’s translation feature allows users to convert text instantly in Messages, Mail, Safari, and Notes. I use it daily to translate Dutch messages into English before posting online or responding to international contacts. Instead of copy-pasting into Google Translate or DeepL, I simply select text, tap “Translate,” and refine the result in seconds. 0:00 /0:15 1× The translate workflow on iOS, iPhone ## Fast, Private, and Seamless System-wide translation keeps workflows smooth and private. Unlike cloud-based services, Apple’s on-device processing ensures sensitive content remains secure. The ability to translate within any app eliminates unnecessary app-switching, making it a fast and efficient tool. 0:00 /0:09 1× The translate workflow on MacOS ## Limitations and Potential Improvements While great for short messages, Apple’s translations struggle with longer texts, often losing nuance and coherence. It also lacks the accuracy of DeepL or Google Translate for complex phrases. For anything beyond a few sentences, I double-check with other tools. Despite its limitations, Apple’s built-in translation is an invaluable feature for multilingual users. It’s fast, private, and seamlessly integrated across Apple devices. With further improvements, it could become even more powerful. 🚫 Not all languages are supported. Apple's Translate app supports translation between Arabic, Chinese (Simplified and Traditional), Dutch, English (UK and US), French, German, Hindi, Indonesian, Italian, Japanese, Korean, Polish, Portuguese (Brazilian), Russian, Spanish (Spain), Thai, Turkish, Ukrainian, and Vietnamese. All these languages support dictation and can be downloaded for offline use. --- ### Further reading - [OS-Level Translations: Enhancing Writing Across Devices](https://rob-hoeijmakers.ghost.io/os-level-translations/) ### Google Gemini's Video AI: What's New and Why It Matters URL: https://rob-hoeijmakers.ghost.io/google-gemini-video-ai/ Last updated: 2026-05-10T08:54:07.000Z Last night, I attended an AI event ([AI Salon Amsterdam](https://aisalon.amsterdam/)) and had an interesting conversation with someone working in **industrial AI video applications**. Naturally, I brought up **Google Gemini’s video analysis**—the speed, the ability to handle both **recorded and real-time video**, and its seamless integration with a powerful language model for **summarization and reasoning**. His reaction? **Unimpressed.** Mine? **Amazed.** That contrast made me reflect: **why does Gemini’s video analysis feel like such a big deal to me, but not to someone deeply embedded in industrial AI?** ## **Industrial AI vs. Google Gemini: A Shift in Perspective** It turns out the difference isn't just about **technology**—it’s about **who it’s for** and **what it enables**. - **Industrial AI video systems** have existed for years. Security cameras can already detect motion, identify faces, and trigger alerts. In manufacturing, AI analyzes production lines for defects. These systems are powerful, but they are **narrowly focused and purpose-built**. - **Google Gemini, on the other hand, is shifting AI video analysis into consumer and general-purpose computing.** It’s not just about making surveillance better—it’s about integrating **video intelligence into everyday tools**, **enhancing computing workflows**, and **making video searchable, actionable, and interactive**. This **shift to new areas** is what excites me. Gemini’s video analysis isn’t just a feature—it’s a **gateway to new workflows and product ideas**. ## **Where This Could Lead: Efficiency + New Growth Areas** ### **1\. From Passive to Active Video Analysis** Right now, most video tools require **manual effort**—you scrub through footage, try to find key moments, and make sense of what’s happening. AI-driven video changes this: - **Wildlife cameras**: Instead of scanning hours of empty footage, **ask AI to find when an animal appears**. - **Security cameras**: Go beyond motion alerts—**"Show me when a person approached the door, but ignore the cat."** - **Baby monitors & pet cameras**: AI could detect **specific sounds, behaviors, or even emotions**. ### **2\. AI That Watches Your Screen and Helps** I realized that **screencasting is also video**—and that opens another set of possibilities: - **AI-assisted tutorials**: Instead of watching an entire how-to video, **ask AI to summarize just the relevant parts**. - **Smart recording**: AI could **auto-capture** only the important moments from meetings, coding sessions, or creative work. - **On-screen assistance**: If Gemini can analyze video **in real-time**, why not let it watch your screen and suggest actions? (e.g., “It looks like you're editing a document—do you need formatting help?”) ### **3\. Beyond Efficiency: New Products & Possibilities** When AI video analysis is **fast, integrated, and intelligent**, it’s not just about saving time. **It unlocks new products:** - **AI-driven highlight reels** (for sports, events, or personal videos). - **AI-assisted coaching** (for workouts, public speaking, or even music practice). - **Elderly care & accessibility** (detecting falls, tracking movements, offering real-time assistance). This **isn’t just about improving what we already do—it’s about thinking differently about what’s possible**. ## **Why This Feels Like a Breakthrough** So why does Google Gemini’s video analysis impress me, while an industrial AI expert sees nothing new? Because **it’s not just about better video AI—it’s about bringing it to new domains, integrating it with powerful reasoning (LLMs), and opening the door to workflows that didn’t exist before**. **It’s the shift from niche applications to general-purpose intelligence.** That’s what makes it exciting. And the more I think about it, the more I see that we’re only scratching the surface of what AI video analysis can do. --- ### Further reading - [Google AI Studio](https://aistudio.google.com/welcome) ### Reflections on a Deep Tech AI Event URL: https://rob-hoeijmakers.ghost.io/reflections-on-a-deep-tech-ai-event/ Last updated: 2026-05-10T08:54:07.000Z The Deep Tech AI Event brought together key organisations like "[Thematic Technology Transfer Artificial Intelligence](https://ttt-ai.nl)," "[Amsterdam AI](https://www.amsterdamai.com)," "[ROM InWest](https://rominwest.nl)," and "[Startup Amsterdam](https://www.iamsterdam.com/en/business/startupamsterdam/about-startupamsterdam)." These groups highlight the collaborative nature of the event, which bridged academic knowledge, industrial innovation, and governmental support. ### Key Learnings As someone relatively new to the world of Deep Tech and AI, I’ve been diving into this rapidly evolving space with a combination of curiosity and excitement. With my background as a digital strategist, I’m looking to invest not just financially, but also with my time, knowledge, and energy. The potential I see in AI is vast, particularly in the areas of Education, Health, and Defense, where its capabilities could bring about profound change. Attending this event was an eye-opening experience, providing invaluable insights into the latest advancements and how they’re shaping industries. Below are the key learnings that stood out to me. 1. **Early-Stage and Scale-Up Support Matters**: Regional development organisations play a pivotal role in helping startups grow by providing financial and infrastructural support. 2. **Balancing IP and Collaboration**: The panel discussions made it clear that finding the right balance between protecting proprietary technology and fostering collaboration is critical for success in deep tech. 3. **Open Source as an Innovation Catalyst**: Open source can act as a powerful enabler, but businesses need to understand (early) its obligations to leverage it effectively. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/01/IMG_0667-1.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/01/IMG_0669.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/01/IMG_0668.jpeg) [Jorn Eiting van Liempt ](https://www.linkedin.com/in/jorneiting/)from [ROM inWest](https://rominwest.nl) gave a great keynote about Funding rounds. ### Inspiring Meetings Two conversations with startups in an Italian table setup (a relaxed, small-group discussion format) stood out during the event: - [**Readler.ai**](https://www.readler.ai): This innovative startup is using advanced speech support algorithms to improve voice interaction systems. Their approach combines the flexibility of open-source tools with proprietary enhancements to deliver superior user experiences. Their vision aligned closely with my interests, and I’m excited to explore potential collaboration. - [**Flux Robotics**](https://www.fluxrobotics.nl/#product): Specialising in AI-driven magnetic navigation for surgery, Flux Robotics exemplifies how deep tech can tackle real-world challenges. By automating surgical tool positioning, they are setting new standards in healthcare precision and efficiency. The event left me energised and inspired by the possibilities of deep tech AI. The mix of structured discussions and informal networking underscored the importance of collaboration in this rapidly evolving field. Whether it’s through time, resources, or (in-kind) partnerships, events like these remind us that innovation thrives when we work together. --- ### Further reading - [Deep Tech AI and Open Source licenses](https://rob-hoeijmakers.ghost.io/deep-tech-ai-open-source-licenses/) 💡 ****Deep Tech** refers to innovations based on advanced scientific or engineering breakthroughs, often requiring long-term development and significant investment. It includes fields like quantum computing, biotechnology, and nanotechnology. ****Deep Tech AI** involves AI technologies that use advanced methods to solve complex problems, often relying on cutting-edge research, powerful computing, and interdisciplinary approaches. Examples include autonomous systems and AI-driven drug discovery. ### Deep Tech AI and Open Source licenses URL: https://rob-hoeijmakers.ghost.io/deep-tech-ai-open-source-licenses/ Last updated: 2026-05-10T08:54:07.000Z Last week, I attended the [Deep Tech Event](https://rob-hoeijmakers.ghost.io/reflections-on-a-deep-tech-ai-event/) in [Amsterdam](https://www.amsterdamsciencepark.nl/locate/opportunities/startup-village/). The energy, ambition, and collaboration on display were truly inspiring. The event brought together innovators, investors, and thought leaders to discuss the future of deep tech, with a specific focus on **"Investments & Intellectual Property in Deep Tech AI."** This panel discussion shed light on the nuances of intellectual property (IP) and open-source software—topics I hadn’t fully appreciated before. ## Panel Discussion: Investing & IP in Deep Tech AI One of the most striking points raised during the panel was that some open-source licences, particularly those in the copyleft category, require businesses to open-source their own software if they use or modify the licensed software as part of their product. This principle ensures that any improvements made by companies are shared back with the community, fostering collaboration. However, it raises important questions: How can businesses adopt open-source tools while protecting their proprietary innovations? And what strategies can they use to navigate these licensing complexities effectively? ## The Two Faces of Open Source Open source was at the heart of the discussion, particularly its dual nature: permissive and copyleft licences. ### Permissive Licences These licences, like [MIT](https://opensource.org/license/mit) and [Apache](https://www.apache.org/licenses/LICENSE-2.0), are designed for flexibility. They allow businesses to use, modify, and incorporate open-source code into proprietary software without requiring them to open-source their own code. The only obligation is to provide proper attribution to the original creators. This model encourages widespread adoption and innovation, making it a favourite for startups and companies seeking scalability without restrictive obligations. ### Copyleft Licences On the other hand, copyleft licences, such as the [GNU General Public License (GPL)](https://www.gnu.org/licenses/gpl-3.0.en.html), take a different approach. They promote a "share-alike" philosophy, requiring any modifications or derivative works that are distributed to also be open-sourced under the same licence. This ensures a continuous cycle of contribution back to the community but may pose challenges for businesses aiming to protect proprietary innovations. These two approaches highlight how open source can serve different purposes: fostering collaboration and ensuring community-driven progress, or enabling flexible use for competitive advantages. [Dynion BakkerDo you have questions about intellectual property for your technical invention? The Netherlands Patent Office gives you information about ways to protect your idea. Our advisors can help you find information in the patent databases. This will save you time and money in developing your technical inventions. Do you have questions about intellectual property for your technical invention? The Netherlands Patent Office gives you information about ways to protect your idea. Our advisors can help you find information in the patent databases. This will save you time and money in developing your technical inventions.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon-3.ico)RVO.nl![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/delen-blue.svg)](https://english.rvo.nl/startup-advisors/dynion-bakker) Panel member Dynion Bakker from The Netherlands Enterprise Agency (RVO) offers his support for startups that have specific questions about IP. ## Use Cases of Open Source: DeepSeek and Llama Two examples of open-source software in action demonstrate these principles: - **DeepSeek**: [DeepSeek-R1](https://rob-hoeijmakers.ghost.io/deepseek-r-1-open-source-ai-model/) is released under the MIT License, a permissive licence that allows users to freely use, modify, and distribute the software, even in proprietary applications, as long as the original licence and copyright notices are retained. This flexibility fosters innovation by removing barriers to adoption and enabling developers to build on existing tools without restrictive obligations. - **Llama**: [Meta’s Llama models](https://rob-hoeijmakers.ghost.io/llama-32-on-a-mac/) operate under a custom licence called the Llama Community License Agreement. While it permits broad use and modification, it imposes specific restrictions, such as prohibiting the use of the model to enhance other large language models. This unique approach blends aspects of permissive licensing with limitations designed to maintain control over how the technology evolves, prompting debates about whether Llama can truly be classified as open source. These examples illustrate how the choice of licence can shape the adoption and impact of open-source tools in both commercial and community-driven contexts. --- ### Further reading - [Reflections on a Deep Tech AI Event](https://rob-hoeijmakers.ghost.io/reflections-on-a-deep-tech-ai-event/) ### DeepSeek R-1: Open-Source AI model URL: https://rob-hoeijmakers.ghost.io/deepseek-r-1-open-source-ai-model/ Last updated: 2026-05-10T08:54:07.000Z **Imagine an AI model that rivals OpenAI’s o1 but costs 13 times less to operate.** That’s exactly what DeepSeek R-1 has accomplished. Developed by Liang Wenfeng’s team at High-Flyer, a Chinese quant-trading firm, this breakthrough is turning heads across the AI community. Marc Andreessen sums it up best: > *"DeepSeek R1 is one of the most amazing and impressive breakthroughs I’ve ever seen – and as open source, a profound gift to the world."* ## What Makes DeepSeek R-1 Special? DeepSeek R-1 leverages techniques like **model distillation** and **sparsity** to achieve its performance edge. With a mere $5.6 million investment for its earlier V3 version, High-Flyer optimized leftover GPU capacity to produce what is now a serious contender in the AI space. The implications are profound. OpenAI and Meta—leaders in closed and open-source AI respectively—are now looking over their shoulders. Meta, once at the top of the open-source game, has reportedly gone into panic mode. OpenAI may also feel the heat, as R-1 and other open-source efforts offer serious competition at a fraction of the cost. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/01/DeepSeek.jpg) Still need to check out Terms of Use and Privacy Policy. ## Why Open-Source Matters As Sebastien Bubeck of OpenAI notes, reasoning models are *“extremely scalable”*. Yet, open-source models like DeepSeek R-1 introduce a democratized dynamic to innovation. Anyone, anywhere in the world, can tinker, improve, and build upon them. The pace of discovery is accelerating as more researchers dive into these open projects. ByteDance, the parent company of TikTok, recently released **Doubao-1.5-pro**, another remarkable model matching GPT-4o’s performance while being **eight times cheaper**. The race is on, and open-source is leading the charge. ## Quick Demo: DeepSeek on iOS Curious how all this translates to practical applications? Let’s take a look at the [**DeepSeek iOS app**](https://apps.apple.com/app/id6737597349), which integrates reasoning models like DeepSeek R-1 into a sleek mobile experience. ## Running DeepSeek Locally For those interested in running DeepSeek R-1 locally, I tried it using [Ollama](https://ollama.com) on a Mac mini M2\. Downloading and running the 14B model was straightforward with the following command: ``` ollama run deepseek-r1:14b ``` Once set up, I tested it by asking about cooking cod. The responses were quick and detailed, demonstrating the model’s impressive reasoning capabilities even on local hardware. Running it locally requires significant resources, but it’s a great option for experimentation or integrating into custom workflows. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/01/DeepSeek-14B-locally-on-a-mac.jpg) DeepSeek 14B locally on a Mac mini ## The Road Ahead While [DeepSeek](https://www.deepseek.com) may be a fast follower rather than a true pioneer, its success proves one thing: **the proliferation of performant models is inevitable.** Businesses and developers now face a rapidly shifting landscape. Open-source will continue to disrupt traditional business models while raising new questions about AI safety. What’s your take on these developments? Could open-source breakthroughs like DeepSeek R-1 redefine the AI landscape? Let me know in the comments! --- ### Further reading - [Deep Tech AI and Open Source licenses](https://rob-hoeijmakers.ghost.io/deep-tech-ai-open-source-licenses/) ### Reversible Computing: A Paradigm Shift in Computer Science? URL: https://rob-hoeijmakers.ghost.io/reversible-computing/ Last updated: 2025-01-23T09:41:41.000Z I heard about **reversible computing** through some experts in the field, and although I now partially understand the underlying processes, I thought it would be good to share what I now know so you can make up your own mind. It is rather complex but also highly promising. Below, I’ll summarise what I’ve learned from a fascinating [YouTube video](https://youtu.be/2CijJaNEh%5FQ) by Anastasia in Tech, which explains the concept in detail (see below). Writing this article is partially my way of processing and partially a way to highlight a potentially transformative idea. It might turn out to be something big, and being there early when it happens could be beneficial. ## Breaking Through the Limits of Traditional Computing For decades, Moore's Law—the idea that the number of transistors on a chip doubles approximately every two years—has driven exponential growth in computing power. This relentless progress has powered the digital revolution, but we are now encountering physical and thermodynamic limits that constrain traditional computing methods. Reversible computing, however, offers a way to break through these barriers. ### The Energy and Heat Problem Modern computing systems consume vast amounts of energy, with billions of devices operating worldwide. Much of this energy is wasted as heat due to the irreversible nature of current computing processes. This is tied to a principle discovered by physicist Rolf Landauer in 1961, known as the *Landauer Limit*. It states that flipping a single bit at room temperature requires a minimal amount of energy. While the energy for one bit is negligible, the cumulative effect across trillions of operations results in significant energy consumption and heat dissipation. ### The Vision of Reversible Computing Reversible computing reimagines this process. Instead of erasing data after each operation, a reversible computer retains it, making it theoretically possible to recover and recycle the energy used in computations. Physicist Richard Feynman and others demonstrated that computation does not inherently require energy if it avoids destroying information. This insight led to the concept of reversible logic gates, which allow computation to be undone, effectively "decomputing" to restore previous states. Unlike conventional logic gates that dissipate energy as heat, reversible gates preserve energy within the system. [🔮 Breaking the energy barrier with reversible computingAs energy toll of AI growth and computation is expanding, the science of reversible computing is more important than ever![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/https-3A-2F-2Fbucketeer-e05bbc84-baa3-437e-9518-adb32be77984.s3.amazonaws.com-2Fpublic-2Fimages-2Faa7453ed-eb02-46d7-b523-74d86b1c1e24-2Fapple-touch-icon-180x180.png)Exponential ViewAzeem Azhar![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/https-3A-2F-2Fsubstack-post-media.s3.amazonaws.com-2Fpublic-2Fimages-2F8dcc397c-a086-4abc-add0-ea3fa8f75082_1600x1224.png)](https://www.exponentialview.co/p/reversible-computing-1) ### How Reversible Computing Works Reversible computing pairs logic gates with resonators—components that store and recycle energy instead of dissipating it. When a reversible gate performs a computation, energy flows into the logic. The resonator then recaptures this energy during the decomputation phase, creating a feedback loop that minimises energy loss. Additionally, adiabatic techniques slowly ramp up voltages and currents to further reduce energy consumption. These methods align perfectly with the requirements of reversible computing, as they depend on preserving information throughout the process. ### Challenges and Opportunities The first reversible computing chips, developed by startups like [Vaire](http://vaire.co), are set to debut in 2025\. While promising, they face several challenges: - **Chip Area**: Reversible gates and resonators take up more silicon real estate than traditional gates. - **Manufacturing Complexity**: Integrating reversible logic and resonators requires innovative engineering. - **Software Development**: Current algorithms are designed for irreversible computing, necessitating new approaches for reversible systems. Despite these hurdles, the potential gains are enormous. By reducing heat dissipation and energy consumption, reversible computing could make vertical stacking of chips more feasible, paving the way for 3D computing architectures. This breakthrough could dramatically enhance the efficiency of applications like AI and machine learning. [The Limits of Current ComputingThe three limits to the performance of current (irreversible) computing.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/10fd5c419ac61637245384e7099e131627900034828f4f386bdaa47a74eae156)MediumDavid Galbraith![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/1-cSczf3MtJ0TyU1Juy4Uduw.png)](https://daveg.medium.com/the-limits-of-current-computing-8b4fca539a4b) ## The Future of Computing Reversible computing represents a radical departure from traditional methods, offering a path to break free from the constraints of Moore's Law. It holds promise not only for silicon-based technologies but also for alternative approaches, such as light-based and DNA computing. While the underlying principles are complex, the potential applications are exciting. The first prototypes will test the feasibility of this idea in practice, and I’m curious to see how much energy can actually be recovered and recycled. Writing this article is my way of sharing this emerging concept and taking the opportunity to learn alongside others. If reverse computing delivers on its promises, it could mark the beginning of a new era in computer science. If this topic resonates with you, feel free to leave a comment (👇) or [follow me](https://rob-hoeijmakers.ghost.io/#/portal/signup) for more insights! ### Enhancing Photos with Apple Intelligence URL: https://rob-hoeijmakers.ghost.io/enhancing-photos-with-apple-intelligence/ Last updated: 2026-04-25T13:16:04.000Z Apple has just rolled out *Apple Intelligence* as part of macOS Sequoia 15.2, the latest version of their operating system. While this feature is also available on iPhones, it hasn’t launched in Europe yet. Being based in the Netherlands, I can’t access it on my phone—and it doesn’t seem likely I will anytime soon. However, I’ve had the chance to play around with it on my M2 Mac, as it’s only available on silicon-equipped Macs. Unfortunately, my trusty Intel iMac doesn’t support it. ## Exploring Apple Intelligence Features Apple Intelligence offers an impressive suite of features. It’s deeply integrated with many existing macOS functions. For instance, I could activate my ChatGPT+ subscription directly from within the system and use it for writing enhancements, summarisation, and other text-related tasks. It even includes tools for image generation and emojis, though I’ll admit—those seem less useful for my needs. ## Introducing the Clean Up Tool in Photos What really caught my attention, though, was the new *cleanup* functionality in the Photos app. If you’ve used Google’s photo tools or dedicated apps like [Topaz Labs](https://www.topazlabs.com/?fbclid=IwY2xjawHvZ-9leHRuA2FlbQIxMAABHUwemrvfXii-ZCupMzjrQDuGXSgBneBKLB%5FgGTXOZkgFxnlY7dmd-5APyQ%5Faem%5F1GG%5F64Xa1CdgAKeauKzMHg), you’ll recognise the concept. The exciting difference here is that it’s built natively into macOS. ## How to Use the Clean Up Tool Here’s how it works: 1. Open the photo library and select the photo you want to edit. 2. Activate the Intelligence feature and enter the edit mode. 3. Select the cleanup tool. The system automatically identifies elements that it suggests for removal, like people in the background, cars, bikes, or even shadows. 4. Tap or click on the suggested elements to remove them instantly. If the tool doesn’t preselect something you want gone, you can manually outline it with a quick swipe or draw. ## Performance and Impressions I’ve found the tool particularly useful for quick touch-ups. It’s not always professional-grade in its results—some spots where elements were removed still show subtle traces—but it’s a significant step forward for anyone looking to clean up images without relying on third-party software. For photographers who need a fast and simple way to tidy up their photos, this feature is a game changer. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/01/IMG_0506.jpeg) ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/01/IMG_0506-2.jpeg) Cleaned Up. ### LLM-Driven Automation and Anthropic’s Computer Use URL: https://rob-hoeijmakers.ghost.io/llm-driven-automation-and-anthropics-computer-use/ Last updated: 2025-01-17T11:03:04.000Z Exploring the forefront of technology can be both exciting and daunting. With the rapid advancements in [large language models (LLMs)](https://rob-hoeijmakers.ghost.io/a-comparative-analysis-of-top-language-models/), tools like Anthropic’s Computer Use have emerged as experimental platforms that showcase the potential of LLM-driven automation. This article serves as a quick, accessible overview for those who may not have the time or resources to experiment with these tools themselves but are curious about what’s possible. ## Anthropic’s Computer Use: Computers Powered by Natural Language At the heart of this exploration is Anthropic’s Computer Use, an experimental system where an LLM can control a computer interface using natural language commands. Imagine assigning tasks to a virtual assistant that navigates browsers, operates spreadsheets, or uses calculators on your behalf. While the technology is far from fully functional or scalable, it provides a glimpse into how computers can be activated and driven by LLMs. To illustrate its capabilities, I’ve included videos demonstrating real-world interactions with a Virtual Network Computer (VNC). These examples highlight both the potential and the current limitations of the technology. It’s [experimental](https://github.com/anthropics/anthropic-quickstarts/tree/main/computer-use-demo), yes, but it’s also fascinating to see how intuitive human-machine interaction could become. Having an LLM drive computer actions through natural language ## The Bigger Picture: LLMs in Automation While Anthropic’s VNC is a compelling case study, it’s part of a much broader landscape of automation tools. Traditional solutions like Apple Automator and make.com offer visually driven workflows, enabling users to automate repetitive tasks without writing complex code. These tools have empowered countless individuals to streamline their workflows. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/01/image-3.png) AI Assistance on make.com really changes the game. As long as you precisely know what you want and roughly how to do it. Now, LLMs are adding an exciting layer to this mix. For example, I’ve used make.com, which incorporates an LLM in a chatbot-style interface to assist users intuitively. With Apple Automator, which I initially found too complex, I leveraged ChatGPT to create a promising local automation for managing monthly invoices with my bookkeeping software. This combination avoids sharing credentials or dealing with [2FA](https://rob-hoeijmakers.ghost.io/two-factor-authentication-2fa/), blending the best of local and intelligent systems. Beyond simply performing tasks, LLMs have the potential to analyze your computer behavior, identify patterns, and suggest or implement automation. For instance, an LLM could observe how you repeatedly format data in a spreadsheet and then create a script to handle it automatically. This represents a significant shift from predefined workflows to adaptive, intelligent systems. Use LLM's to copilot you with on device automations. ## The Role of Video in Automation Screen recording and screencasting are valuable tools in this space. For those experimenting with or showcasing automation, videos help bridge the gap between abstract concepts and tangible results. By watching an LLM in action—navigating menus, solving problems, or completing assignments—viewers can better grasp its capabilities and limitations. Videos also serve as excellent documentation for complex setups, making them more accessible to broader audiences. As automation tools grow in sophistication, the ability to demonstrate their functionality visually becomes increasingly important. In this regard, I’ve been exploring [**Google’s AI Studio**](http://aistudio.google.com), which is absolutely spectacular and promising. I plan to write about it soon, as it offers unique insights into how these technologies can evolve. Use video assistance on setting up interactive workflows ## Why This Matters The intersection of LLMs and automation is not just a technical curiosity; it’s a preview of how technology could reshape work and productivity. By understanding the possibilities today, we can better prepare for the tools of tomorrow. This whistle-stop tour is designed to give you a quick taste of what’s possible without the need to set everything up yourself. Whether it’s Anthropic’s Computer Use, visually driven, AI enhanced tools like make.com, or the integration of LLMs into existing workflows, the future of automation is closer than you think. [Stay tuned for more explorations](https://rob-hoeijmakers.ghost.io/#/portal/signup) and insights into this rapidly evolving field. I’d love to hear your thoughts, so feel free to leave a comment below. And if you’re curious about anything specific, let me know—I’m always happy to dive deeper! An extra video showing off the video capabilities of Google Gemini. ### A Better Way to Export Kobo Notes and Highlights URL: https://rob-hoeijmakers.ghost.io/a-better-way-to-export-kobo-notes-and-highlights/ Last updated: 2026-04-14T13:22:23.000Z A while back, I shared [a detailed article on extracting notes and highlights from your Kobo device](https://rob-hoeijmakers.ghost.io/kobo-highlights-and-notes-to-your-mac/). At the time, it involved a complex, almost hack-like process to get the job done. While the method worked, I quickly realised that it wasn't practical for most users—too many steps, too much room for error, and frankly, it felt a bit like cracking into your own device. The process I shared required connecting your Kobo to your computer, navigating through hidden folders, and manually pulling files to access your notes. It was a solution, but it wasn’t elegant. ## The Email That Changed Everything Fast forward to last week: I received an email from a developer, Up Chen, who had read my original article. He introduced himself and casually mentioned that he’d come up with a simpler, much more user-friendly way to achieve the same goal. Intrigued, I decided to give his solution a try—and I have to say, I’m impressed. This developer has created a lightweight browser-based tool that streamlines the entire process. Here’s how it works: 1. Connect your Kobo to your computer (Mac or Windows) using a USB cable. 2. Open Chrome (the tool is currently optimised for this browser). 3. Point the tool to the drive where your Kobo is connected. That’s it. From there, the software does the heavy lifting—directly in your browser. It reads the data locally (so nothing gets uploaded), and you can quickly navigate through your books and export your notes and highlights. The export options include plain text or [Markdown](https://rob-hoeijmakers.ghost.io/using-markdown-for-text/), which is perfect for reusing your notes however you like. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/01/Kobo-notes-and-highlights.jpg) The main screen with e-books lined up. ## No Installs. No Accounts. No Cost. What makes this tool even more remarkable is its simplicity. You don’t need to install anything, create an account, or worry about hidden fees. It’s free, lightweight, and works seamlessly. For now, Up Chen is offering the tool completely free of charge, though he’s considering adding a subscription or donation model in the future. If you’ve struggled with my original method—or never even attempted it because it sounded too technical—this is the solution for you. ## Credits Where They’re Due A huge thanks to Up Chen for reaching out and sharing his work. I love seeing innovation like this, especially when it makes life easier for everyday users. If you want to try the tool, you can find it on his product website: Up Chen also features his product on [Product Hunt](https://www.producthunt.com/products/kobo-note-up), so feel free to check it out and show your support there. ## Quick demo ### A Busy Start to the Year: Exploring AI, Agents, and Automation URL: https://rob-hoeijmakers.ghost.io/a-busy-start-to-the-year-exploring-ai-agents-and-automation/ Last updated: 2026-05-26T19:37:25.000Z While I didn’t manage to finish anything completely, I gained valuable insights that will shape my projects in the weeks ahead. Here’s a recap of what I’ve been up to. ## Discovering Gemini Video One of the standout moments this week was experimenting with [Gemini Video (Google AI Studio)](https://aistudio.google.com). This tool uses AI to analyse video content with remarkable precision. From screencasts to real-time footage, its capabilities are nothing short of impressive. I conducted a small experiment by filming my bookshelf, having Gemini scan the video for text recognition (OCR). It not only identified the titles but categorised them, demonstrating its potential for organising visual data. This left me eager to explore its applications further, from UX analysis to creative storytelling. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/01/image-2.png) Here Gemini analysis a video of a toy train riding under a couch ## Deepening My Understanding of AI Agents I also dived into the world of AI agents, starting with an intriguing article defining what makes a "true agent." [AgentsIntelligent agents are considered by many to be the ultimate goal of AI. The classic book by Stuart Russell and Peter Norvig, Artificial Intelligence: A Modern Approach (Prentice Hall, 1995), defines the field of AI research as “the study and design of rational agents.”![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/favicon-1.ico)Chip HuyenChip Huyen![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/2-agent-pattern.png)](https://huyenchip.com//2025/01/07/agents.html?utm%5Fsource=tldrnewsletter) Key factors include its environment—be it digital or physical—its memory, and its ability to act autonomously. There is a clear theme emerging: AI is evolving to become more agent-like, with increasing modalities to provide it with agency. > If your “AI agent” only reacts from an external cron, it’s a regular LLM. > > If your “AI agent” has no external memory, it’s a regular LLM. > > Giving an API wrapper hooks, or web access doesn’t make it more of an agent than any other response engine. > > AI agents will be \*huge\* but 90%… > > — Adam Cochran (adamscochran.eth) (@adamscochran) [December 31, 2024](https://twitter.com/adamscochran/status/1874117425085444108?ref%5Fsrc=twsrc%5Etfw) When AI operates in a digital environment, like the web, it can leverage multiple pathways. It might rely on vision-based inputs, such as analysing visual content or interfaces, or interact directly through APIs. This versatility is what makes modern AI agents so powerful and adaptable. [Building effective agentsA post for developers with advice and workflows for building effective AI agents![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/apple-touch-icon-2.png)![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/b05cf65de663b0b93909dee5071c73b273a3cef3-2560x1344.png)](https://www.anthropic.com/research/building-effective-agents) ## Testing Anthropics’ Browser Tools Another highlight was experimenting with [Anthropics’ browser extension](https://rob-hoeijmakers.ghost.io/llm-driven-automation-and-anthropics-computer-use/), a tool designed for natural language interaction with the web. It’s still a bit slow but promising in concept—imagine having AI assist you in tasks like looking up information, managing emails, or automating processes. [LLM-Driven Automation and Anthropic’s Computer UseAnthropic’s Computer Use, make.com, and Automator: See how LLMs simplify automation, analyze tasks, and redefine productivity. Watch examples and learn more!![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-11.png)Rob Hoeijmakers (robhoeij)Rob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_9049.jpeg)](https://rob-hoeijmakers.ghost.io/llm-driven-automation-and-anthropics-computer-use/) ## Exploring APIs and Automation with Make.com Finally, I turned my attention to automation using tools like [Make.com](https://make.com). While trying to streamline tasks like retrieving OpenAI billing invoices, I encountered some challenges. Yet, I was struck by how AI simplified the process, filling in gaps where my technical knowledge fell short. This combination of abstraction and precision is a great help for complex workflows. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/01/image-1.png) Make.com - AI Assistant (Beta) \--- This week has been a mix of exploration and learning, setting the stage for exciting projects in the coming months. Stay tuned as I continue experimenting, and let’s keep the curiosity alive! *Do you have experiences with these tools or suggestions for what I should try next?* [*Share your thoughts—I’d love to hear from you.*](https://rob-hoeijmakers.ghost.io/about/) P.s. And if you still want to read further, here is a great update by [Ethan Mollick on AGI and Agents](https://www.oneusefulthing.org/p/prophecies-of-the-flood?utm%5Fsource=substack&publication%5Fid=1180644&post%5Fid=154445396&utm%5Fmedium=email&utm%5Fcontent=share&utm%5Fcampaign=email-share&triggerShare=true&isFreemail=true&r=49ph0&triedRedirect=true). ### The upgraded iPhone Calculator URL: https://rob-hoeijmakers.ghost.io/the-upgraded-iphone-calculator/ Last updated: 2026-04-25T08:31:28.000Z Kick off the year by exploring the upgraded iPhone and Mac calculator, an update that coincided with iOS 18, initially overlooked but worth revisiting. This handy tool now includes features that simplify everyday tasks. Quick demo of the improved iOS calculator. ## New "Convert" Functionality The **Convert** feature allows quick conversions for: - **Currency**: Convert between currencies with up-to-date rates. - **Weight**: Switch between pounds and kilograms effortlessly. - **Temperature**: Toggle between Celsius and Fahrenheit. - **Energy**: Convert calories to joules in a tap. Simply choose a category, input a value, and let the app do the rest. ## Switching Between Modes Easily switch between **Basic** and **Scientific** modes: 1. Open the Calculator app. 2. Tap the calculator icon to change modes. 3. Access the Convert feature in both modes. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2025/04/iphone-calculator.jpeg) Here is where you open up the conversions. ## Mac Integration The Mac version offers the same seamless functionality, from currency conversions to energy calculations, ensuring consistency across devices. ## Why It Matters Stock apps like the Calculator are essential because they’re always at hand when you need quick conversions. These moments often involve time-sensitive tasks like splitting bills, budgeting during shopping, or double-checking recipes while cooking. The Calculator’s reliable and accessible features ensure you can handle these conversions without delay, making it an indispensable tool for everyday efficiency. ### From Mifi to eSIM: Finally Catching Up with Mobile Connectivity URL: https://rob-hoeijmakers.ghost.io/from-mifi-to-esim-finally-catching-up-with-mobile-connectivity/ Last updated: 2026-05-10T08:54:08.000Z It’s taken me a while, but I’ve finally caught up with the eSIM revolution—and I’m glad I did. After years of juggling Mifi routers and prepaid SIM cards, I decided to test an eSIM for my trip to Switzerland. The experience was so seamless and convenient that it left me wondering why I didn’t make the switch sooner. Here’s how it went and why I’m leaving my old connectivity solutions behind. ## Why I Finally Switched to an eSIM For years, I relied on Mifi routers to stay connected during trips, and prepaid SIM cards were my go-to for mobile data. While these worked, they came with their share of challenges—extra devices to carry, [fiddling with physical SIMs](https://rob-hoeijmakers.ghost.io/from-physical-sim-to-esim-on-iphone/), and the uncertainty of finding a plan that worked well abroad. When I learned more about eSIMs and their convenience, I decided it was time to make the leap. 💡 ****Summer holiday 2025 update:** Used eSIM and worked like a charm. ordered and activated one for my wife’s phone and one for my own phone (extra GB’s). Tip: you can use a [Europe wide one](https://maya.net/plans/europe), it includes Switzerland. ## The Process: Easy, Quick, and Affordable ordered an eSIM online and received a QR code to set it up. The installation process was straightforward, and the eSIM activated right on schedule. For €8, I got 5GB of data, which turned out to be fast and reliable for my typical needs. The only adjustment I had to make was turning off roaming data when I crossed the German-Swiss border, after which the eSIM worked flawlessly. The simplicity of the process made me realise how much time I used to waste on physical SIMs and routers. [Maya Mobile - Travel eSIM - International Data PlansSay goodbye to roaming fees, and enjoy unlimited high-speed 4G & 5G mobile data in 200+ countries worldwide. Choose your travel eSIM and activate in 5 minutes!![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/icon-2.png)Maya Mobile LogoMaya Virtual, Inc.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/maya-mobile-embed-graphic-1.jpg)](https://maya.net) I bought my eSIM from maya.net ## Goodbye Mifi and Prepaid SIMs Now that I’ve experienced the convenience of an eSIM, I can’t see myself going back to the old ways. No more carrying extra devices or worrying about physical SIM cards. eSIMs are not just affordable—they’re efficient, reliable, and a far better fit for modern travel. If, like me, you’ve been hesitant to make the switch, trust me: it’s worth catching up. The future of connectivity is here, and it’s never been easier to embrace. ⚠️ eSIMs are wonderfully convenient, but recent research shows a hidden trade-off: some travel eSIM providers route data through unexpected countries (including China), exposing traffic to unfamiliar jurisdictions and surveillance risks. A reminder that digital convenience often hides infrastructural choices we don’t see. [Read more](https://www.itnews.com.au/news/travel-esims-secretly-route-traffic-over-chinese-and-undisclosed-networks-study-619659). ### Further reading - [A Mifi hotspot for my mountain holidays](https://rob-hoeijmakers.ghost.io/a-mifi-router-for-my-mountain-holidays/) ### When Macs Are No Longer Safe: Lessons from a Trojan Incident URL: https://rob-hoeijmakers.ghost.io/when-macs-are-no-longer-safe-lessons-from-a-trojan-incident/ Last updated: 2026-05-26T19:38:14.000Z Here’s the story of an international entrepreneur’s close call with the **AMOS Trojan** and the lessons we can all learn. ## A Seemingly Trustworthy Opportunity The entrepreneur operates a crypto-enabled gaming project with global reach. Recently, they were approached by a prominent venture capital firm. Over several weeks, the VC firm established trust through professional communications on platforms like Twitter, Telegram, and email. Initial meetings appeared genuine, with well-branded materials and relevant discussions. However, during a critical call, the firm claimed that key stakeholders, based in China, preferred using WeChat for the discussion. They provided a link to a WeChat Teams application, urging its quick installation. ## The Trojan Horse Revealed Though the WeChat Teams link appeared legitimate, the entrepreneur and their team grew suspicious. Upon downloading the software, the application bypassed standard macOS security prompts. One team member ran the file through VirusTotal, a service that scans files for malware. The result? It was flagged as the AMOS Trojan, a so called stealer. Realising the potential threat, the entrepreneur immediately disconnected their Mac from the internet and reformatted the system to ensure complete removal of the Trojan. ⚠️ ****The Gravity and Reach of AMOS Trojan**: The AMOS Trojan is not just a typical malware; it represents a calculated threat targeting high-value systems, often using advanced social engineering tactics. Its ability to bypass macOS security measures and exploit critical data, such as financial credentials and sensitive documents, underlines its severity. The Trojan’s global footprint, combined with its use of sophisticated disguises, highlights the need for heightened vigilance in both personal and professional digital environments. ## Social Engineering at Its Finest This incident underscores the growing sophistication of social engineering tactics. The attackers used urgent scenarios to bypass critical thinking, presented professional branding and realistic communications, and exploited the team’s trust and eagerness to move forward. Social engineering is particularly dangerous because it targets human behaviour, not just technology. ## Lessons for macOS Users Key takeaways for individuals and organisations: - **Scrutinise software requests.** Avoid downloading or installing software in haste and verify the legitimacy of applications with tools like VirusTotal. - **Isolate your system.** Disconnect from the internet if a threat is suspected and use a sandbox or virtual machine to test unknown software. - **Adopt proactive security measures.** Keep macOS updated with the latest security patches, use trusted antivirus and malware detection tools, and regularly update passwords [while avoiding keychain storage](https://rob-hoeijmakers.ghost.io/mac-and-ios-password-manager/) (which I will research further). - **Be wary of social engineering.** Stay alert for red flags, such as urgency or missing video feeds during calls, and remember that even professional-looking communications can be deceptive. ## The Bigger Picture The AMOS Trojan highlights the rise of malware targeting macOS. Attackers are leveraging AI and malware-as-a-service to create accessible and effective tools. While Macs were once considered safe, their growing popularity makes them attractive targets. For international businesses, particularly in sensitive sectors like crypto, heightened vigilance is essential. The cost of a breach can be immense. ## Final Thoughts This close call with the AMOS Trojan reminds us that no platform is invulnerable. By adopting proactive measures and staying alert to social engineering tactics, macOS users can protect themselves from evolving threats. Stay safe, stay informed, and always question the links you click. --- ### Further reading - [MacOS users beware! AMOS Trojan](https://xavie.mirmon.co.uk/blog/macos-users-beware-amos-trojan) 🦠 ****Malware-as-a-Service (MaaS)**: The AMOS Trojan exemplifies the growing market for Malware-as-a-Service, where malicious tools are sold or rented like any other software product. This model makes it easier for less technically skilled attackers to launch sophisticated campaigns, reducing barriers to entry for cybercrime. MaaS platforms are well-organised, offering customer support, regular updates, and even user-friendly interfaces, which drastically increases the global reach and impact of malware like AMOS. ### The Pinch That Proves It's Not Symmetry URL: https://rob-hoeijmakers.ghost.io/the-pinch-that-proves-its-not-symmetry/ Last updated: 2026-04-25T13:17:09.000Z Photography has this wonderful ability to challenge what we see—or think we see. A recent photo of mine sparked a lot of comments about its “symmetry,” but here’s the thing: it’s not symmetrical. And that’s precisely what I love about it. The shot in question was of an Amsterdam street, taken from above, with its rooftops, windows, and reflections guiding the eye into the distance. It has a single-point perspective—your eye naturally travels along the street to the depth of the frame. But symmetry? Look closer. It’s an illusion. I actually worked on the composition to make it feel harmonious. The split line wasn’t originally centred; it was at two-thirds horizontally. By cropping the image to a square and placing the split line dead in the middle, I created a stronger sense of balance. But balance isn’t the same as symmetry. It’s composition that makes it work. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/12/IMG_8674-2.jpeg) Split line on two third of the photograph. This brings me to a tool I’ve recently fallen in love with: **the Pinch feature in** [**Threads**](https://www.threads.com/@robhoeij). It lets you divide a photo into two perfect halves, effectively splitting the image and encouraging you to examine it differently. When I used it on this photo, it clarified what I’d been feeling all along: this isn’t about symmetry—it’s about harmony in composition. [Rob Hoeijmakers (@robhoeij) on ThreadsPinch to match 👉👈![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Pcnemah90K8.png)Threads![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/470334493_449278691552960_4742375094138919066_n.jpg)](https://www.threads.net/@robhoeij/post/DDhEVf5sV0z?xmt=AQGz%5FrLWoou4X1UcxKVznhcjFprgxsXcpLuaryo06iskFw) ## Splitting to See More By pinching the photo, you can immediately see how different the two halves are. On the left, there’s the charming geometry of Amsterdam’s brick facades, their warm lights glowing against the grey day. On the right, the sharp angles of the roof tiles and the glossy reflection in the gutter. The two sides contrast and complement each other, but they’re far from mirror images. 0:00 /0:05 1× The Threads pinch feature at work What the Pinch feature does is guide your eye to the composition itself. It strips away the illusion of symmetry and makes you appreciate the balance and the depth created by the **single-point perspective**. It’s fascinating how a small edit can reveal so much more about a photo’s structure. ## Why Composition Matters More Than Symmetry Symmetry can be beautiful, but for me, it’s the intentional composition that gives a photo its soul. By centring the split line, I wasn’t aiming for perfection or duplication—I was creating a visual rhythm that guides the viewer. The centred perspective draws you in, and the contrast between the two halves keeps you engaged. Using the Pinch feature reinforced this idea. It not only lets me experiment but also helps me see my own work with fresh eyes. When the halves are split, the composition becomes even more apparent, and the illusion of symmetry falls away. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/12/IMG_8674-1-1.jpeg) ## Experimenting with Pinch For photographers who love playing with depth, contrast, or abstract ideas, this feature in Threads is worth exploring. It’s not just about creating something new—it’s about discovering what’s already there. I’ve included the video above to show how the Pinch feature works (touch screen only) and how I used it on this photo. If you’ve ever wondered how to challenge visual assumptions, or if you’re just looking for a fresh way to interact with your own photos, give it a try. For me, this journey started with a simple photo of a rainy Amsterdam street. But through composition, perspective, and a little pinch of creativity, it became something much more. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/12/ChatGPT-split.jpg) Tip: Ask ChatGPT to split it for you ### Setting up a WhatsApp Community URL: https://rob-hoeijmakers.ghost.io/setting-up-a-whatsapp-community/ Last updated: 2026-04-25T13:16:00.000Z Over the past few weeks I've been trying out something new that I'd like to share with you: setting up a [WhatsApp Community](https://rob-hoeijmakers.ghost.io/what-is-a-whatsapp-community-an-introduction/). This is an interesting way to bring different groups of people together under one roof, and I have applied it myself for this own project around chat for companies. In this article I will show you how to set up and manage such a community yourself. ### What is a WhatsApp Community? Before we dive into the depths, a brief explanation. A [WhatsApp Community](https://rob-hoeijmakers.ghost.io/what-is-a-whatsapp-community-an-introduction/) is an umbrella platform in which you can manage multiple[ WhatsApp groups](https://rob-hoeijmakers.ghost.io/whatsapp-groups-vs-communities/). Imagine: you have one main community and within that you can create different subgroups, each with a specific topic. Ideal for a neighborhood association, a project team, or – as in my case – a network of professionals. ⚠️ In WhatsApp Communities, the groups functionality in particular poses a privacy challenge: within a group, participants can see each other's name and phone number. This can be a problem for companies, as they often have strict rules around sharing customer and employee data. This makes WhatsApp groups less suitable for applications where anonymity or data protection is crucial. ### Setting up the community Setting up a WhatsApp Community is easier than you think. When I started, this is how I went about it: 1. **Start by creating a new community**: Open WhatsApp, go to the community section and choose "Create a new community". You can immediately enter a name and description that make it clear what the community is about. In my case, I chose "Rob Hoeijmakers (Robhoeij)". 2. **Add Groups**: After creating your community, you can immediately start adding subgroups. I created groups for “General,” “Digital Life,” and “Photography.” Each of these groups has its own focus, which helps keep discussions and information organized. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/12/image-1-1.png) ### Invite people Now comes the fun part: bringing your community to life by inviting people. This can be done in several ways: - **Direct invitation**: From your contacts you can add people directly to the community. For example, I added Rob Hoeijmakers. This is very easy by clicking on his name and pressing “Add”. - **Shareable link**: Do you want people to be able to choose whether or not to participate? You can also generate a link that you can share via WhatsApp or other channels. This is useful if you want to invite a larger group without having to add everyone manually. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/12/image-2-1.png) [WhatsApp: Groups vs. CommunitiesLearn how to strike the right balance between openness and organization to optimize your WhatsApp experience, using a tennis club as a metaphor.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-8.png)Rob Hoeijmakers (robhoeij)Rob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_8100-1-1-1.jpeg)](https://rob-hoeijmakers.ghost.io/whatsapp-groups-vs-communities/) ### Management of communications As an administrator, you have several options to ensure the community functions properly: - **Making Announcements** : You can send messages to the entire community via the Announcements group. This is a one-way group that only admins can post to, so important information doesn't get lost in the conversation. - **Customize Group Images and Descriptions** : You can easily manage the look and feel of your groups by customizing the group images and descriptions. This keeps everything professional and organized. ### **Membership and group management** As an administrator of a WhatsApp community, you can determine who can add members. This can be set at two levels: - **Adding members** : You can set it so that only admins can add members, or all members have this ability. This allows you to maintain control over who has access to the community, which is especially important in professional or larger environments. - **Adding Groups** : Admins also have the option to control who can add new groups to the community. Again, you can choose to keep these rights exclusive to admins, or open them up to all members. This prevents the community from becoming cluttered by uncontrolled group creations. ### The user experience As a member of a community, you have a few simple yet powerful options: - **Discover Groups** : Members can easily browse through the different groups and join the one that interests them most. - **Manage Notifications** : Each member can choose which notifications they want to receive, so they only get updates from the groups that are relevant to them. ### Want to try it yourself? Curious? You can try it yourself by joining my test community. [Click here to join now and see how it works.](https://chat.whatsapp.com/JgNe9r1bJZuDQUDRyBhakO) 💬 You can only set up a community and group via the personal WhatsApp app and not the business WhatsApp. You can participate with both platforms. ### Beyond keyboards and screens URL: https://rob-hoeijmakers.ghost.io/beyond-keyboards-and-screens/ Last updated: 2026-05-10T08:54:08.000Z A few weeks ago, I reconnected with [Bram Donkers, a UX expert](https://www.aratio.nl) I had met over 20 years ago. We crossed paths again at an AI conference, where we discovered a shared interest in how large language models (LLMs) and AI are transforming human interaction with technology. That reconnection sparked a conversation that led to this experiment: writing an article entirely through dictation, embracing the idea of a future without keyboards or screens. ## Voice as a Natural Interaction Medium For decades, keyboards and screens have dominated digital workflows. However, as Bram and I discussed, voice represents a more intuitive and human-centered interaction method. It removes the friction of typing, allowing ideas to flow naturally. Speaking feels unencumbered by the mechanics of traditional input methods, enabling professionals to focus on substance rather than form. [From Typing to TalkingFrom speech synthesis to dictation, I explore how evolving AI tools make communication swifter and more natural. Dive into my journey through spoken and written digital modes.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-1.png)Rob Hoeijmakers (robhoeij)Rob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_8495-1.jpeg)](https://rob-hoeijmakers.ghost.io/from-typing-to-talking/) ### Two Distinct Workflows Voice workflows can be broadly divided into two approaches, each catering to specific needs: 1. **Directive Workflow**: Dictation works best for well-defined tasks or topics where the structure is clear. Professionals can efficiently capture ideas, instructions, or drafts without breaking focus. 2. **Exploratory Workflow**: For brainstorming or tackling open-ended questions, engaging in dialogue—with an LLM or a peer—is more effective. These conversational workflows help uncover new perspectives and refine thoughts. While exploratory voice interaction has significant potential, it often requires active focus, which limits multitasking capabilities. ### Efficiency Gains Voice-first workflows offer several advantages for online professionals: 1. **Speed**: Speaking is often faster than typing, making voice an efficient way to capture thoughts and complete tasks. 2. **Flexibility**: Voice interaction liberates professionals from screens, allowing them to dictate while walking or multitasking, provided they maintain clarity and focus. ## Voice and Conversational UX Voice interfaces excel in exploratory workflows, fostering creativity and dynamic problem-solving. Conversational UX—a design approach that prioritises natural, human-like interactions—enables users to: - Engage in Socratic dialogues with AI tools. - Iterate on ideas in real time. - Receive adaptive, meaningful feedback. This interactive design not only enhances productivity but also creates a more enjoyable and immersive user experience. ## Challenges of Voice-First Interfaces Despite its benefits, voice technology poses challenges that must be addressed for broader adoption by online professionals: ### Speed and Comprehension - **Listening vs. Reading**: Consuming information through voice is slower than scanning text, making it less practical for quick information intake. - **Dictation Skills**: Effective voice input requires practice in pacing, clarity, and incorporating punctuation naturally. ### Privacy Concern - Cloud-based voice processing raises questions about data ownership and security. - Local-first solutions are crucial for ensuring privacy, particularly for sensitive information in professional contexts. ### Interface Fragmentation The current landscape of voice input tools—spanning operating systems, apps, and third-party platforms—creates inconsistency and confusion. Unified systems are needed to streamline the experience. [Whisper by OpenAI: Transforming Speech-to-Text TechDive into the world of Whisper by OpenAI. See how this groundbreaking speech-to-text tool is changing business communication and tech innovation.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-2.png)Rob Hoeijmakers (robhoeij)Rob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_8014-1-1.jpeg)](https://rob-hoeijmakers.ghost.io/openais-whisper-speech-to-text-technology/) ## The Path Forward: Improving Voice Interfaces Our discussion identified several key areas for improvement in voice-first technology: ### 1\. Unified Voice Capture A centralised, OS-level voice capture feature could simplify workflows. By processing voice locally and directing transcripts to preferred tools, professionals can avoid the fragmentation of current systems. ### 2\. Personalisation and Flexibility Voice interfaces should adapt to individual needs. Customisable settings for tone, speed, and tool integration would enhance usability and efficiency. ### 3\. Ethical Design Developers must prioritise transparency and control in voice data processing. Local-first options and clear user controls build trust, especially for applications handling sensitive information, such as healthcare or confidential business contexts. Vision technology. ## The Future of Voice for Online Professionals As voice technology matures, it has the potential to revolutionise workflows for online professionals. By eliminating the need for keyboards and screens, voice-first workflows can enhance creativity, efficiency, and accessibility. However, success depends on addressing critical challenges such as privacy, interface fragmentation, and user adaptation. Reconnecting with Bram reaffirmed the transformative potential of voice technology. For professionals like us, adopting voice is not merely about embracing a new tool—it’s about rethinking how we engage with technology to align with our natural ways of thinking and communicating. Bram and I really hit it off, and we decided to continue our conversations with a focus on[ healthcare-specific environments](https://rob-hoeijmakers.ghost.io/the-ai-revolution-in-medicine-gpt-4-and-beyond/). Stay tuned for more insights as we explore the possibilities in this critical domain. The shift to voice is already underway. Are you ready to speak your future into existence? [Stay tuned](https://rob-hoeijmakers.ghost.io/#/portal/signup). --- ### Further reading - [Playing in the interactive sandbox](https://rob-hoeijmakers.ghost.io/playing-in-the-interactive-sandbox/) - [LiveDrop: Offline Data Sharing](https://rob-hoeijmakers.ghost.io/exploring-livedrop-offline-data-sharing/) ### How I’m Rethinking Social Media for 2025 URL: https://rob-hoeijmakers.ghost.io/rethinking-social-media-for-2025/ Last updated: 2026-05-10T08:54:08.000Z A few weeks ago, something unexpected happened: my Instagram followers surged. For a moment, I felt that rush of validation that social media often dangles in front of us. But almost as quickly, questions began to bubble up. How do people even find me? Why is one post seen by thousands and another buried? This moment sparked a decision to take my online presence seriously—not just as scattered pieces on different platforms, but as a unified whole. That’s when the idea of "[robhoeij](https://rob-hoeijmakers.ghost.io/)" as a personal brand came to life. But before I could move forward, I needed to understand how social media platforms really work—and how I can work them. ## How Platforms Decide What We See Social media feels personal, but behind the scenes, it’s not. What you see in your feed isn’t determined by who you follow but by what the platform thinks will keep you scrolling. This isn’t just speculation—it’s the very design of modern social media. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/12/Algo-s.jpg) Feed complete Here’s how it works, in simple terms: - **Algorithms** analyze user behavior. - They rank every post based on how likely it is to engage you. - Posts that keep people hooked get amplified. It’s like walking into a crowded room where everyone is shouting. The algorithm listens for the loudest, most engaging voices and makes them even louder. That’s why my posts don’t always reach my followers, even though they chose to follow me. On TikTok, this idea is taken to the extreme. Your followers barely matter; the platform’s entire focus is on showing the most engaging content possible. Instagram and Facebook have followed suit. That’s the game now: **engagement, not followers**. ❗ While followers aren’t the ultimate measure of success, they undeniably matter. For many, they represent credibility and community, but more importantly, they are essential for algorithms. Without a decent following, even the best content is less likely to gain traction, as algorithms favour accounts with established audiences. It’s a reality worth acknowledging: followers might not define your worth, but they do influence your chances of being seen. ## **My Strategy for Gaming the System** Once I realized how platforms work, I stopped worrying about my follower count. Instead, I started thinking about how to create a web of interconnected platforms that amplify each other. Here’s the plan: 1. **Unify Under “robhoeij”** Instead of treating blogging, posting, and photography as separate projects, I’m bringing everything under one personal brand. Whether you’re reading my blog, scrolling through Instagram, or checking out my Facebook page, it should all feel like part of the same story. 2. **Leverage Each Platform’s Strengths** Each platform I use serves a unique purpose, allowing me to tailor my content and engagement style to its strengths while building a cohesive presence across my digital ecosystem. Here's how I approach each one: - [**Instagram**](https://www.instagram.com/robhoeij) for visual storytelling. - [**Facebook**](https://www.facebook.com/robhoeij) for a different demographic, giving me an opportunity to write longer stories to accompany my photos, and a place to showcase landscape-oriented photos. - [**WhatsApp**](https://wa.me/31628129030) for communities, groups, and channels. I’ve launched an AI chatbot here and see enormous potential in its unique mix of intimacy, functionality, and scalability. - [**LinkedIn**](https://www.linkedin.com/in/rob-hoeijmakers) for professional insights. - [**Blogging**](https://rob-hoeijmakers.ghost.io/) for deep dives and reflections like this, where I also use my photos as the default image and plan to blog more about photography alongside my insights on digital life and AI. - [**Threads**](https://www.threads.com/@robhoeij)**/**[**X**](https://x.com/robhoeij) for quick thoughts, banter and behind-the-scenes insights. I also like to do polls. - [**Website**](https://rob-hoeijmakers.ghost.io/) to serve as the central hub of everything—explaining who I am, showcasing my photos, and offering prints for sale. I'll make sure to create consistency around my brand in color, style, and imagery. - [**YouTube**](https://youtube.com/@robhoeij): I will split the current hosting of video's and stand alone video's. I will make more stand alone items on digital life and photography. 1. **Cross-Promote Content** Each platform will feed the others. For example, I might tease a blog post on Instagram Stories or link my Facebook updates back to my blog. This way, my content works together, not in isolation. 2. **Repurpose with AI** Using [LLMs](https://rob-hoeijmakers.ghost.io/a-comparative-analysis-of-top-language-models/), I can adapt my content for each platform. A long blog post can become a punchy Instagram caption, a tweet, or even a short video script. This saves time and ensures my voice stays consistent. ## Balancing Algorithms with Authenticity Of course, there’s a risk in playing this game: it’s easy to lose yourself in the race for engagement. Platforms reward catchy hooks and viral trends, but they don’t care about authenticity. That’s why my goal is to balance the two. I’m not here to chase trends blindly. Instead, I’ll focus on sharing stories that resonate with me, while also being smart about how I package them. After all, algorithms reward engagement, but humans reward authenticity. 💡 I primarily focus on two ecosystems: [Meta](https://webstrategiesblog.com/tag/meta/) (Facebook, Instagram, Threads, WhatsApp) and Google (through [SEO for search traffic](https://webstrategiesblog.com/tag/seo/)). While AI models like ChatGPT are gaining traction, I don’t yet see them as a significant source of personal traffic. Similarly, platforms like X and LinkedIn feel less aligned with my approach, offering limited potential for the type of engagement I value. My energy remains concentrated on ecosystems that deliver tangible results and align with my vision. ## Looking Ahead Social media is constantly evolving, and as platforms crumble or pivot, so must we. But one thing stays true: at the heart of any strategy is the simple act of connecting with others. That’s what drew me to social media in the first place, and it’s what will keep me here. With “[robhoeij](https://rob-hoeijmakers.ghost.io/),” I’m ready to create something cohesive, meaningful, and maybe even a little rebellious. A personal brand that doesn’t just follow the algorithm but works with it—and around it. 📹 One more thought, in 2025 I will work more with video. Ideally prompted from text. As a reader I am not a fan of video, but the signs are clear more and more information is sent through video and audio. ## P.s. Peak Social Media passed, 8th January On the first week of January Meta announced that they will do a major change to their platforms. It shows that we are renting space and the house lord can change the rules on a whim. 🫨 First week of January and Meta announces major change of direction: [More Speech and Fewer Mistakes](https://about.fb.com/news/2025/01/meta-more-speech-fewer-mistakes/) I was annoyed until I read this message from futurist Sinéad Bovell who explains we passed peak social media a while ago and I realised I am busy **to preserve and conserve a gone by era**. So I now decided to let go, maintain but put energy in new forms of expression and connection, even though it is not clear what that will be. But I am confident it will be emerging in 2025. --- ### Further reading - [Sinéad Bovell (@sineadbovell) on Threads](https://www.threads.net/@sineadbovell/post/DEiZkJKThJF?xmt=AQGzB2flgUA7i-eUHmSVCCt6Wae1VpCXrz20qVUyILBVUA) ### Agents and Agency URL: https://rob-hoeijmakers.ghost.io/agents-and-agency/ Last updated: 2025-05-13T07:24:39.000Z It all started with a discussion. Someone raised the question: "Isn’t ‘AI agent’ just a fancy marketing term for chatbot?" It seemed a simple enough point at first. After all, chatbots are everywhere—they answer our queries, help with customer support, and manage appointments. But as the conversation unfolded, I found myself digging deeper, questioning what we really mean by “agent” and “agency” in the digital sense. At the same time, I stumbled upon a fascinating example: a team of AI agents running a hedge fund. This wasn’t a simple chatbot responding to questions—it was a system of interconnected AI agents, each with its own specialised role, working together to analyse market data, assess risks, and make trading decisions. It was a vivid example of how far we’ve come from the static chatbots of just a few years ago. That got me thinking: what’s really happening here? Are we witnessing the birth of something fundamentally new? And what does this mean for how we work, delegate, and lead in a digital world? ## The AI Agent: A New Kind of Worker? The term "agent" implies more than just a tool. It carries the idea of autonomy—a system that can act on your behalf, making decisions and taking action without constant supervision. But unlike humans, these agents are powered by [Large Language Models (LLMs)](https://rob-hoeijmakers.ghost.io/a-comparative-analysis-of-top-language-models/), which give them the ability to understand complex instructions, maintain context, and even collaborate with other agents. This is where the distinction between **chatbots** and **AI agents** becomes clear: - A chatbot is reactive, responding to predefined inputs. - An AI agent is proactive, capable of initiating actions and collaborating with other systems. When LLMs entered the scene, they added a layer of intelligence that chatbots could never achieve. These systems can now: - **Interpret nuanced instructions**: Moving beyond rigid scripts to understand complex and ambiguous human language. - **Access external tools**: Connecting to APIs to retrieve data, make calculations, or complete tasks. - **Make decisions**: Based on probabilistic reasoning, they can suggest or even act on the best course of action. But the hedge fund example I’d seen wasn’t just about one smart agent. It was about **teams of agents**—each with a specialised role—working together to achieve a shared goal. ## A Team of Agents: The Hedge Fund Example Let me take you back to that hedge fund. Here’s how it worked: 1. **Market Data Agent**: The scout. It gathered raw market data and calculated technical signals, preparing the groundwork. 2. **Quant Agent**: The strategist. It analysed the data and decided whether to buy, sell, or hold. 3. **Risk Manager Agent**: The sentinel. It evaluated the trading signal against risk thresholds. 4. **Portfolio Manager Agent**: The leader. It synthesised all the inputs to make the final trading decision. Each agent operated independently but relied on the outputs of the others to perform its role. It was a textbook example of **collaboration**, not just automation. And yet, the system as a whole behaved like a cohesive team, much like human professionals working together to solve a complex problem. This example struck me as more than just clever programming. It felt like a new paradigm for how we could think about AI: not as isolated tools, but as ecosystems of intelligence, capable of scaling human workflows in ways we’ve never seen before. > I made a real-world AI hedge fund team. > > It has 4 agents: > 1 • market data agent > 2 • quant agent > 3 • risk manager agent > 4 • portfolio manager agent > > You can run all of my code below. > > No prior coding experience is required. > > The agents work sequentially, passing messages to… [pic.twitter.com/AMAyrEKJZr](https://t.co/AMAyrEKJZr) > > — virat (@virattt) [November 23, 2024](https://twitter.com/virattt/status/1860443712490274996?ref%5Fsrc=twsrc%5Etfw) ## Agency and Control: Lessons in Leadership The hedge fund story also surfaced deeper questions about **control and delegation**. As humans, we’re used to delegating tasks to others—but we still retain ultimate responsibility. AI agents introduce a similar dynamic: - How much agency are you willing to give away? - How do you maintain control over the system without micromanaging every decision? This reminded me of **situational leadership**: the idea that you adapt your leadership style depending on the maturity of the team. With AI agents, it’s a similar process: 1. **Directing**: When the agent is first implemented, you monitor it closely, providing specific instructions. 2. **Coaching**: As the agent gains proficiency, you guide it with periodic feedback. 3. **Supporting**: You step back, allowing the agent to act autonomously while staying available for advice. 4. **Delegating**: The agent operates independently, requiring only occasional oversight. The difference is that with AI agents, this process is compressed. They learn and adapt much faster than humans, but they still require careful calibration. And the ultimate responsibility—ethical and operational—remains with you. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/12/IMG_9472-Medium.jpeg) Situational leadership, dr. Paul Hersey ## The Bigger Picture: Ecosystems of Intelligence The hedge fund example shows what happens when agents collaborate sequentially. But what if they worked **in parallel**? What if multiple agents tackled different aspects of a problem simultaneously, sharing insights in real time? This is where the concept of **digital ecosystems** emerges. Instead of linear workflows, we can imagine: - **Parallel Collaboration**: Agents working on various facets of a task simultaneously, such as one analysing data, another generating ideas, and a third optimising outcomes. - **Meta-Agents**: A higher-level coordinator (human or AI) that ensures alignment, resolves conflicts, and refines strategies. - **Emergent Properties**: When agents collaborate, the system gains capabilities no single agent could achieve, such as adaptability, robustness, and creativity. This ecosystem model is particularly exciting because it mirrors how humans work in teams—leveraging diverse strengths while aiming for a shared goal. ## My Personal Reflection As I thought more about this, I realised how much this connects to my own work and aspirations. Like many, I want to express myself, collaborate meaningfully, and balance control with delegation. AI agents offer a glimpse of what’s possible: systems that amplify our capabilities without replacing our humanity. But they also challenge us to rethink what it means to lead. Just as the portfolio manager in the hedge fund example synthesises inputs to make decisions, we must act as stewards of these digital ecosystems. We set the vision, guide the process, and ensure that the agents operate in alignment with our goals and values. At the heart of this shift is a question of trust: **How do we trust systems that act autonomously?** For me, the answer lies in periodic reflection and recalibration. Just as you review a team’s performance, you must review your agents—ensuring they continue to serve your intent. ## The New Paradigm: AI as Collaborative Partners We are entering an era where AI agents are not just tools but **co-creators**. They will handle the mundane and the complex, freeing us to focus on what truly matters. But this also demands a shift in mindset: - From controlling to guiding. - From managing tasks to orchestrating systems. - From using tools to collaborating with digital partners. This is not just a technical evolution—it’s a cultural one. And as someone navigating this change, I find it both daunting and exhilarating. The future will belong to those who can balance agency and control, embracing AI not as a threat, but as a partner in creativity, productivity, and leadership. ## Conclusion The story of AI agents running a hedge fund is just one example of how digital agents are evolving beyond simple chatbots. They are becoming part of collaborative, intelligent ecosystems that augment our abilities and reshape our workflows. The real power lies not just in what these agents can do alone, but in how they work together—and with us—to create something greater. The challenge ahead is to find the right balance of delegation, control, and trust. As we embrace AI agents as partners, we must remember that the ultimate responsibility for their actions rests with us. By doing so thoughtfully, we can unlock new possibilities and redefine how we create, lead, and connect in the digital age. *P.s. this has turned out to be a longer article then expected, it mostly means I am still working on more clear insights. And I will dig in deeper. Great to see you made it to here, thank you, and consider* [*to subscribe*](https://rob-hoeijmakers.ghost.io/newsletter/)*.* ### Facebook Protect URL: https://rob-hoeijmakers.ghost.io/facebook-protect/ Last updated: 2026-04-24T18:29:41.000Z The notification arrived: *"Your account requires extra security through Facebook Protect."* This all started when [a photograph of mine went viral for the second time](https://rob-hoeijmakers.ghost.io/viral-photo/), leading to a surge in followers on Instagram. At first, it seemed like a simple formality, but it sparked deeper thoughts about how intertwined our digital identities and security have become—especially within the Meta ecosystem. For those managing public or influential accounts, platforms like Facebook and Instagram are more than social networks; they are digital extensions of our lives. Safeguarding them is essential, and this is where **Facebook Protect** comes into play. ## What is Facebook Protect? Facebook Protect is a security program for accounts that Meta sees as high-risk, like those of journalists, public figures, or users with large followings. Its primary goal is to shield these accounts from hacking attempts by requiring [**two-factor authentication (2FA)**](https://rob-hoeijmakers.ghost.io/two-factor-authentication-2fa/). This extra layer ensures that even if someone gets your password, they cannot access your account without a second verification step. In my case, I already had 2FA enabled, so onboarding into Facebook Protect was straightforward. It was more about confirming my account met Meta's security standards rather than setting up new protections. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/12/Facebook-Protect-E-mail.jpg) Facebook Protect E-mail ## The Simplicity of Compliance For those already taking security seriously, Facebook Protect feels like a nudge toward good habits. The onboarding process includes: - Confirming 2FA is enabled. - Reviewing recovery options. - Checking for old, insecure third-party apps. These steps were quick for me, serving as a reminder that small actions, like enabling 2FA, make a big difference in securing our presence online. ## Identity, Security, and Meta's Ecosystem What struck me most was the underlying message: *Your digital identity is valuable, and protecting it is a shared responsibility.* Meta’s push for security highlights how our lives converge in the digital space. For someone like me, with a growing Instagram following, this shift feels significant (also see: [when locked out of this digital space](https://rob-hoeijmakers.ghost.io/locked-out-banned-instagram-threads/)). My Instagram presence is growing steadily, and even though I log in separately to Facebook and Instagram with different credentials, they are linked in **Meta's Account Centre**. This means the platforms are interconnected, providing overlapping security features even if I access them independently. Accounts are not just personal—they are platforms for influence and connection. Facebook Protect extends seamlessly to Instagram, even if there is no equivalent program by name. Due to the connection between my accounts in Meta's Account Centre, protections apply across both platforms, reflecting Meta’s commitment to unified digital security. Instagram uses similar safeguards, like 2FA and suspicious activity alerts. If accounts are linked, protections overlap, reinforcing Meta's vision of a connected digital identity. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/12/Facebook-Advanced-Protection.jpg) The Facebook Advanced Protection flow ## A Moment for Reflection Enabling **Facebook Protect** made me realise how much we take for granted in our digital lives. Platforms like Facebook and Instagram are so embedded in our routines that we rarely think about the security measures behind them. Practically speaking, it seems that enabling two-factor authentication (2FA) is key for Meta. Personally, I prefer using [an authenticator generator on macOS and iOS](https://rob-hoeijmakers.ghost.io/mac-and-ios-password-manager/), which makes the process easy and secure. Beyond the practicalities, there was also an emotional side to enabling Facebook Protect that surprised me. I must admit, the experience made me feel a bit vain—as if the need for extra security was a kind of recognition of my online presence. It felt like Meta was acknowledging that my account was important enough to protect, which brought with it a small sense of validation. However, it's important to remember that this sense of recognition is ultimately misplaced. I am, after all, [just a tenant on this platform](https://rob-hoeijmakers.ghost.io/bluesky/)—subject to its rules and changes. At the same time, it made me wonder: How do you really know when your account is breaking through the clutter and gaining significance? It’s easy to look at follower counts or engagement metrics, but that doesn't always reflect real impact. Being prompted to use Facebook Protect was, in a strange way, a signal that my digital identity had reached a new level—one worth protecting. This realisation was both affirming and humbling, reminding me of the growing responsibilities tied to an online presence that reaches more people. This analysis also made me realise I could do much more with my photos on Facebook. To leverage this potential, I plan to set up a public page to showcase my work while keeping my personal account private. To be continued ... ### The Power and Popularity of Podcasting URL: https://rob-hoeijmakers.ghost.io/power-and-popularity-of-podcasting/ Last updated: 2026-04-24T18:35:30.000Z Podcasting has grown from a niche format to one of the most influential mediums of storytelling and communication. Last week, I had the opportunity to explore this at a podcasting conference held at the [**National Holocaust Museum in Amsterdam**](https://jck.nl/locatie/nationaal-holocaustmuseum). The event, aimed at professionals in musea, was impeccably organised, and the venue—a beautiful museum with an excellent auditorium—provided a meaningful backdrop for the day. I discovered the conference through [Marvin Jacobs, whom I met personally](https://rob-hoeijmakers.ghost.io/ai-is-transforming-podcast-production/), and his newsletter sparked my curiosity. Attending with two friends, I stepped into a medium I don’t naturally gravitate toward, and it turned out to be a truly insightful experience. While I’m more comfortable with written formats, the day gave me a new perspective on why podcasts resonate so deeply. From the intimate connection fostered by earbuds to the serial nature of episodic storytelling, podcasting revealed itself as an exciting and multifaceted medium. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/12/0C75B120-339E-497B-849F-7F35164749BF_1_201_a.jpg) The day also included a panel discussion. ## Inside the Layers of Storytelling Podcasts One of the highlights of the day was Marvin Jacobs’ deep dive into how his firm, [Audio Agency Airborne](https://www.airborne.audio/en), creates storytelling podcasts. He walked us through every step of the process, from concept to production, and revealed the intricate layers that bring these podcasts to life: - **Recorded Interviews:** Authentic voices and stories form the foundation of a compelling narrative. - **Leader Music:** A recognisable intro sets the tone and brand of the podcast. - **The Host:** The anchor who guides the listener through the story. - **Sound Design and Soundtrack:** Carefully crafted to create atmosphere, evoke emotion, and add depth to the narrative. These elements make storytelling podcasts distinct from other styles. For example, “chat casts” (two personalities talking conversationally), interview-style podcasts (think Joe Rogan), single-person explorations (audio essays), or dramatic documentaries all have unique formats and approaches. Marvin’s insights highlighted how storytelling podcasts stand out by weaving multiple layers into a cohesive and immersive experience. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/12/017825FC-E38D-4DDD-B06E-2D78BEB82B42_1_102.jpeg) Marvin explains and shows the layers in storytelling podcast ## Podcasting’s Future: Blurring the Line with Video As podcasting continues to evolve, the line between audio and video is becoming increasingly blurred. Short video snippets from podcasts, often accompanied by captions and quotes, are transforming how podcast content spreads on social media. These visual elements extend the reach of podcasts, making them more shareable and engaging for wider audiences. This development raises intriguing questions about the future of podcasting. Will audio-first remain the norm, or will video elements become a standard part of the format? These are topics I hope to explore further, potentially through a conversation with [**Professor Martin Spinelli**](https://profiles.sussex.ac.uk/p82794-martin-spinelli), whose insights and workshop at the conference were particularly inspiring. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/12/Podcast-snippet.jpg) Example of a podcast video snippet on X ## A Reflection on Podcasting’s Role in Communication For me, attending this event was a chance to step out of my comfort zone and engage with a medium I’ve often overlooked. **Podcasts offer a unique blend of intimacy, creativity, and narrative depth that can teach us valuable lessons about connection and engagement.** Listening is often an intimate experience, and the podcasts—crafted in a sober, direct, and yet lively way—become emotional experiences rather than intellectual ones. While I plan to dive deeper into the topics of video integration (vodcasts) and shareable quotes (Social Snippets), this day was a reminder that even as formats evolve, the power of **storytelling** remains constant. Whether it’s through words, sounds, or visuals, the goal is always the same: to connect with people in meaningful ways. *If you liked this article, make sure to* [*subscribe for updates*](https://rob-hoeijmakers.ghost.io/newsletter/)*.* ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/12/IMG_9350.jpeg) As an afterthought, I noticed again the benefits of a smaller symposium. Learned so much. ### Meta Ads Payments: Credit Cards, Invoicing, and a Crash Course in Transaction Types URL: https://rob-hoeijmakers.ghost.io/meta-ads-payments-credit-cards-invoicing/ Last updated: 2026-05-10T08:54:09.000Z When I launched an Instagram ad campaign, I expected creativity and strategy to be my main focus—not figuring out the ins and outs of how Meta Ads processes payments. Instead, I found myself learning a lot more about how credit card transactions work, the quirks of Meta's payment system, and even how different kinds of holds or authorisations show up on my account. Here’s what I learned and how it might help you if you're running ads or just want to understand credit card transactions better. ## The Start: Choosing a Payment Method for Meta Ads When starting my campaign, I opted to pay via **credit card**. Meta, now operating Instagram and Facebook advertising, offers several payment methods, including credit cards, bank transfers, and direct debits. I chose credit card for the flexibility, assuming it would be as simple as making a regular online purchase. Little did I know that Meta’s system isn’t quite that straightforward. While it is efficient and trustworthy, the way they handle **authorisations**, **reservations**, and **final payments** involves multiple steps that can be confusing at first. ## Micro-Reservations: The Surprise of Small Holds Once my ad campaign was live, I quickly noticed small, recurring **€2 reservations** on my credit card. Even though I had set a payment threshold of €100, Meta didn’t immediately take that amount. Instead, they started by placing small holds—€2 here, €5 there—as if testing the waters before incrementing. Meta actually **captured** these small amounts, meaning the funds were formally taken from my account. After capturing, it took a while before my bank updated the status of these transactions from reserved to paid. Each of these authorisations came with an email notification, giving me the sense of being constantly bombarded by payment updates, even though no actual money was leaving my account. It turns out that this process is Meta’s way of ensuring everything works smoothly before they allow you to build up larger balances. But as a new advertiser, it did feel a bit chaotic at first. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/11/Meta-Ads-payment-threshold-and-billing-date.jpg) Meta Ads: Payment threshold and monthly billing date ## Authorisations vs Reservations: What Are They? This experience with Meta taught me a lot about the difference between **authorisations** and **reservations**. An **authorisation** is the first step: it’s essentially Meta checking whether the money is available. It doesn’t move the funds yet, but it places a **hold** on them—reducing my available credit. A **reservation**, in this context, is what happens after the authorisation. It ensures that the money is still set aside for Meta and cannot be spent elsewhere, but it hasn’t formally left my account. This meant I had less money available, even though the amount wasn't officially deducted. And unlike a typical online purchase, there was no immediate transfer to Meta. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/11/Meta-Pay-Ads.jpg) Lost of small payments and reservations ## Capturing the Payment: When Does the Money Leave? Eventually, after accumulating spending to meet my €100 threshold, Meta **captured** the payment. This is when the reserved amount formally leaves your account, becoming an actual payment. However, this didn’t appear in my bank account immediately. There was a delay—sometimes a day or more—between Meta processing the payment and it showing up on my credit card statement. This lag can make it feel like you’re unsure whether the payment has been processed or not. Understanding this process made it clear why hotels often tell guests, “It’s just an authorisation, no payment necessary.” The same concept applies here—the money isn’t gone, just held until the transaction completes or is released. [Payment requests: Your digital payment assistantA brief overview of how to properly deploy and manage payment requests. A powerful addition to your digital routines.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Logo-Rob-Hoeijmakers-Square-Dark-Glow-3-1-4.png)Web StrategiesRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_6560-1-1536x1152.jpeg)](https://rob-hoeijmakers.ghost.io/payment-requests-your-digital-payment-assistant/) ## Invoicing with Meta: The Monthly Recap Meta’s payment system starts with a low threshold, resulting in frequent small charges. Each payment generates an invoice, an email notification, and a status update in Meta marking it as 'paid'. As you continue to advertise, the threshold gradually increases. This means you’ll see fewer individual payments and invoices over time. Eventually, if the threshold surpasses your total monthly ad spend, Meta consolidates the charges and bills you only once per month. This single monthly invoice summarises all ad expenses during the billing cycle, providing a more streamlined record for tax reporting and bookkeeping. While the system transitions from frequent payments to a consolidated monthly charge, it’s essential to track individual transactions early on. The initial phase, with its numerous small payments, offers detailed visibility into your spending. Later, the consolidated invoices make bookkeeping simpler, provided you’ve kept a record of earlier transactions. This approach works well once you understand how the system evolves, but initially, the frequent charges and low thresholds can be a bit confusing. Over time, however, the system becomes more efficient, especially for advertisers with consistent or lower ad spend. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/11/2024-11-26_13-43-59.jpg) Small charges, all have a transaction and an invoice. ## Key Takeaways for Advertisers 1. **Meta Ads Start Small**: Meta’s initial use of small authorisations is a way to build trust and ensure funds are available, but it can feel overwhelming with all the notifications. 2. **Understand Authorisations and Reservations**: These are not final payments, but they do reduce your available balance, which can be confusing if you’re not expecting it. 3. **Payment Capture Lags**: Even after Meta decides to capture the funds, it can take time to appear on your bank statement. Don’t panic if there’s a delay. 4. **Monthly Invoices Simplify, but Track Closely**: The monthly invoice is convenient, but it means you need to be organised throughout the month to match payments accurately. ## Final Thoughts Using a credit card for Meta Ads taught me a lot more about payment systems than I expected. Their approach, which includes frequent small authorisations, delayed captures, and monthly consolidated invoicing, is efficient but not intuitive at first. Understanding how different transaction types—like authorisations, reservations, and deposits—work can help you better manage your finances, especially if you’re running campaigns or dealing with high-value transactions. If you’ve had similar experiences or felt confused by Meta’s payment process, I hope this overview helps you navigate it more confidently. Feel free to share your thoughts or questions—I’d love to hear about your experiences too! 💰 Small transactions, like the frequent €2 reservations from Meta Pay, can quickly eat into the monthly transaction limit of bookkeeping software like [MoneyBird](https://moneybird.com). This can lead to exceeding your plan's capacity, causing unnecessary costs or disruptions. The other thing I noted is that even when the bank says it is a reservation, the amount will appear in the bookkeeping software. ### Further reading - [Understanding the layers behind digital payments](https://rob-hoeijmakers.ghost.io/understanding-digital-payments-layers/) - [How Money Moves](https://rob-hoeijmakers.ghost.io/how-money-moves/) ### The Power of Sequences URL: https://rob-hoeijmakers.ghost.io/the-power-of-sequences/ Last updated: 2026-04-25T13:17:12.000Z I’ve been diving into AI and large language models (LLMs) for a while, and I keep coming back to this big question: what is language? On the surface, it seems simple, but it has forced me to dig deep into the way written words work—for us, as humans, and for the technology that now processes those words. I’m not an expert explaining how things are; instead, I’m trying to grasp something fundamental here. And that’s why this exploration is so important to me. It’s about seeing beyond the buzzwords and understanding what really drives these systems. It all started when I came across a tweet by [Richard Socher](https://t.co/XGzuiLr1Em). He pointed out that **the name "Large Language Model" is actually a misnomer**. That struck me as a huge insight. These models aren’t just about language. They are better described as **Large Neural Sequence Models**—a term that reveals much more about what they actually do. They don’t only work with words; they handle any kind of sequential data, from natural language to biological sequences, musical notes, or even images. That realisation created a mental shift in me—one that felt almost seismic. > "Large language model" is a bit of a misnomer. > > I call them large neural sequence models, applicable to any domain with abundant sequential data: > • Natural language > • Programming languages > • Biological sequences (proteins) > • Pixel sequences > > Current LLMs are just the… [pic.twitter.com/OyD29VK8t0](https://t.co/OyD29VK8t0) > > — Richard Socher (@RichardSocher) [November 24, 2024](https://twitter.com/RichardSocher/status/1860745433758527747?ref%5Fsrc=twsrc%5Etfw) ## Sequences: The Heart of It All At the core of it all lies this idea of a sequence: an ordered set of elements that carry meaning. **Language is inherently sequential because it unfolds over time, whether spoken or written.** It’s surprising how this concept spans so many different domains. Think about language—words and letters need to be in the right order for a sentence to make sense. If I write, “the cat chases the mouse,” you get a clear picture. But if I switch it to, “the mouse chases the cat,” suddenly everything changes. The sequence is what defines the meaning. The same applies to biology. DNA and proteins are just sequences of molecules, but the order of these molecules is what defines their structure and function. It’s like a code for life itself, where shifting a single component can completely change the outcome. Music, too, is fundamentally sequential. A melody unfolds note by note, and the specific order of those notes creates the mood, the story, the emotion. Even images can be thought of as sequences. At first, this idea seemed a bit abstract to me—after all, pictures are static, right? But if you think about an image as a grid of pixels, each pixel has a specific place, a particular value that contributes to the whole. Videos go one step further—frames arranged over time, creating movement. It’s all sequences, and the relationships within those sequences are what give them meaning. ## Understanding Large Neural Sequence Models The key to these models is that they are designed to understand these sequences. They aren’t locked into just processing language; they can take in any kind of ordered data. The architecture they use, often involving transformers, allows them to pick up on patterns and relationships. They learn that “the cat chases the mouse” means something specific about the roles of “cat” and “mouse,” and they can adapt this understanding to other contexts—whether that’s predicting the next note in a song, designing a protein, or generating a coherent image. The magic of these models lies in their ability to abstract the idea of a sequence. They don’t come pre-programmed with rules about language or music or biology. Instead, they learn directly from the data itself. They figure out that there’s a kind of syntax to DNA, that musical notes have a rhythm, that pixels have a structure. And the more I think about this, the more I see the power in treating all these different forms of data as sequences. ## A Shift in Perspective Thinking about these models as “sequence specialists” rather than “language specialists” has changed how I understand them. It made me realise why the same architecture can be used to create music, generate images, or even make sense of biological data. It’s all about sequences, and that’s where their true power lies. By refining this one fundamental approach, researchers have opened up so many possibilities—designing proteins, composing original music, generating lifelike images, and more. For me, this isn’t just an abstract concept; it’s a personal revelation. It deepens my understanding of language, especially the written form, and helps me see AI in a different light. These models aren’t just tools for language—they are windows into how meaning is built from sequences, in everything from a sentence to a melody to an image. ### Wrap Up: The Key, the Essence This journey has been about more than just understanding AI models; it's been about uncovering the deeper significance of sequences. Realising that language is just one type of sequential data—and that these models can handle so many other forms—changed my perspective entirely. The insight, because of the reframing, made the models more tangible in my mind. They feel less abstract, less like a black box, and more like something I can relate to on a deeper, even physical level. It’s not just an intellectual understanding but something that I can almost feel in my body—an abstraction that now feels more real and approachable. By reframing LLMs as Large Neural Sequence Models, I now see their true capacity: they aren’t just about processing words but about understanding and predicting sequences of any kind. This insight has made me appreciate how flexible and powerful these models are. They aren’t limited to human language; they are capable of abstracting and finding meaning in anything that unfolds over time or through ordered elements. The shift in my thinking has been profound. It's shown me that the power of AI isn't just about language—it’s about recognising and building meaning from sequences, whatever form they take. That’s the core insight I'll carry forward: it's all about the sequences. ### A Thousand Compresses: How a Video Slowly Disappears URL: https://rob-hoeijmakers.ghost.io/a-thousand-compresses-how-video-slowly-disappears/ Last updated: 2026-04-25T13:16:40.000Z Sometimes, a simple demonstration cuts through layers of technical jargon to reveal the essence of a concept. This happened to me recently with video compression. It's a topic I've [explored before](https://rob-hoeijmakers.ghost.io/audio-encoding/), but let’s be honest—talking about codecs, encoding, and decoding can feel a bit abstract. It’s a world of technical terms, hardware versus software debates, and questions like, “Which codec should you use?” But recently, I came across a brilliant demonstration that brought it all into focus. ## The Thousand Upload Experiment Marques Brownlee did something that caught my attention. He uploaded and downloaded a video with audio to YouTube Shorts and TikTok—*not once, not twice, but a thousand times*. The result? A fascinating visual representation of compression in action. Each time the video was processed, there was some lossy quality loss, but it wasn’t noticeable at first. After four or five rounds, you could still recognise the original. But after a thousand cycles? The result was hilariously unrecognisable—almost entirely purple, a glitchy ghost of the original content. > I downloaded and uploaded to YouTube Shorts AND TikTok 1000 times each to see how bad processing would get. Fascinating results 👀 [pic.twitter.com/Qubo6A6GfN](https://t.co/Qubo6A6GfN) > > — Marques Brownlee (@MKBHD) [November 2, 2024](https://twitter.com/MKBHD/status/1852750641031950502?ref%5Fsrc=twsrc%5Etfw) It was both funny and enlightening, like watching the life cycle of your favourite T-shirt. When you wash clothes a few times, the wear and tear are imperceptible. But after years of use, the colours fade, the fabric thins, and the item eventually becomes unwearable. Video compression follows a similar principle: the lossy compression losses accumulate, and the quality degrades with repeated cycles. The same effect can be seen with [JPEG images](https://rob-hoeijmakers.ghost.io/webp-versus-jpg/)—each time you re-save, they lose quality due to cumulative compression, eventually leading to noticeable degradation. 👕 ****Technical Callout**: YouTube Shorts primarily uses VP9 and H.264 codecs, while TikTok relies on H.264\. Both platforms apply lossy compression to optimise streaming, sacrificing quality for efficiency. With each re-encoding, visible degradation occurs as data is discarded, resulting in a progressive decline in video fidelity. ### Project: AI Chatbot for RSV Awareness through Web and WhatsApp URL: https://rob-hoeijmakers.ghost.io/awareness-via-chatbot-collaboration-with-resvinet/ Last updated: 2025-12-17T18:27:56.000Z ## A Partnership with ReSViNET Over the past few months I have been working on a project that has been both special and challenging: developing a chatbot for [ReSViNET](https://resvinet.org/). For those who are not familiar with this organisation: ReSViNET is committed to reducing the impact of the Respiratory Syncytial Virus (RSV). This virus causes severe pneumonia and can be fatal, especially for vulnerable groups such as infants, the elderly and people with underlying conditions. The impact can be enormous, especially in areas where medical care is less advanced. ## The Origin of the Project and Technical Innovations ReSViNET approached me with the question if I could develop an AI-driven chatbot that would provide education about RSV. The goal was to inform users about infection prevention, the latest developments in treatments and upcoming vaccines. We decided to collaborate with doctors from the UMC Utrecht for this, which resulted in an extensive knowledge base. In addition, we use the technology and expertise of [Enreach](https://www.chatvoorbedrijven.nl/ms-teams-whatsapp-dialox/) to realise the chatbot and the linked channels. What made this project really special was the use of [**Retrieval Augmented Generation (RAG)**](https://rob-hoeijmakers.ghost.io/exploring-retrieval-augmented-generation-rag/) . This means that we don’t just rely on AI models, but also carefully extract information from our knowledge base to provide users with accurate and detailed answers. When we started, we were still working with GPT-3.5, but we have since upgraded to GPT-4o mini, which resulted in even better results. ### Multilingual chatbot One of the unique aspects of the chatbot is that it is available in multiple languages, thanks to the language capabilities of OpenAI’s models. This allows us to truly serve a global audience. This is where AI is particularly powerful: until now, it has certainly not been possible to provide personalized information at this scale in so many languages. It is a unique opportunity for companies and organizations that operate globally. ## Multichannel Approach: From Web to WhatsApp In order to reach as many people as possible, we decided to make the chatbot available not only through the ReSViNET website, but also through **WhatsApp** . This turned out to be a smart move, especially in countries where many people have skipped the step to computers and are using mobile devices directly. There are [about 3 billion WhatsApp users worldwide](https://rob-hoeijmakers.ghost.io/whatsapps-global-reach/), which gives us a huge reach. A great thing about WhatsApp is that users can continue their conversations. Imagine someone asks a question, gets an answer, and then comes back a week later with a follow-up question—the chatbot remembers the conversation. Especially when people are confronted with the disease, for example with their child, the world is in disarray and there are many concerns but also developments. It is precisely at such moments that there is a need for a developed conversation that can be approached practically. This creates an ongoing dialogue that adds a lot of value to the user experience. ![RSV Chatbot on mobile: the website, the chatbot page and WhatsApp.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/11/RSV-Chatbot---Mobile-flow-1.jpg) RSV Chatbot on mobile: the website, the chatbot page and WhatsApp. ### QR-codes In addition, we consciously use [**QR codes**](https://rob-hoeijmakers.ghost.io/qr-code-explained/) to facilitate the transition from the physical to the digital world. Imagine a hospital in Indonesia hanging up a QR code that allows patients to immediately start a conversation with the chatbot to get answers to their questions. This not only makes access to information accessible, but also easy to distribute. ## The Learning Chatbot Although the chatbot can already answer many questions, this remains a continuous learning process. We monitor the interactions to see where the chatbot falls short and add new information when necessary. We also expect developments regarding the availability of vaccines and expect questions about this. The knowledge base offers the possibility to be up-to-date, actually much more than the general [AI language models](https://rob-hoeijmakers.ghost.io/a-comparative-analysis-of-top-language-models/) . The great thing is that we don't know exactly what questions users will ask in the future, but by continuously learning and improving, we can increasingly respond to their needs. ## Experience It Yourself! Are you curious? I invite you to have a conversation with the chatbot yourself via the ReSViNET website or via WhatsApp. It is a unique opportunity to experience how AI and human insight come together to make a difference in healthcare. [RSV Chatbot - ReSViNETDiscover the RSV chatbot for instant support on respiratory syncytial virus. It’s your go-to resource for understanding symptoms, prevention tips, and available treatments, all through a simple, user-friendly chat interface.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/cropped-Favicon-RESVINET-270x270.png)ReSViNETRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/Aankondiging-Chatbot-ReSViNET-3.png)](https://resvinet.org/patient-network/chatbot/) *I will write more about this in the coming months and keep you updated on the effect.* [*Sign up for the newsletter*](https://rob-hoeijmakers.ghost.io/#/portal/signup) *to stay updated.* ### From Typing to Talking URL: https://rob-hoeijmakers.ghost.io/from-typing-to-talking/ Last updated: 2026-05-26T19:32:09.000Z For as long as I’ve been fascinated by the digital realm, the intersection of language and technology has intrigued me. The digital world thrives on words—whether spoken or written. Yet, speaking predates writing by millennia, shaping human interaction long before the written word took hold. This led me to explore how information is structured, especially as we see a resurgence of oral culture with advancements in AI technologies like [Automatic Speech recognition and text-to-speech (TtS)](https://rob-hoeijmakers.ghost.io/asr-speech-to-text-dictation-and-transcription-whats-the-difference/). In my exploration, I’ve found that while text remains compact, easier to store, and more searchable, spoken language carries nuances—intonation, emotion, and even the subtleties of body language—that text often loses. This shift back towards an oral mode of communication, facilitated by AI, intrigued me, especially since it blurs the lines between speaking and writing. ## The Return to Oral Culture Through AI With tools like ASR and TtS, we’re witnessing a partial return to an oral culture in our digital spaces. I wanted to understand how I could seamlessly switch between the two modes—speech and text. This investigation turned into a fascinating journey. I’ve come to realise that dictation, while faster than typing, works best when you have a clear idea of what you want to say. If your thoughts are more fluid or exploratory, recording and then transcribing might still be the better approach. ## The Tools Behind My Workflow I’ve experimented with various tools to find what works best for my needs: 1. **Transcription**: I use three methods for transcription: local, cloud-based, and for video calls. Locally, I use [OpenAI’s Whisper](https://rob-hoeijmakers.ghost.io/openais-whisper-speech-to-text-technology/), which keeps everything on my device, ensuring privacy and control over my data. For cloud-based transcription, I use an application called [Transcribe](https://rob-hoeijmakers.ghost.io/transcribe-audio-on-macos-and-ios/), which is cross-device and offers flexibility in accessing my transcriptions wherever I need them. For video calls, I use [Fireflies](https://rob-hoeijmakers.ghost.io/fireflies-ai-transcriptions/), which automatically records and transcribes my calls, including speaker tracking—a feature that differentiates who said what. This combination has been incredibly useful for projects requiring detailed dialogue analysis. 2. **Speech Synthesis**: Beyond transcription, I’ve delved into the world of speech synthesis. I even created a [clone of my voice](https://rob-hoeijmakers.ghost.io/my-experiment-with-voice-cloning/), allowing me to generate spoken content that feels more personal and connected to my style. This is particularly helpful for projects where I need to update content regularly. For instance, in screencast videos, where the visuals change frequently, using TtS to sync updated narration with new footage is much more efficient than re-recording everything. ![Spoken input work flow in short: Record Voice Memo, transcribe with Transcribe, send text to ChatGPT.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/11/Free-flow-audio-work-stream.jpg) Spoken input work flow in short: Record Voice Memo, transcribe with Transcribe, send text to ChatGPT. ## Dictation: Efficiency and Built-In Features One of the biggest revelations for me was the sheer efficiency of [dictation](https://rob-hoeijmakers.ghost.io/asr-speech-to-text-dictation-and-transcription-whats-the-difference/). I’ve found that dictation is at least three times faster than typing, especially for longer texts. While typing can help with concentration, dictation feels more natural and fluid when capturing initial thoughts. Although dictation has limitations—like not being able to work in complete silence—it benefits from improved local computing power, which also makes it more private. Ironically, typing on a phone, despite being tactile, often leads to physical strain, whereas dictation is more comfortable and efficient in many situations. I’ve also embraced the convenience of OS-level transcription features on iOS and macOS. Having StT functionality integrated directly into the system means I can access it anywhere, whether I’m working on my laptop or my phone. This kind of consistency is a game-changer. Although I continue to experiment with platforms like [ChatGPT’s](https://rob-hoeijmakers.ghost.io/chatgpt-new-voice-features/) transcription capabilities, I appreciate the flexibility of being able to dictate wherever inspiration strikes. [ASR, Speech-to-Text, Dictation, and Transcription: What’s the Difference?Just like picking the right writing tool, choosing the right speech technology makes all the difference—here’s how to decide.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-36.png)Rob Hoeijmakers (robhoeij)Rob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_4550.jpeg)](https://rob-hoeijmakers.ghost.io/asr-speech-to-text-dictation-and-transcription-whats-the-difference/) ### Navigating Voice Commands and Interaction Beyond transcription and synthesis, I’ve been exploring voice commands on both desktop and mobile devices, using tools like Siri and VoiceOver. These experiments are still in their early stages, but I’m convinced that as AI continues to evolve, the way we interact with our devices will shift dramatically. We’re moving away from traditional Graphical User Interfaces (GUIs) towards more natural, voice-driven Application Programming Interfaces (APIs). I plan to delve deeper into this in future posts. ## A New Way of Working Ultimately, my journey into StT and TtS has taught me that these technologies aren’t just about efficiency; they represent a new way of working, one that’s more in tune with the natural flow of human communication. By blending spoken and written modalities, we can communicate more effectively, streamline our workflows, and adapt to the evolving digital landscape. I’m still learning, still practising, but I see a future where speech and text coexist seamlessly, allowing us to focus less on the tools themselves and more on the messages we want to share. [Subscribe to be updated.](https://rob-hoeijmakers.ghost.io/newsletter/) --- ### Further reading - [The Silent Revolution: How AI is Transforming Speech Recognition](https://rob-hoeijmakers.ghost.io/speech-recognition/) 💬 One of the challenges with dictation software is its inconsistent handling of ****proper nouns** like company names, city names, and personal names. These systems often misinterpret or misspell unique names, leading to extra corrections. Surprisingly, most dictation tools lack a personalised dictionary feature where users could store frequently used names for more accurate recognition. Another issue arises with ****multilingual users**: while some software attempts to detect the spoken language automatically, it’s not always reliable. This can lead to frustration when switching between languages, as the system might default to the wrong one. A truly seamless experience would involve automatic language recognition on a sentence-by-sentence basis, or at least a faster way to switch languages on the fly. ### How AI is Transforming Podcast Production: A Personal Take URL: https://rob-hoeijmakers.ghost.io/ai-is-transforming-podcast-production/ Last updated: 2026-04-25T13:50:05.000Z Imagine you no longer need to painstakingly write minutes or take notes during a meeting. Instead, every word is [automatically captured, and key points are summarized for you](https://rob-hoeijmakers.ghost.io/review-of-transcription-software-tactiq-vs-firefliesai/). That’s part of my workflow now, thanks to AI, and it’s also a major transformation happening in podcast production. Recently, I had a conversation with Marvin Jacobs from Airborne where we discussed exactly how AI tools are reshaping this space. [![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/11/2024-11-10_15-13-07.jpg)](https://www.airborne.audio/en) Audio Agence Airborne ## From Chaos to Story: The Workflow Let’s start with how a typical podcast episode comes to life. The process begins with collecting audio material—interviews, sound bites, and ambient recordings bases on an outline. Often, as Marvin explained, they’ll record interviews that may last up to two hours, with multiple guests like researchers or experts, depending on the theme of the episode. Imagine you have hours of content from three different speakers, and your goal is to condense all of that into a cohesive half-hour episode. This is where AI steps in. Once all the raw audio is recorded, it’s transcribed using tools like [**Whisper**](https://rob-hoeijmakers.ghost.io/openais-whisper-speech-to-text-technology/) or [**AmberScript**](https://www.amberscript.com/en/). These AI transcription services convert the entire conversation into text, complete with time codes. While these tools make it easy to refine and improve the transcript—since errors are often present—the underlying audio remains unchanged. Marvin put it nicely: “It’s actually a text editor that syncs with the audio.” From there, the team refines the outline into a more detailed script, which forms the backbone of the final product, defining the structure and flow of the episode. 🔤 ****Text-Based Editing** For true text-based audio editing, tools like [Descript](https://rob-hoeijmakers.ghost.io/descript/) are used, where changes to the transcript directly alter the audio, offering a seamless way to refine both text and sound. This makes it easy to decide which parts of a conversation to keep, which to cut, and how to refine the material—all while having the option to listen back to the exact point in the audio to ensure nothing important is lost. Airborne employs this method periodically. I've also experimented with [Detail.io](https://detail.co/duo/text-based-video-editing), which offers a similar text-driven approach to video editing, making complex edits more intuitive and efficient. ## Extractive vs. Abstractive Summaries During our chat, we touched on something that’s fundamental to working with AI in creative processes—the difference between extractive and abstractive summaries. I explained it to Marvin like this: *An extractive summary is like what we learned in primary school, where you highlight what’s important—essentially a quote pulled from the material.* Meanwhile, an abstractive summary is about rephrasing and synthesising meaning, capturing the core ideas and transforming them into something new. Marvin and his team use AI for both. They create extractive summaries to get a shortlist of key points, almost like a puzzle piece collection, which then allows them to decide on a narrative. But the real creative part comes when they start working on the *abstractive summaries*. They ask, *“What’s the story we’re trying to tell?”* and *“What’s the message for our audience?”* AI helps to pull out highlights, but it’s the human touch that brings the real meaning and flow to the podcast. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/11/1-SZsIV6bLqCHLXgWw5in7bA.webp) Extractive versus Abstractive Summaries. Image source: [Abstractive Health](https://www.abstractivehealth.com/?) ## Adding the Personal Touch: Voice and Emotion What stood out for me during our discussion was how even in a highly tech-driven workflow, the human element remains irreplaceable. For example, Marvin mentioned the importance of having a recognisable voice tied to the podcast: *“We usually choose hosts who speak to the imagination, someone with a genuine connection to the topic,”* he said. AI might help speed up production, but there’s still something fundamentally intimate about knowing who you’re listening to. It’s not just about summarising and editing. AI is also used for other parts of production—like generating show notes or even enhancing sound effects and creating an immersive atmosphere. Marvin gave an example of [creating podcasts for KLM](https://www.airborne.audio/en/work/the-journey) where they used sound design to enrich their storytelling, evoking the sound of Buenos Aires or the crunching snow of Alaska. Although this project predated AI, it illustrates the value of immersive sound design. [Airborne - Case The JourneyPodcast for KLM Royal Dutch Airlines. Unforgettable stories from travelers about ‘the trip that changed everything’. Awarded multiple times at home and abroad.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/apple-touch-icon-1.png)The Journey29floris03![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/Social-image-Airborne-1.png)](https://www.airborne.audio/en/work/the-journey) In more recent projects, like '[Dieren laten van zich horen](https://www.airborne.audio/nl/pages/dieren-laten-van-zich-horen)' for Albert Heijn, they used extensive sound effects (SFX) to create a vivid experience. Today, tools like [Wondercraft.ai](https://www.wondercraft.ai) can even generate SFX based on prompts, making it easier and faster to add those layers of immersion. [Audio agency AirborneThrough the ears, into the heart | Audio agency Airborne is de meest ervaren en bekroonde podcastspecialist in Nederland. Met aansprekende en onderscheidende verhalen laten wij mensen naar merken luisteren.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/apple-touch-icon.png)Dieren laten van zich horen![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/Social-image-Airborne.png)](https://www.airborne.audio/nl/pages/dieren-laten-van-zich-horen) ## AI as a Creative Collaborator My favourite insight from the conversation was this: **AI is not about replacing the human role in podcasting; it’s about enhancing it**. I often walk through Amsterdam, earbuds in, listening to transcripts read aloud while crafting my blog posts. AI gives me the freedom to create without being tied to a desk. Similarly, for Marvin and his team, AI takes over the tedious parts of production, letting them focus on the creative work—like turning abstractive insights into compelling narratives. Marvin and I both agreed that the true magic lies in balancing the efficiency AI offers with the uniquely human capacity for creativity and connection. It’s about working smarter and having more fun while doing it. \--- This is a glimpse into how AI is being used behind the scenes in podcast production today. If you’re a digital professional interested in using these kinds of tools, think of AI as your co-pilot. It can do the heavy lifting, so you can focus on what you love most—telling stories that matter. [Follow me for future updates.](https://rob-hoeijmakers.ghost.io/newsletter/) ### My Experiment with Voice Cloning URL: https://rob-hoeijmakers.ghost.io/my-experiment-with-voice-cloning/ Last updated: 2026-04-25T13:16:06.000Z Have you ever wished your digital content could sound a bit more like you? Standard text-to-speech (TTS) voices are useful but lack a personal touch. These TTS solutions do a great job at converting text to sound, but they often miss the unique character that a personal voice can add. When I heard about [voice cloning technology](https://rob-hoeijmakers.ghost.io/exploring-the-potential-of-ai-avatars/), I wondered if I could bring my own voice into the digital world with the efficiency of AI. That’s where I decided to put [ElevenLabs](https://elevenlabs.io) to the test. I want to share how I experimented with transforming static texts into engaging content using my own voice clone. I’ll cover the workflow, tools, and why I think it’s a fascinating experiment for anyone interested in digital content creation. ## Experimenting with Voice Cloning These technologies have improved drastically due to advances in artificial intelligence. Both OCR and speech synthesis, including voice cloning, now use sophisticated AI models that make the processes much more accurate and natural. This improvement is what makes experimenting with them so exciting and accessible today. I wasn’t trying to solve a specific problem. Instead, I was curious—how good could a digital clone of my voice be? How natural would it sound when reading aloud? I wanted to test the quality and explore the possibilities of using my voice in digital content. Here is my workflow, keep in mind I did the voice cloning upfront: 0:00 /0:42 1× First OCR via iOS, then copy and past ## Enter ElevenLabs: Bringing My Voice to Life I found my answer in ElevenLabs, which offers voice cloning. The cloned voice captured my tone and cadence, adding authenticity that no generic AI voice could. Suddenly, I could read my writing without recording every word—it was my voice, just turbocharged. The cloned voice is also versatile. I’ve used it in tools like [HeyGen](https://www.heygen.com/), [KapWing](https://www.kapwing.com) and [Synthesia](https://www.synthesia.io/), making voiceovers straightforward. It’s a powerful way to create consistent voice content across platforms. ## My Workflow: OCR to Voice Cloning For this specific workflow, I focused on using text-to-speech on my mobile phone. Listening to audio is a personal experience, and I wanted something I could easily carry with me, rather than being tied to a desktop. My workflow starts with capturing text via OCR on iOS. I grab text from books, notes, or screenshots, then send it to ElevenLabs to generate an audio version in my voice. It’s fast and efficient, saving me from manual recording. Imagine taking a book quote and, in minutes, having an audio version of *you* reading it aloud. This has made my content creation faster and much more personal. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/11/ElevenLabs.jpg) While making screen shots I discovered you can also scan straight from the ElevenLab app. ## Beyond Voice Cloning: The Power of ElevenLabs ElevenLabs isn’t just for quick demos. It’s powerful for reading longer texts like PDFs, e-books, and webpages. It’s great for consuming information on the go and makes content accessible for those who prefer listening. What’s also fascinating is that ElevenLabs makes it possible to read and listen at the same time. This combination might be the key for me to learn more deeply from a text and stay more engaged with it. There’s also an option to embed a read-aloud player directly into text, which could be great for blog posts (minimal tier for this is Creator). Blending written and spoken content seamlessly is something I’m excited to explore. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/11/image-4.png) Audio Native example from ElevenLabs. You can embed such a player in your webpage. ## Final Thoughts Voice cloning might seem futuristic, but it’s here and easy to use. With ElevenLabs, I’ve been able to experiment with bringing my content to life, adding a personal touch that’s often missing in AI voices. What really struck me is that the digital clone, while not analog as such, is essentially a collection of parameters based on voices in general. It’s not trained specifically on *my* voice in the way a traditional recording would be, but instead configures the general model of voices to emulate *my* specific vocal characteristics. This approach makes it flexible and powerful, capable of producing a version of my voice that feels direct and natural. Whether for videos, audio articles, or just playing with new tech, voice cloning has huge potential. If you’re a digital professional looking to make your content stand out, this is worth exploring. [Let me know if you’ve tried anything similar—I’d love to hear about your experiences!](https://rob-hoeijmakers.ghost.io/about-rob-hoeijmakers/) ### The difference between slides and photos URL: https://rob-hoeijmakers.ghost.io/the-difference-between-photos-and-slides/ Last updated: 2026-04-25T13:16:05.000Z There is nothing more magical than the moment when light shines through a slide. The room darkened, the projector humming, and then that sudden burst of color on the screen. It was a moment of pure magic, when ordinary images were transformed into bright, vibrant, colorful tableaus. Literally a window on the world. But what makes slides so unique, so different from photographs? The answer lies in the fascinating world of **additive and subtractive color systems** . ## The magic of slides My first experience with slides was a revelation. Unlike photographs, which seem to absorb colors, slides seemed to radiate colors. This has everything to do with how slides create colors. Slides use an additive color system, where light is shone through transparent colors of cyan, magenta, and yellow to create the final color. The light shining through these colors creates vibrant and saturated images that appear to come alive on the projection screen. ## Colors in a printed photo Photographs, on the other hand, use a subtractive color system to create color. This process absorbs and reflects light rather than emitting it. In photographs, the colours are formed by the way in which the white light of the environment is absorbed and reflected by the dyes cyan, magenta and yellow on the paper. The result is often less vibrant and less saturated than with slides. ![A photo greatly enlarged.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/11/DraggedImage.png) A photo greatly enlarged. ## Two worlds, two systems The additive and subtractive color systems are two sides of the same color coin. In **the additive system** , as used in slides, white light is created by combining all the colors, while black results from the absence of light. In **the subtractive system** , as we see in photographs, the opposite is the case. White is the result of the absence of coloring, while black is created by combining all the colors. ![Additive and subtractive color mixing.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/11/DraggedImage-1.png) Additive and subtractive color mixing. ## From vibrant slide to muted photo The difference between additive and subtractive color systems presents interesting challenges, especially when converting a slide to a photo. The vibrant, bright colors of the slide can appear muted and different when printed on paper. This is an essential consideration for anyone working with both digital and printed images. Understanding the difference between these two color systems has deepened my fascination with the power of color and light. Technology and techniques change, but the fundamental principles of light and color remain. And it is this knowledge that allows us to create images that are not only visually appealing, but also convey deeper meaning and emotion. Because whether it is a slide or a photo, every image tells a story – a story written in color. The difference between a photo is also a nice comparison for the difference between [colors on a screen](https://webstrategieblog.nl/het-wonder-van-het-scherm/) and colors in print. It is precisely on the screen that slides come to life again and are more or less animated by light. ![Slide of Amsterdam, sixties. Not my own slide but that of an American tourist.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/11/DraggedImage-2.png) Slide of Amsterdam, sixties. Not my own slide but that of an American tourist. Slide is part of the book [Amsterdam in kleur, 1950-1970](https://www.bol.com/nl/nl/p/amsterdam-in-kleur/9300000181393870/). ## Next slide! Fun fact, once upon a time, presentations were given with a slide projector, where physical slides were projected onto a screen. Although the technology has evolved, we still refer to 'slides' in digital presentations like PowerPoint. This term refers to that old way of presenting and shows how old techniques continue to leave their mark on modern tools. I found a funny, and for older readers recognizable, fragment with slides. 0:00 /0:21 1× Short clip from the movie **Catch me as you can* to show slides as a presentation tool. ### Mastering the Digital Shift: From Video to Text URL: https://rob-hoeijmakers.ghost.io/from-text-to-video-and-back-again/ Last updated: 2025-04-17T09:13:02.000Z The digital world is evolving at a rapid pace, and one thing I feel strongly about is the need to stay proficient in all major media forms: text, images, audio, and video. We're witnessing **a notable shift from text and images to more dynamic formats like audio (such as podcasts) and synthetic video**. While I’m still shaping my perspective on this trend, it’s undeniable that we’re moving into an era dominated by new ways to communicate. If you want to stay relevant in your work or on social media, it’s becoming increasingly important to **master how to engage with video content**. This means being able to take a video and repurpose it—to respond to it, translate it, or analyse its contents. With the rise of AI-generated video, I predict we’ll face an ever-growing bombardment of video content. It’s easier and cheaper to produce than ever before, more democratic in its accessibility, and very much in favour with younger audiences. Here is a synthetic video with AI Avatar that got me started thinking about organising my workflow: Example video from [AI Avatars speaking through video](https://webstrategiesblog.com/exploring-the-potential-of-ai-avatars/) rather than sending textual information. ## Embracing the Challenge In light of this, I challenged myself to work from video to text using both my mobile device (iOS) and my computer (macOS). My goal is to get comfortable not just creating and sharing video, but also efficiently turning that content into actionable insights or further creative output—both in English and Dutch. In this article, I want to share my current workflows for video-to-text conversion, with a promise to dive deeper into [text-to-video workflows](https://rob-hoeijmakers.ghost.io/exploring-the-potential-of-ai-avatars/) in the near future. ## My Workflow for Video to Text (Mobile) I use an iPhone running iOS for my mobile workflow. This setup allows me to leverage a variety of apps that help streamline the video-to-text process. ### 1\. Find and Download Video Content The first step is to identify the video I want to work with. Getting a grip on a video is much harder than getting hold of a text—perhaps due to copyright issues or technological constraints like streaming and storage. It has been made deliberately challenging to get a copy of video content. Once I do find the video, it can be on Instagram, YouTube, X, TikTok etc, I use a tool called [**Video Lite**](https://apps.apple.com/nl/app/video-lite/id1598997500?l=en-GB) to download it onto my device (iPhone). *(I would love to have an app for mobile and desktop (iOS/Mac) that can download video's, store them in iCloud, make them searchable and transcribe them for the language they are in.)* ### 2\. Organise Video Files Using Folders After downloading, I organise video files in dedicated folders within iCloud. I need to move video's for this out of Video Lite by saving them to files. For videos I only need for transcription, I use the Downloads folder. Videos I want to save go into an iCloud folder named “Video.” This setup ensures seamless syncing across devices and simplifies sharing between apps, without cluttering up other storage locations. ### 3\. Transcribe the Video To transcribe the video, I use an app called [**Transcribe**](https://rob-hoeijmakers.ghost.io/transcribe-audio-on-macos-and-ios/). It’s reliable and user-friendly for both Dutch and English content, but ideally, I would expect Apple to offer a native solution for this step—handled locally on the phone or computer, rather than relying on the cloud. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/11/image-1.png) Transcription workflow through ****Transcribe**. This also uploads to the cloud. ### 4\. Export the Transcript for Further Work Once the transcription is complete, I export the text and bring it into ChatGPT. This is where the real creative work begins—refining the message, analysing the content, or even generating new ideas based on the transcript. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/11/image-2.png) Send the transcription from Transcribe to ChatGPT app ## Final Thoughts This workflow is still a work in progress, and I’m experimenting to see how I can make it more streamlined. As we continue to embrace AI and video content, I believe adaptability is key. It’s not just about creating content but learning how to wield it in various forms, from video to text and back again. In the next part of this series, I’ll share my approach [to creating synthetic video content from text](https://rob-hoeijmakers.ghost.io/exploring-the-potential-of-ai-avatars/)—an area that’s becoming more accessible, cheaper, and surprisingly versatile. [Stay tuned](https://rob-hoeijmakers.ghost.io/newsletter/), and I’d love to hear about your workflows for video and text. Are you finding new ways to repurpose content across formats? --- P.s. YouTube provides transcripts to most video's as well. They make it a bit hard to fetch these texts but there is a Browser Plugin (Chrome and Safari) for it called [YouTube Summary by Glasp](https://glasp.co/youtube-summary?ref=yt%5Fsummary). Personally I am not a big fan of plugins and extensions. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/11/image-3.png) The example video with transcription through extension. I found out that I also can drop video's into [Fireflies](https://rob-hoeijmakers.ghost.io/fireflies-ai-transcriptions/) and they provide transcription and analysis. You can also export the text als SubRip Subtitle (SRT). ### OS-Level Translations: Enhancing Writing Across Devices URL: https://rob-hoeijmakers.ghost.io/os-level-translations/ Last updated: 2026-05-10T08:54:09.000Z Lately, I've been diving deeper into translations as part of my content creation process. For blog articles, I've often relied on DeepL, ChatGPT, and Google Translate—each bringing its own strengths to the table. But recently, I've stumbled upon something else that has truly shifted my experience: **Apple Translate**. ## Apple Translate's Evolution Translation is often a two-way process: consuming text in foreign languages and translating from your own language into another. In this case, I've been focusing more on the latter—translating my own content into other languages to reach a broader audience. I mostly post to social media from my phone, which makes translation tools on mobile devices particularly interesting. Apple Translate, in its early days, felt very much like a basic clone of Google Translate. There was an app, it worked fine, and you could even translate a web page in Safari. Functional, but nothing revolutionary. Then I discovered—or rather, my wife pointed out—a feature that makes Apple Translate a powerful tool in my workflow: the ability to translate text directly in any input field on iOS and macOS. Now, when I'm drafting a [social media post in Dutch on X (formerly Twitter)](https://x.com/robhoeij), I can simply copy the text, switch over to [Threads](https://www.threads.net/@robhoeij), paste, and seamlessly translate it without leaving the text field. A quick edit and it’s ready to go. This convenience has streamlined my cross-platform posts, reducing friction and saving time. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/11/Translate-from-OS.jpg) Translation on Operating System (OS) level ## System-Level Integration and Its Impact This made me reflect on the importance of where translation functionalities reside— whether at the particular or general level. For translation (and soon, AI functionalities), having these tools at a general level makes so much more sense. Ideally, these tools should be local — working without an internet connection — and consistent across both mobile and desktop. The fact that Apple Translate now functions directly within any input field means that it’s not restricted to a specific app but is integrated at the system level, which greatly enhances usability and accessibility. Other examples where having functionalities at a general level significantly improves the user experience include: 1. **Spell Check and Grammar Tools**: System-wide spell checkers and grammar tools ensure consistency across all apps, making it easier for users to get assistance whether they are writing an email, a blog post, or a social media update. These tools function everywhere, reducing friction and improving the quality of content. 2. **Password Managers**: Password managers like [Apple's Password Manager](https://rob-hoeijmakers.ghost.io/mac-and-ios-password-manager/) provide secure access across all apps and devices without requiring users to switch to a separate app. This general-level integration makes managing credentials simple and efficient. 3. **Clipboard Management**: Apple's Universal Clipboard allows users to copy text or images on one device and paste them on another, making it a seamless experience. This general-level functionality supports productivity across devices. 4. **Offline Maps and Translation**: Tools like offline maps and translation services offer consistency and accessibility without an internet connection. This local, system-wide capability enhances privacy and reliability, similar to what Apple Translate has achieved with offline text translation. 5. **Voice Control and Dictation**: [Voice control and dictation capabilities ](https://rob-hoeijmakers.ghost.io/asr-speech-to-text-dictation-and-transcription-whats-the-difference/)integrated across devices provide universal access to voice commands. This system-level functionality, similar to translation, means users can interact effortlessly regardless of the app they are using. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/11/image.png) Translate on MacOS. Highlight, select and choose. ## Key Takeaways **The evolution of Apple Translate into a system-level tool** shows the immense value of integrating key functionalities across devices. **Whether it's translation, spell check, password management, or clipboard features**, having these tools universally accessible enhances user experience, productivity, and consistency by reducing friction and making interactions seamless. Next up, I'm planning to dig deeper into the quality of translations across these tools, especially for longer formats like blog posts. Each of these translation tools has its quirks, and it's interesting to see how they handle nuanced expressions, especially when I need translations to maintain a specific tone or voice. Exploring these aspects could shed light on how best to leverage them for creative, engaging content. [Stay tuned and subscribe.](https://rob-hoeijmakers.ghost.io/newsletter/) --- ### Further reading - [OS-Level Translation: A Multilingual Productivity Hack](https://rob-hoeijmakers.ghost.io/os-level-translation-a-multilingual-productivity-hack/) ### WhatsApp: Groups vs. Communities URL: https://rob-hoeijmakers.ghost.io/whatsapp-groups-vs-communities/ Last updated: 2026-04-25T13:16:41.000Z In the digital age, communication is all about managing social dynamics. WhatsApp offers two main structures: **Groups** and **Communities** , each suited for different types of interactions. Let's explore this dynamic with the metaphor of a tennis club. ## WhatsApp Groups: One Court, One Game A WhatsApp group is like a small tennis court where everyone plays together. All members share the same space, which makes communication focused and inclusive. This works well for smaller, tight-knit groups. As the group grows, however, it can feel crowded—like too many players on one court. For simple, focused discussions, a single group is ideal. ## WhatsApp Community: Multiple Courts, Multiple Games A WhatsApp community is like a larger tennis club with multiple courts. Different groups focus on different topics—some plan matches, others learn new techniques, while others just chat. This structure allows for specialization, but requires more management to keep everything organized. Communities provide a richer experience, but require careful balance to ensure clarity and inclusivity. ## Four types of Clubs Just as tennis clubs have different rules, WhatsApp has four main group setups: - **Private Club** : Admins invite members directly, no approval needed. Selective but easy to grow through trusted connections. - **Exclusive Club** : Admins invite, new members must be approved. Strictly controlled for exclusivity. - **Open Social Club** : Anyone can invite, no approval needed. Great for fast growth, but less control. - **Friendly but Selective Club** : Anyone can invite, but admins approve. Balances openness with oversight. | | **Admins Invite** | **Anyone Can Invite** | | ------------------- | ------------------ | ------------------------------- | | **No Approval** | **Private club** | **Open Sociale Club** | | **Approval Needed** | **Exclusive Club** | **Friendly but Selective Club** | [Setting up a WhatsApp CommunityLearn how to efficiently manage a WhatsApp community with features like announcements, member management and group management. Provide structure and clear communication.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-6.png)Rob Hoeijmakers (robhoeij)Rob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_5845-1-1.jpeg)](https://rob-hoeijmakers.ghost.io/setting-up-a-whatsapp-community/) ## Find the right balance The key to a successful WhatsApp group or community is understanding the social dynamics. Whether it’s a single group or a multi-layered community, finding the right balance between openness and structure helps foster meaningful conversations and stronger connections. Choose the format that fits your goals—whether you want a small, trusted group or a diverse community with multiple interests. With the right approach, you can make your WhatsApp space thrive. 💬 This WhatsApp Community is part of my blog 'Robhoeij'. [Join my groups on digital life and photography.](https://chat.whatsapp.com/JgNe9r1bJZuDQUDRyBhakO) The group starts as an 'Open Social Club' where everyone can freely join. Later it will be a 'Friendly but Selective Club': everyone can still invite members, but as an administrator I approve new registrations. This way I keep the community spam-free and maintain the friendly atmosphere. ### Exploring the Potential of AI Avatars URL: https://rob-hoeijmakers.ghost.io/exploring-the-potential-of-ai-avatars/ Last updated: 2026-04-25T13:17:09.000Z AI avatars are changing how we create and deliver content, offering adaptable spokespersons that can work across audiences and languages. I recently experimented with tools like Kapwing and Heygen, and saw how they can make content creation faster and more versatile. This article explores the potential, strengths, limitations, and opportunities of AI avatars. ## Why AI Avatars are exciting AI avatars offer many benefits for marketers, educators, and communicators by enabling consistent, dynamic, and scalable videos. The key strength is **adaptability**—avatars can change what they say and speak different languages. For a company with customers in multiple countries, AI avatars eliminate the need for costly re-shoots, voiceovers, and translations. This flexibility makes them highly promising, especially for multilingual audiences. ## Experimenting with Kapwing and Hygen I explored Kapwing and Hygen, which offer a variety of characters to suit different brand personalities. I experiment with these tools out of curiosity, but mostly for the benefit of my clients. My focus was on the Dutch language, and while the results were impressive, they still need improvement. The pronunciation lacked the nuance of a native speaker, but these tools are evolving quickly. Both platforms offer options to integrate text directly into the video or provide subtitles (SRT). Including subtitles ensures accessibility and helps with SEO when publishing content online. Here is the example with [Kapwing](https://kapwing.com): 0:00 /0:17 1× Persona Brian at a desk video And this I produced with [HeyGen](https://heygen.com): 0:00 /0:28 1× AI Avatar, HeyGen ## Next Steps: Creating my own AI Avatar Next, I plan to create an AI avatar using my own voice and image to explore personalisation. I currently don't have a personal use case where I could replace something existing. However, there are options for growth that I am now considering. I see content moving from text to video, and if I had my own avatar, I would probably use it to explain more about myself, rather than filming everything. My aim now is to publish a short video bio first. I plan to mix content—show who I am in real life and then hand it over to my avatar. Can a digital version of myself be convincing? Will it convey emotion and personality effectively? I currently am looking into [Synthesia](https://www.synthesia.io/?via=rob-hoeijmakers) and will share in the coming weeks. [Best AI Video Generator - Start Creating FREE AI Videos NowCreate AI videos from text with AI video generator. Get the most advanced AI avatars and voiceovers in 140+ languages. Try the free AI video generator today!![](https://cdn.prod.website-files.com/65e89895c5a4b8d764c0d710/664f0f482fa5a4d527d892bc_Favicon-Web-Security%201.png)Start Creating FREE AI Videos Now![](https://cdn.prod.website-files.com/65e89895c5a4b8d764c0d710/664dff84b972812764843b0f_NEW_OG.gif)](https://www.synthesia.io/?via=rob-hoeijmakers) ## Potential use cases and costs AI avatars could be used for customer support videos, personalised sales pitches, and multilingual marketing campaigns. I hope to identify which types of content work best, balancing automation with human connection. AI tools require subscriptions or per-use payments, which can add up. However, compared to hiring actors, filming crews, and editing teams, AI avatars are far more affordable. For small projects or budget-conscious businesses, they are a compelling alternative, and costs will likely decrease as technology matures. ## My conclusion AI avatars are not yet perfect, but they are improving fast. Their adaptability makes them valuable for businesses seeking to expand reach affordably. I look forward to sharing more, including my experiences creating a digital version of myself. What seems experimental today could soon become the norm. [Subscribe for updates.](https://rob-hoeijmakers.ghost.io/newsletter/) ### EduAI: An inspiring example of a WhatsApp Community URL: https://rob-hoeijmakers.ghost.io/eduai-an-inspiring-example-of-a-whatsapp-community/ Last updated: 2026-04-25T13:17:07.000Z In today's digital world, niche communities are becoming increasingly important for professionals who want to collaborate, share knowledge, and explore new developments. A prime example of this is the EduAI community, a WhatsApp group focused on the responsible use of AI in public education in the Netherlands and Belgium. With over 1000 members and seven specialized subgroups, EduAI is a model of how a community around two crucial themes - AI and education - can effectively organize and thrive. In this article I will dive deeper into this community, how it is set up, what its success factors are, and what others can learn from it. ## Background of EduAI EduAI was founded with the goal of supporting education professionals in understanding and responsibly applying AI in their daily practice. The community quickly became a hub for discussion, exchange of ideas, and collaboration, both in the Netherlands and Belgium. The founders had the vision to create a platform where members could not only receive updates and information, but also actively contribute to discussions and knowledge sharing. ![WhatsApp Community Overview Screen EduAI](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/12/2024-10-22_13-01-25-1.jpeg) WhatsApp Community Overview Screen EduAI ## Organization of the Community The power of EduAI lies in the organized structure of the community. With seven different subgroups, it provides a clear separation of functions and conversations, which increases engagement and relevance for members. Here are some of the key groups within the community: - **EduAI 1️⃣ Updates:** This is the central group for important announcements and updates. Only posts with links and summaries are posted here. Members can respond with an emoticon, but further discussions are encouraged in other groups to maintain clarity. - **EduAI 2️⃣ Discussion:** As the name suggests, this is the group where members can have extensive discussions about AI and education. Not only are fun tools shared, but the emphasis is on in-depth substantive conversations and the connection to education. - **EduAI 3️⃣ Promotion:** In this group, members can promote their own projects, tools, and initiatives related to AI and education. This provides a platform for members to make their work visible and inspire others. - **EduAI 4️⃣ Prompting:** This is a pilot group specifically focused on the topic of prompting, an important aspect within AI. The group is investigating the possibility of setting up a database for prompts, which could serve as a valuable resource for education professionals. In addition to these four groups, there are three other subgroups that focus on specific topics or needs within the community, further strengthening engagement and collaboration. ## Engagement and reach With over 1000 members, EduAI is a vibrant and active community. Group discussions are diverse and frequent, as evidenced by the hundreds of messages exchanged daily. Members range from AI experts to education professionals, bringing together a wide range of perspectives and experiences. These interactions have led to new collaborations, innovative ideas, and practical applications of AI in education. ![An event for the WhatsApp community](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/12/IRL-1.jpg) An event for the WhatsApp community ## Impact and examples EduAI has already had a significant impact on the field of AI and education, with several members organizing trainings and workshops that emerged from discussions within the community. In addition, new AI tools and methodologies have been developed and implemented in schools, with direct input and feedback from the community. A specific example is the 'Prompting' pilot group, which explores the possibility of a prompts database that could increase the efficiency and effectiveness of AI in educational settings. [WhatsApp: Groups vs. CommunitiesLearn how to strike the right balance between openness and organization to optimize your WhatsApp experience, using a tennis club as a metaphor.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-7.png)Rob Hoeijmakers (robhoeij)Rob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_8100-1-1.jpeg)](https://rob-hoeijmakers.ghost.io/whatsapp-groups-vs-communities/) ## Give it a try The EduAI WhatsApp community is a textbook example of how digital platforms can be used effectively to bring professionals together around urgent and complex themes such as AI and education. The organized structure, active involvement, and focus on both substantive discussions and practical applications make this community particularly successful. Want to be part of this exclusive group? Here you can [join through the website](https://edu-ai.nl/). 💬 I have also set up a WhatsApp Community for Robhoeij. You can join that directly. You can [try it out via this link](https://chat.whatsapp.com/JgNe9r1bJZuDQUDRyBhakO), preferably from your mobile. ### Turning Radio Mentions into Digital Text with Whisper: My Friday AI Experiment URL: https://rob-hoeijmakers.ghost.io/whisper-on-my-mac/ Last updated: 2026-05-26T19:36:59.000Z Fridays are perfect for experimenting with AI, machine learning, and all things tech. This time, I decided to try out [Whisper](https://rob-hoeijmakers.ghost.io/openais-whisper-speech-to-text-technology/), a library by OpenAI that can transcribe and translate audio, and I wanted to run it locally from my browser. It turned out to be a surprisingly easy setup. I had another nice coincidence that added a fun twist: I was mentioned on Dutch radio because of [a photograph of mine that went viral](https://rob-hoeijmakers.ghost.io/viral-photo/) and was covered by [CNN](https://arabic.cnn.com/travel/article/2024/10/10/amsterdam-viral-image-rob-hoeijmakers). That meant I had the perfect audio sample to experiment with! ## Setting up Whisper To start with Whisper, I used a GitHub file that walked me through the process of installing it on my Mac via the Terminal. After setting up the server, I opened it up in my browser, and there it was: a simple interface for uploading audio files to be transcribed or translated. > Run OpenAI's new Whisper Turbo model 100% locally in your browser with Transformers.js! ⚡️ > > Transcribe 2 minutes of audio in \~12 seconds! 🤯 > > Demo + source code 👇 [pic.twitter.com/Xj0sxwWeCX](https://t.co/Xj0sxwWeCX) > > — Xenova (@xenovacom) [October 1, 2024](https://twitter.com/xenovacom/status/1841080068593905960?ref%5Fsrc=twsrc%5Etfw) It worked locally on my machine, which I loved—keeping everything private, just between me and my tech. [GitHub - xenova/whisper-web at experimental-webgpuML-powered speech recognition directly in your browser - GitHub - xenova/whisper-web at experimental-webgpu![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/pinned-octocat-093da3e6fa40.svg)GitHubxenova![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/whisper-web)](https://github.com/xenova/whisper-web/tree/experimental-webgpu) ## The Radio Recording Then came the next part of my experiment. Since I had just been mentioned on the radio, I wanted to use that recording. I do have to be honest: I recorded the broadcast audio with my phone from my computer. Not the most professional way, but sometimes you just go with what's at hand. Once I had the audio, I fed it into Whisper via my browser, and it started transcribing the broadcast—my words and all. Video and screencast I made with TechSmith Snagit. I experimented further and uploaded the recording to [Fireflies.ai](https://rob-hoeijmakers.ghost.io/fireflies-ai-transcriptions/), my trustworthy notetaker. That transcription, in the cloud, was perfect. It even recognises the difference in the voices and splits them into Speaker 1 and Speaker 2, which you can change to names when necessary. It's fun to see how these different tools and formats can intertwine—from broadcast radio to an audio recording on my phone, to an AI transcription, and finally to a digital audio file shared online. It’s all about flipping between different digital media, making the most of what's available. Upgraded sound. Downloaded the podcast of the show, trimmed it with Audacity. ## Conclusion This Friday experiment turned into something more than just playing with AI. It was a way to connect all sorts of digital media in one cohesive flow, and honestly, it was a lot of fun. I think these moments, where tech meets the personal, are where [AI](https://rob-hoeijmakers.ghost.io/tag/ai/) really shines. [https://hoeijmakers.net/openais-whisper-speech-to-text-technology/](https://rob-hoeijmakers.ghost.io/openais-whisper-speech-to-text-technology/) ### Exploring ChatGPT's New Voice Features: More Hype than Substance? URL: https://rob-hoeijmakers.ghost.io/chatgpt-new-voice-features/ Last updated: 2026-05-10T08:54:09.000Z Recently, OpenAI released advanced voice capabilities for [ChatGPT](https://rob-hoeijmakers.ghost.io/try-out-chatgpt-yourself/), and I decided to give it a try. Living in the Netherlands, I hit a small snag—the feature wasn’t available here, possibly due to privacy restrictions in the [EU](https://rob-hoeijmakers.ghost.io/tag/europe/). So, I used a [VPN](https://rob-hoeijmakers.ghost.io/virtual-private-network-vpn/) to bypass this limitation and finally got access. ## **Emotional Voices: A Gimmick or a Feature?** The experience left me with mixed feelings. One of the big selling points is that the voices now carry more emotion. But, to be honest, I wasn’t particularly impressed by that. I don’t need an AI that sounds empathetic—I need one that gives me clear, concise answers. Adding emotional tone to responses feels more like a gimmick than a practical improvement. Another notable feature is that you can interrupt the AI while it's speaking, which is indeed a step forward. It makes the interaction feel a little more natural. But overall, these updates aren't the game-changer I was hoping for. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/10/ChatGPT-Advanced-Voice.jpg) ChatGPT - Advanced voice ## **Platform Inconsistencies** What really struck me, though, is the discontinuity across platforms. On the Mac app, which I also access via VPN, I’m still getting the older voice version. It’s confusing and inconsistent, and it honestly detracts from the experience. OpenAI seems to struggle with these inconsistencies, which only adds to my impression that they aren't quite there yet as a UX-focused company. For now, I keep seeing ChatGPT as more of an engine than a completed machine. It serves as a generic solution for my questions, which I often refine or finish in other apps. ## **Canvas and Memory Features** I was, however, impressed by the [memory feature ](https://rob-hoeijmakers.ghost.io/chatgpts-memory/)for now. It shows a lot of promise, though I will need to see how it behaves as it grows and handles more complex information. About Canvas, I am still in doubt a bit. To be honest, I did not have the bandwidth to test it in depth. I'm not a programmer, but I use ChatGPT with articles—for instance, this one—and I’m already growing a bit more enthusiastic about it. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/10/ChatGPT-Canvas.jpg) ChatGPT Canvas ## **Third-Party Opportunities** It made me realise that trying to cater to every need with a single tool is a tall order, and perhaps OpenAI's current approach isn't the best at doing so. I believe there’s a lot of room for third-party suppliers to take these foundational capabilities and build more user-focused solutions that align better with specific use cases. For now, I’m left feeling that while the tech is impressive in scope, it’s not quite hitting the mark for what I personally need—clear, reliable, and straightforward user experiences without the extra noise. 💡 ****Voice Capabilities of OpenAI** The voice capabilities of OpenAI are impressive as it is. They are now also available as [Realtime API](https://openai.com/index/introducing-the-realtime-api/). Pretty expensive in daily use but with a great potential to enhance your digital offering. --- ### Further reading - [ChatGPT suddenly sounds completely different](https://rob-hoeijmakers.ghost.io/chatgpt-suddenly-sounds-completely-different/) ### The Role of CDNs in Page Speed URL: https://rob-hoeijmakers.ghost.io/cdn-and-page-speed/ Last updated: 2026-04-25T13:17:06.000Z In my experience with different content delivery networks (CDNs) across multiple platforms, I’ve learned a few lessons about what really impacts page speed. For my Ghost blog (this website), the integrated CDN works impressively fast, while for [my WordPress site](https://webstrategieblog.nl), I ended up switching from the [Hostinger](https://hostinger.com?REFERRALCODE=1ROB567)\-offered CDN to **QUIC.cloud**. This switch was primarily because QUIC.cloud integrates seamlessly with LiteSpeed, providing a level of optimisation that just wasn’t there with my hosting partner’s default option. ### Why CDN Choice Matters With Ghost, the CDN is pre-configured and works straight out of the box, optimising delivery without much hassle. Ghost's integrated CDN effectively manages the delivery of static assets, which contributes to a smooth and fast user experience. In contrast, my [WordPress](https://rob-hoeijmakers.ghost.io/how-about-wordpress/) site required a more hands-on approach. I wasn’t entirely satisfied with the CDN that came with my hosting package—it felt like something was missing in terms of control and speed consistency. That’s when I explored **QUIC.cloud**, which works closely with the LiteSpeed Cache plugin to create a unified optimisation setup. **QUIC.cloud** offers deep integration with WordPress through LiteSpeed, allowing server-level caching and optimisation tailored to the platform. This integration makes a significant difference, allowing for full-page caching, dynamic content optimisation, and other advanced techniques that go beyond what a generic CDN usually offers. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/10/report2.jpeg) Page Speed report. All good! ### Lessons on Page Speed Through all this experimentation, one lesson stood out: **page speed improvements come primarily from effective caching and a well-configured network setup** rather than reducing image and page sizes alone. While image compression and lightweight page designs certainly help, they can't compensate for slow delivery mechanisms. A good CDN with effective server-side caching means the difference between average load times and a blazing-fast experience. Here’s what I found really matters: - **Server-side Caching**: The ability to cache HTML and database queries directly on the server before they even hit the CDN ensures that repeat visitors get served content as quickly as possible. This is where LiteSpeed’s integration with QUIC.cloud excelled. - **Optimised Network Configuration**: CDNs like QUIC.cloud provide edge servers in strategic locations. My WordPress audience is mainly in the Netherlands, and having Amsterdam as the key point of presence ensures that latency is minimal. The efficient routing and low-latency response times trump even the best optimised images when it comes to speed perception. - **Reducing Server Strain**: By offloading traffic to edge servers, the origin server can handle fewer requests and provide faster, more reliable performance, especially for dynamic content. ### Final Takeaway If you’re serious about optimising page speed, the takeaway from my own journey is clear: **prioritise a strong server-level caching solution and an optimised CDN** ### What is a WhatsApp Community? An Introduction URL: https://rob-hoeijmakers.ghost.io/what-is-a-whatsapp-community-an-introduction/ Last updated: 2026-04-25T13:16:37.000Z WhatsApp, the ubiquitous messaging platform, has introduced a new feature in 2022 that could change the way we organize digital groups: **WhatsApp Communities** . This addition provides an innovative and structured way to communicate with multiple groups simultaneously, which can be particularly valuable for all kinds of organizations and interest groups. ## What are WhatsApp Communities? WhatsApp Communities allows users to manage multiple groups under one umbrella. Imagine you have multiple WhatsApp groups, each of which is a variation on a larger theme — for example, neighborhood groups, sports teams, or project teams. Communities allow you to bring these groups together into one centrally organized space, where communication and information exchange can be more efficient. What makes this feature special is the ability to not only centralize group communication, but also segment it. This means that within a Community you can create different subgroups, each with their own specific focus, while still retaining the benefits of an overarching structure. ![A WhatsApp Community in English and Dutch](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/12/DraggedImage-2.png) A WhatsApp Community in English and Dutch ## Why WhatsApp Communities are Special The introduction of communities on WhatsApp offers several advantages that distinguish the platform from other social networks or group communication tools: 1. **End-to-End Encryption** : As with all WhatsApp communications, messages within Communities are also protected with end-to-end encryption. This makes the platform particularly suitable for confidential or sensitive information exchange, such as in healthcare or legal circles. 2. **Usability** : WhatsApp already has a huge user base, which often makes the transition to using communities easy. Many people are already familiar with the platform, which makes adoption easier. 3. **Flexibility and Organization** : Within communities, you can easily create new subgroups, manage members, and share messages that are relevant to all groups within the community. This provides a level of organizational flexibility that is often lacking in other platforms. ## Possible applications of WhatsApp Communities ### 1\. **Educational Environments** In schools and universities, WhatsApp Communities can serve as a powerful tool to connect students, teachers, and administrators. Within a school, different subgroups can be created for specific classes, subjects, or even extracurricular activities. This makes it easy to make important announcements that reach everyone, while still allowing for smaller, more focused discussions within subgroups. ### 2\. **Corporate Communications** For businesses, especially those with a decentralized structure, Communities can help streamline communication across multiple levels. Think of a company with multiple departments or branches; each department can have its own subgroup, while still allowing for company-wide announcements and information sharing. ### 3\. **Non-profit Organizations and Communities** Nonprofits can use Communities to organize volunteers, donors, and other stakeholders. Subgroups allow them to plan targeted actions and provide information to specific segments of their community, while the main group serves as a focal point for general communications. ## Critical comments on WhatsApp Communities While WhatsApp Communities is a promising new feature, there are a few considerations users should keep in mind: ### 1\. Dependence on WhatsApp By using WhatsApp Communities, you make yourself technically and content-wise dependent on the platform. This means that all communications and data remain within WhatsApp’s environment, under the company’s terms and policies. As with Facebook Groups, changes to these terms, the privacy policy, or even technical limitations of the platform can have a direct impact on the way Communities function. This can lead to issues around data management, platform dependency, and limitations in the degree of control you have over your own data and communication settings. ### 2\. **Restricted Uses for WhatsApp Business** Currently, WhatsApp Communities is primarily aimed at consumer users and has limited integration with WhatsApp Business. This means that businesses that rely on WhatsApp Business may not be able to take full advantage of Communities. The absence of business-specific features, such as advanced automation or customer support within Communities, may be a barrier for some organizations. ### 3\. **Not Everyone Uses WhatsApp** While WhatsApp is one of the most popular messaging apps in the world, not everyone uses the platform. This can be a challenge for organizations or groups that want to benefit from Communities, but have to deal with members or participants who do not have access to WhatsApp, for example due to privacy preferences or technical limitations. In such cases, it may be necessary to continue using alternative means of communication, which negates the benefits of centralization. ⚠️ In WhatsApp Communities, the groups functionality in particular poses a privacy challenge: within a group, participants can see each other's name and phone number. This can be a problem for companies, as they often have strict rules around sharing customer and employee data. This makes WhatsApp groups less suitable for applications where anonymity or data protection is crucial. ## WhatsApp Communities in Action WhatsApp’s introduction of Communities could have a significant impact on the way digital communities organize and communicate. By providing scalability and the ability to share information securely and in a structured manner, this feature could help drive increased efficiency and engagement across different types of communities. A recent poll among my followers on X showed that many people have heard of WhatsApp Communities, and a small percentage are even active in one. This shows that while the feature is not yet widely known, there is definitely interest. ![Poll on X to WhatsApp Communities](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/12/DraggedImage-1.png) Poll on X to WhatsApp Communities To better understand and utilize this new feature, I will soon [set up a WhatsApp Community](https://rob-hoeijmakers.ghost.io/setting-up-a-whatsapp-community/) myself . This will not only be an opportunity to experiment with the possibilities of the platform, but also to share experiences and insights. While you don't have to actively participate, I want to keep you updated on developments. I will also highlight a successful existing community so we can learn from best practices together. ### Llama 3.2 on a Mac URL: https://rob-hoeijmakers.ghost.io/llama-32-on-a-mac/ Last updated: 2025-07-20T14:25:06.000Z This Friday afternoon, I spent some time experimenting with Meta’s open-source [LLM](https://rob-hoeijmakers.ghost.io/a-comparative-analysis-of-top-language-models/), Llama 3.2 (3B). After installing Llama 3.2 on my Mac Mini, complete with a web UI via Docker, I found the setup surprisingly quick—it only took a couple of minutes thanks to this helpful video tutorial. I used the 3B version, that one is without vision. Steps to install Llama 3.2\. The speed and the privacy of running a local language model are definite advantages. It even provides useful code snippets when needed. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/09/DraggedImage-3.png) That said, it lacks the [memory functionality](https://rob-hoeijmakers.ghost.io/chatgpts-memory/) and multimodal capabilities I’m used to with ChatGPT, so for now, it’s still incomparable to closed models like that. However, for what it is—free, fast, and private—it’s a solid option. Short demo of Llama 3.2 running locally Below, I’ll share a quick article generated by Llama 3.2, 3B, itself, along with my thoughts on the installation process and performance. --- Story written by Llama 3.2, locally installed on my Mac. ## Unleashing the Power of Open-Source: Installing Llama 3.2 on My Mac Mini with Docker ### Introduction As an online professional, I'm always on the lookout for innovative tools to boost my productivity and creativity. Today, I embarked on a delightful Friday afternoon project – installing Open-Source Llama 3.2 (3.2B) on my trusty Mac mini, courtesy of Docker. The excitement is palpable as I dive into this new adventure, comparing it to my current default LLM, ChatGPT 4. ### The Setup I started by downloading and installing the Open-Source Llama 3.2 (3.2B) on my Mac mini. This wasn't a straightforward process, but with the help of Docker, I managed to get it up and running smoothly. To simplify the experience, I leveraged Docker's ease of use to create a containerized environment for Llama. ### The OpenWeb UI After successfully installing Llama, I installed the OpenWeb UI through Docker, which provided me with an intuitive interface to interact with the model. The user-friendly design made it easy to navigate and explore the capabilities of this powerful language model. ### First Impressions I must say, I'm thrilled to be using Open-Source Llama 3.2 (3.2B) on my Mac mini! The experience is incredibly smooth, with minimal latency and impressive response times. The OpenWeb UI has streamlined interactions, allowing me to effortlessly access the model's features. ### Comparison to ChatGPT 4 As I began exploring Llama's capabilities, I couldn't help but compare it to my current default LLM, ChatGPT 4\. Here are some initial observations: - **Accuracy:** Both models delivered impressive accuracy rates, with minor variations in precision depending on the topic and context. - **Response Time:** Llama (3.2B) responded almost instantly, whereas ChatGPT 4 took a fraction of a second longer. - **Contextual Understanding:** While both models demonstrated remarkable contextual understanding, I noticed that Llama excelled in handling multi-turn conversations. ### Conclusion My experience with Open-Source Llama 3.2 (3.2B) on my Mac mini has been nothing short of exhilarating! The ease of use and impressive capabilities have left me eager to explore more advanced features and applications for this powerful language model. As I continue to delve into the world of open-source AI, I'm excited to share more insights and discoveries with you, my fellow online professionals. Stay tuned for future updates on this journey! Let's get started on this fascinating adventure together! --- [Running a Local LLM on Your iPhoneI explored how far mobile AI has come by running LLMs directly on my iPhone — no cloud, no upload. Here’s what I learned from testing Haplo AI.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-163.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_5988.jpeg)](https://rob-hoeijmakers.ghost.io/running-a-local-llm-on-your-iphone/) I wrote about the Open Source strategy from Meta: [Future predictions for AI’sExplore the evolving landscape of AI: are we pushing the limits of technology or just scratching the surface?![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-164.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_2355-2.jpeg)](https://rob-hoeijmakers.ghost.io/future-predictions-for-ais/) ### The Mac and iOS Password Manager URL: https://rob-hoeijmakers.ghost.io/mac-and-ios-password-manager/ Last updated: 2026-04-25T13:17:06.000Z As someone who manages hundreds of online accounts, I know firsthand how much time can be wasted trying to keep track of passwords. With Apple’s latest releases for iOS and macOS, there’s finally a more comprehensive and user-friendly solution to this common problem. From password management to [two-factor authentication](https://rob-hoeijmakers.ghost.io/two-factor-authentication-2fa/) and [passkeys](https://rob-hoeijmakers.ghost.io/passwords-passkeys/), Apple has consolidated these features into a single, cohesive interface that simplifies security for both personal and professional use. Personally, I believe that an integrated, cross-device management of authentication details with 2FA is not only a huge factor in staying safe online but also key to improving day-to-day productivity. 🔐 My wife and I, both online professionals, primarily use Apple devices and Safari. This setup works for us due to its seamless integration, but it may not be ideal for everyone, especially if you use other platforms. ## Evolution of Password Management on Apple Devices Over the years, password management on Apple devices has gone from simple storage solutions to a robust system that integrates with the entire Apple ecosystem. From syncing passwords across devices to auto-filling credentials, Apple has made security more accessible for the everyday user. With [each iteration](https://rob-hoeijmakers.ghost.io/2fa-within-apple-keychain/), Apple’s approach to securing online accounts has evolved, making it easier for people to manage their accounts without sacrificing security. With this new update, the password manager feels even more streamlined, providing a unified hub for managing not only passwords but also account details and Wi-Fi networks. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/09/image-2.png) The Apple Password app on different devices ## What’s actually changed? On the surface, it might not seem like much has changed with the new password and authentication updates in iOS and macOS. The core features were already there, but the real improvement lies in the **consistent, unified experience** provided by the new interface. Previously, you had to know where to look for each function, often navigating through various settings to manage passwords, Wi-Fi, or authentication methods. Now, everything is brought together in one place, making it much easier to access and manage your security settings without the need to dig around. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/09/1200x600wa.png) The Apple Password app's icon ## Family Password Sharing One of the most significant new features is the ability to securely share passwords with family members. This has been a game-changer for me, both personally and professionally. My wife and I are both online professionals, and the need to share access to various accounts is a daily occurrence. This new feature allows us to do this securely and easily, without the usual hassle of constantly sending each other passwords. Whether it's sharing streaming services or professional tools, this new family password-sharing functionality is both convenient and secure, making account management smoother than ever. 💡 You can also share incidental accounts quickly through AirDrop, which has saved the day for us more times than I can count. It’s a fast and secure way to exchange credentials when needed. ## Enhanced Two-Factor Authentication: Prioritising security [Two-factor authentication (2FA)](https://rob-hoeijmakers.ghost.io/two-factor-authentication-2fa/) has long been an essential tool in protecting online accounts, and the new updates make it easier than ever to manage your 2FA settings. Importantly, Apple has made it simpler to implement more secure methods of 2FA based on local calculations, rather than relying on SMS-based authentication, which is known for vulnerabilities. For those of us who prioritise security, the more organised interface makes managing these settings straightforward and intuitive. I’ve been using this system to ensure my most important accounts are protected with the highest level of security available. ## Passkeys: A promising but cautious step forward Passkeys are a new authentication method designed to reduce reliance on traditional passwords by using unique digital keys. While they offer enhanced security, I’m still a bit hesitant to fully embrace them. One concern is the risk of potentially locking myself out of accounts if things go wrong, particularly if a browser or device doesn’t support them. For example, I recently encountered an issue where passkeys didn’t work on Safari, and I had to switch to Chrome to log in. While the [technology is promising](https://rob-hoeijmakers.ghost.io/passwords-passkeys/), I’m taking a [cautious approach](https://rob-hoeijmakers.ghost.io/transitioning-to-passkeys/) until I fully understand the potential risks and have a clearer sense of its reliability across platforms. ### My personal conclusion In my view, the password management system built into Mac and iOS is a highly comprehensive and well-integrated solution for the entire lifecycle of authentication. It’s not just an app but a deeply embedded feature that works seamlessly across Apple devices, offering robust security and convenience. I strongly recommend it for its ease of use and thoughtful design, though I do recognise the limitations that come with being tied to the Apple ecosystem. ## ### Anki, Human Memory in a Digital World URL: https://rob-hoeijmakers.ghost.io/human-memory-in-a-digital-world/ Last updated: 2026-05-10T08:54:10.000Z Is memorising information still important to you? For me, it’s essential. As information flows more freely and rapidly, the real challenge is ensuring that you make it your own. I recently wrote about [ChatGPT’s memory](https://rob-hoeijmakers.ghost.io/chatgpts-memory/), which I really like for its ability to recall and manage details across conversations. But now, I want to shift the focus to human memory and the role it plays in navigating today’s overwhelming flow of information. ## My approach to memorising Memorising for me is a multi-step process, built on several foundations: using flashcards, making mind maps (MindNode), reading from paper or [an e-reader](https://rob-hoeijmakers.ghost.io/reading-webpages-on-your-kindle/), [taking notes](https://rob-hoeijmakers.ghost.io/kobo-highlights-and-notes-to-your-mac/), [writing blogs](https://rob-hoeijmakers.ghost.io/blog-writing-workflow/) and even having conversations with [ChatGPT while I walk](https://rob-hoeijmakers.ghost.io/this-week-chatgpt-learned-to-draw-see-and-talk/). Each of these elements plays a part in ensuring that the information I encounter sticks with me. But in this article, I want to share the most effective tool in my process: my **flashcards app Anki** and the workflow that supports it. ## My flashcards workflow: From reading to retention When I come across an important document, such as a report, I often read it on my e-reader to maintain focus. While reading, or after I’ve finished, I upload the document to ChatGPT or Google NotebookLM to help break down and summarise the key points. I then ask it to generate flashcards based on this content. ![Creating a stack of flash cards from a document with ChatGPT.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/09/DraggedImage-2.png) Creating a stack of flash cards from a document with ChatGPT. These flashcards are exported as a CSV file and imported into Anki, my go-to app for spaced repetition learning. By reviewing these flashcards regularly, I reinforce the material over time, making sure the most important information is retained. ![Flash cards in Anki, synced through iCloud. ](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/09/DraggedImage-1.png) Flash cards in Anki, synced through iCloud. This process is powerful because it combines active engagement—through reading, summarising, and creating flashcards—with the spaced repetition technique, which optimises memory retention. And above all, it is so smooth, it is fun! ## Why this workflow works By combining reading, digital tools, and flashcards, I create a system that balances technology with human memory. The digital tools streamline the process of summarising and creating learning materials, while Anki helps ensure that the knowledge is reinforced at the right intervals. The added layer of review and repetition solidifies my understanding, turning fleeting information into lasting memory. --- ### Anki and Space repetition For me it really works, especially if I need to enter a new space of knowledge. It helps you sifting out things that stick quickly and things you find hard to remember. **Anki** is a flashcard-based tool designed to help users retain information more effectively. It uses spaced repetition, a scientifically proven method of learning that schedules reviews of flashcards at intervals tailored to maximise long-term memory retention. The idea is simple: you review information just before you're about to forget it, which strengthens your ability to recall it in the future. The software looks a bit clunky and basic, but it has everything you need. It evens syncs through iCloud and so I can kill time with it when on the road. [Anki - powerful, intelligent flashcardsAnki - a program which makes remembering things easy.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/logo.svg)powerful, intelligent flashcards![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/logo.svg)](https://apps.ankiweb.net) **Spaced repetition** is based on the principle that we learn best when information is reviewed over time, with increasing gaps between each review. When a concept is recalled easily, the gap between reviews lengthens. When it's harder to recall, the interval shortens. This approach optimises the timing of your reviews, preventing overload and ensuring that the most important information is always at the top of your mind. Anki's algorithm automates this process, adjusting the review schedule for each card based on your responses, making it a highly efficient tool for memorisation. --- ### Further reading - [Spaced repetition - Wikipedia](https://en.wikipedia.org/wiki/Spaced%5Frepetition) ### Europe’s digital dilemma and new commission URL: https://rob-hoeijmakers.ghost.io/europes-digital-dilemma-and-new-commission/ Last updated: 2026-05-10T08:54:10.000Z **Europe has long prided itself on its progressive stance on privacy, fairness, and ethics in the digital space.** Regulations like the [GDPR](https://rob-hoeijmakers.ghost.io/gdpr/) have been hailed as victories for data protection and digital sovereignty, while newer frameworks such as the [DSA](https://rob-hoeijmakers.ghost.io/europes-dsa-and-the-new-contours-of-online-freedom/), [DMA](https://rob-hoeijmakers.ghost.io/eu-digital-markets-act/), and the upcoming [AI Act](https://rob-hoeijmakers.ghost.io/learning-to-work-with-the-eu-ai-act/) aim to create a fairer, more controlled digital market and ensure ethical standards for emerging technologies. Yet, while these policies reflect a noble intent—protecting citizens’ rights and curbing the dominance of Big Tech—they may also act as anchors on the region’s technological innovation. Compared to the US and China, where advancements in AI, cloud computing, and digital platforms are driven by risk-taking and agility, Europe seems to be lagging behind. The question is: why? This article explores the delicate balance [Europe](https://rob-hoeijmakers.ghost.io/tag/europe/) must strike between protecting its citizens and fostering an environment that encourages rapid digitalisation, innovation, and competitiveness. With the AI Act setting strict ethical standards for artificial intelligence, the Digital Markets Act targeting monopolistic behaviours, and the GDPR safeguarding privacy, Europe faces a complex challenge: how to regulate without stifling the growth it needs to thrive in the global digital economy. ## Regulating Innovation: How GDPR, DSA, DMA, and the AI Act may be holding Europe back One could argue that Europe's privacy and digital regulations have become a double-edged sword. On one hand, they shield citizens from unwanted intrusions, especially from global tech giants. On the other, they may be isolating Europe from the very innovation hubs that could help it grow. The GDPR, along with the newer [DSA](https://rob-hoeijmakers.ghost.io/europes-dsa-and-the-new-contours-of-online-freedom/) and [DMA](https://rob-hoeijmakers.ghost.io/eu-digital-markets-act/) frameworks, initially designed to protect individuals and ensure fair competition, can also be viewed as tools of economic defence—keeping American and Chinese companies at arm’s length. The upcoming AI Act adds yet another layer of oversight, potentially slowing down AI advancements in Europe, while other regions surge ahead. However, this protectionism may come at a steep price. The technological ecosystems of the US and China are already racing ahead, with fewer regulatory hurdles. Could Europe’s focus on stringent regulation be stifling its own potential, particularly in key areas like AI, digital services, and cloud infrastructure? As Europe doubles down on its regulations, such as the AI Act that imposes specific requirements on AI development, it risks becoming a bystander in the race for digital leadership. [Draghi report: How Europe’s Digital Lag threatens innovationA personal reflection on how EU regulations impact innovation and competitiveness, drawing on Mario Draghi’s report and my experiences as a European entrepreneur.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/size/w256h256/2024/05/Logo-Rob-Hoeijmakers-Square-Dark-Glow-3-1.png)Web StrategiesRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/size/w1200/2024/09/IMG_5643.jpeg)](https://rob-hoeijmakers.ghost.io/europe-between-innovation-and-regulation/) ## The Role of the new commission: A turning point? With the new European Commission taking shape, there is an opportunity to reassess this course. The next five years will be critical for Europe's digital future. Will the new Commission maintain the regulatory-heavy approach, or will it introduce reforms that encourage innovation while still safeguarding privacy? The path they choose could make or break Europe’s digital transition. This Commission holds the potential to either shift Europe towards a more agile, innovation-driven future, or continue along the current trajectory of over-regulation. It is crucial that policymakers recognise the importance of nurturing not just privacy, but also digital infrastructure, cross-border collaboration, and entrepreneurship if Europe wants to remain competitive. The new Commission is being formed under the leadership of Ursula von der Leyen, but it still needs to be approved by the European Parliament. I will be following this process closely and provide an update in the future. [The eight biggest surprises from Ursula von der Leyen’s new CommissionFrom Austria’s unexpected assignment to Spain’s mega portfolio, the next Commission of Ursula von der Leyen has Brussels talking. #EuropeDecoded![](https://www.euronews.com/apple-touch-icon-180x180.png)EuronewsJorge Liboreiro![](https://static.euronews.com/articles/stories/08/73/74/74/1200x675_cmsv2_917c7a7f-68c5-5931-a799-85a736a0b41e-8737474.jpg)](https://www.euronews.com/my-europe/2024/09/17/the-eight-biggest-surprises-from-ursula-von-der-leyens-new-european-commission) --- ### New Rules of the Digital Game: DSA, DMA, and the AI Act To understand the regulatory landscape shaping Europe’s digital future, it’s essential to look at three significant frameworks: the [**Digital Services Act (DSA)**](https://rob-hoeijmakers.ghost.io/europes-dsa-and-the-new-contours-of-online-freedom/), [**Digital Markets Act (DMA)**](https://rob-hoeijmakers.ghost.io/eu-digital-markets-act/), and the **AI Act**. These regulations aim to create fair, safe, and transparent digital spaces by addressing online platform accountability, monopolistic control by major tech companies, and ethical AI deployment. I will delve into each of these in more detail later on, exploring how they affect not only the big tech firms but also European entrepreneurs and the broader digital ecosystem. Stay tuned as I unpack the implications and potential challenges these regulations bring. 1. [**Digital Services Act (DSA)**](https://rob-hoeijmakers.ghost.io/europes-dsa-and-the-new-contours-of-online-freedom/): The DSA focuses on creating safer digital spaces by imposing obligations on online platforms, especially very large ones, to tackle illegal content, disinformation, and ensure transparency in advertising and algorithms. It establishes a framework that holds online platforms accountable for the content they host, ensuring more responsibility towards users and society. The DSA's goal is to protect user rights while maintaining a balance with innovation. 2. [**Digital Markets Act (DMA)**](https://rob-hoeijmakers.ghost.io/eu-digital-markets-act/): The DMA aims to prevent unfair practices by large online platforms identified as “gatekeepers”—companies that have a strong market position and control over digital ecosystems. The law requires gatekeepers like Apple, Google, and Meta to comply with specific obligations to ensure competition and fairness in digital markets. Non-compliance can lead to massive fines or even break-up of companies in extreme cases. The DMA is primarily designed to open up markets, promote competition, and reduce monopolistic control. 3. [**AI Act**](https://rob-hoeijmakers.ghost.io/learning-to-work-with-the-eu-ai-act/): The AI Act is Europe’s legislative proposal to regulate artificial intelligence (AI). It categorises AI systems based on their level of risk—from minimal to unacceptable—and applies corresponding regulatory obligations. High-risk AI systems (e.g., in health, law enforcement, or employment) will face strict transparency and accountability requirements. The act is intended to ensure AI systems are safe, ethical, and transparent, but some worry that it could slow down AI innovation in Europe compared to other regions like the US and China. ### Further reading - [Europe - Rob Hoeijmakers](https://rob-hoeijmakers.ghost.io/tag/europe/) - [The Digital Markets Act is Here and It’s Changing How Apps Work in the EU](https://rob-hoeijmakers.ghost.io/eu-digital-markets-act/) - [From Resistance to Strategy: What the EU’s Digital Acts Mean for the Work I Do](https://rob-hoeijmakers.ghost.io/eu-digital-acts/) - [Europe’s DSA and the New Contours of Online Freedom](https://rob-hoeijmakers.ghost.io/europes-dsa-and-the-new-contours-of-online-freedom/) ### Draghi report: How Europe’s Digital Lag threatens innovation URL: https://rob-hoeijmakers.ghost.io/europe-between-innovation-and-regulation/ Last updated: 2026-05-10T08:54:10.000Z As a European professional working in the digital space, I find myself in a unique and often frustrating position. Most of my projects rely heavily on American software and cloud services, not because I prefer them over European alternatives, but because there simply aren’t comparable options. The tools I use are progressive, intuitive, and they make it easy to add real value to the world—whether it's in information sharing or content creation. But there’s a catch. All these advancements are constantly offset by the bitter side of doing business in the [EU](https://rob-hoeijmakers.ghost.io/tag/europe/): compliance. As a digital professional, I often feel torn. On one hand, I want to innovate, to push boundaries, and to create something that improves people's lives. On the other, I have to constantly navigate a sea of regulations that can stifle that very innovation. While the American tech scene seems to race forward, Europe's approach feels bogged down by layers of rules that don’t always fit the pace of change we're experiencing. It’s hard not to feel like we, as European professionals, are being squeezed into a corner, constantly having to balance progress with compliance. ## The Digital Gap and Europe’s competitive struggles It’s not just a personal frustration. A recent report by Mario Draghi lays bare the wider issues Europe is facing. The gap between the EU and global leaders like the US and China, particularly in digital innovation, is widening. While the US and China surge ahead in foundational technologies like AI, cloud computing, and quantum computing, Europe lags behind. The reasons are many—fragmented markets, overregulation, and a lack of cohesive investment in high-tech sectors—but the impact is the same: a continent that is struggling to keep up with the pace of global digital transformation. The report highlights how Europe’s slow adoption of technology, combined with regulatory burdens, has caused tech companies to look elsewhere for investment and scale-up opportunities. Meanwhile, European professionals like myself find ourselves increasingly dependent on external services just to keep up with the demands of our work. The EU’s ambition to close this gap is there, but without significant reforms, it feels like that ambition may fall short. ## Looking forward: A chance for change However, there is some hope on the horizon. With the [election of a new EU Commission](https://rob-hoeijmakers.ghost.io/europes-digital-dilemma-and-new-commission/), there’s a real opportunity to rethink our approach to innovation and regulation. The next few years could be crucial in determining whether Europe can rise to meet the challenges of the digital age or continue to fall behind. In a recent article, I explored some of the upcoming shifts in the political landscape and what they could mean for entrepreneurs like myself—perhaps this momentum will lead to the reforms we so desperately need. [Europe’s digital dilemma and new commissionEurope faces a pivotal moment in its digital strategy, with the new EU Commission balancing innovation with regulation to shape the continent’s future.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-86.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_1248.jpeg)](https://rob-hoeijmakers.ghost.io/europes-digital-dilemma-and-new-commission/) ### Key Innovation Points from the Draghi Report The *Draghi Report* on the future of Europe’s economy provides critical insights into the region's struggles with digital innovation. While European policies have historically focused on privacy and regulation, the report reveals how these approaches may be hindering the growth needed to compete in the global digital landscape. I believe the report is well worth reading for anyone interested in understanding these challenges. However, if time is a concern, you can always listen to it as a podcast—I've written an article on how to [easily convert PDF documents into audio form](https://rob-hoeijmakers.ghost.io/from-pages-to-podcasts/). [EU competitiveness: Looking aheadRead Mario Draghi’s report on the future of European competitiveness. Its findings will contribute to the Commission’s work on a new plan for Europe’s sustainable prosperity and competitiveness.![](https://commission.europa.eu/profiles/contrib/ewcms/themes/ewcms_theme/images/favicons/ec/favicon.svg)European Commission![](https://commission.europa.eu/sites/default/files/styles/ewcms_metatag_image/public/2024-09/Draghi-report_featured-media.png?itok=RdMzKhzg)](https://commission.europa.eu/topics/strengthening-european-competitiveness/eu-competitiveness-looking-ahead%5Fen) 1. **Innovation Deficit**: The report highlights that Europe’s digital infrastructure is falling behind other major global players, particularly the US and China. This lack of digital progress is cited as a significant factor in Europe’s economic stagnation. 2. **Policy vs. Innovation**: European policymakers have focused on regulation, often at the expense of creating an environment that nurtures innovation. The heavy regulatory framework, while protecting citizens' rights, risks stifling digital growth and technological advancement. 3. **Limited Investment in Tech**: Europe’s investment in high-growth tech sectors, like AI and digital platforms, lags behind global competitors. There is a clear need for increased venture capital and support for startups if Europe is to close the gap in these key areas. 4. **Fragmented Digital Market**: The report points out that Europe’s digital market remains fragmented, making it difficult for tech companies to scale across the continent. Harmonising digital regulations and creating a more unified market are essential for fostering innovation. 5. **Skilled Workforce Shortage**: Europe faces a shortage of workers skilled in digital technologies, further exacerbating the innovation deficit. The report recommends improving education and training in digital skills to meet the demands of the modern economy. 6. **The Role of AI**: Artificial intelligence is seen as a cornerstone of future innovation, but Europe’s progress in this area is sluggish. The report stresses the importance of integrating AI across industries to drive future growth and competitiveness. --- ### Further reading - [From Pages to Podcast with Google NotebookLM](https://rob-hoeijmakers.ghost.io/from-pages-to-podcasts/) ### From Pages to Podcast with Google NotebookLM URL: https://rob-hoeijmakers.ghost.io/from-pages-to-podcasts/ Last updated: 2026-05-26T19:31:54.000Z Today, I stumbled upon a tool that has completely shifted how I think about content transformation. [Google's NotebookLM](https://notebooklm.google.com/), still in beta, offers something truly remarkable: the ability to not only interact with long-form content through chats but to actually convert that content into something entirely new—a podcast. This was no small task either. I tested it on [a 60-page report on the future of the European economy by Mario Draghi](https://rob-hoeijmakers.ghost.io/europe-between-innovation-and-regulation/). Within moments, the tool summarized the report and produced a fully narrated podcast, with voices so natural, I had to remind myself this was a machine. In this post, I’ll walk you through my experience with NotebookLM, from document upload to podcast creation, and share how you can try it out yourself. ## What is NotebookLM and why does it matter? NotebookLM is part of Google’s continuing push to make information more accessible and interactive. At its core, it allows you to upload a document and interact with it as if you were having a chat. You can ask questions, get summaries, and receive key insights directly from the text. While this is interesting on its own, what really caught my attention was the tool’s ability to turn long-form documents into podcasts. We’ve all seen tools that summarize content or let you chat with a document, but NotebookLM goes one step further by converting dense reports into audio content—complete with multiple voices that sound astonishingly natural. This opens up a world of possibilities for those of us looking for alternative ways to consume information on the go, or for those who simply prefer listening over reading. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/09/DraggedImage.png) ## Transforming a 60-page report into a podcast As a test, I uploaded Mario Draghi’s 60-page report on the future of the European economy. Within minutes, NotebookLM summarized the key points and offered to convert the text into a podcast with two distinct voices. The setup was straightforward: no complex configuration or voice selection was needed. The magic happened when I hit play. The voices sounded so natural that I could easily imagine myself listening to a professional podcast. Gone are the days of robotic-sounding text-to-speech. This was different. The voices flowed smoothly, with appropriate pauses and even subtle inflections that made it hard to believe it wasn’t a human narrating. The whole experience was seamless, from start to finish, and I could already imagine using this for any number of professional or personal projects. ![Google's NotebookLM with the PDF, the chat interface and the generated audio.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/09/image.png) Google's NotebookLM with the PDF, the chat interface and the generated audio. ## How you can try it: My experiment results I’ve uploaded the result of my experiment to SoundCloud so you can hear the podcast version of the [Draghi report](https://rob-hoeijmakers.ghost.io/europe-between-innovation-and-regulation/) yourself. You’ll find the link below, along with a downloadable version of the original report for those who want to dive into the full document. The podcast from NotebookLM as podcast on Soundcloud. [The PDF with the report on the future of European competitiveness](https://commission.europa.eu/document/download/97e481fd-2dc3-412d-be4c-f152a8232961%5Fen?filename=The%20future%20of%20European%20competitiveness%20%5F%20A%20competitiveness%20strategy%20for%20Europe.pdf) Whether you’re a researcher, content creator, or simply someone who prefers audio, tools like NotebookLM are paving the way for a more flexible approach to how we consume content. I’m excited to see where this technology goes and will be keeping a close eye on how it evolves. In the meantime, I encourage you to give it a try and see for yourself what this new wave of content transformation can offer. 💡 NotebookLM is your personalized AI research assistant powered by Google's most capable model, **Gemini 1.5 Pro*. Check out my [comparison of LLM's](https://rob-hoeijmakers.ghost.io/a-comparative-analysis-of-top-language-models/). ### ChatGPT’s Memory URL: https://rob-hoeijmakers.ghost.io/chatgpts-memory/ Last updated: 2026-04-25T13:16:41.000Z ChatGPT has introduced a new memory feature that enhances its ability to personalise interactions. As a frequent user, I’ve had the chance to explore this feature in the past week. In this article, I’ll share what I’ve learned about how ChatGPT’s memory works, its benefits, and why you might want to consider privacy settings. This feature is currently rolling out globally to ChatGPT Plus users—like myself in the Netherlands. ## How does ChatGPT’s memory work? ChatGPT’s memory allows it to retain information from past interactions, so you don’t have to keep repeating yourself. For instance, if I mention that I enjoy coffee, ChatGPT will remember that preference in future conversations. This leads to more personalised responses and a smoother experience overall. Here I make memory of my love for coffee. Here you see making memories in action ## Managing your memory: full control One of the best aspects of this feature is the transparency it offers. ChatGPT shows you what information it’s stored, and you can easily manage that data. You can review, update, or even delete memories whenever you like. This flexibility ensures that you remain in control of what ChatGPT remembers about you, keeping things personal but not intrusive. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/09/2024-09-07_10-20-45.png) My updated ChatGPT memory ## Privacy concerns: what you should know By default, ChatGPT uses your data to improve the model’s training. However, if you’re concerned about privacy, there’s an option to switch off data sharing for training purposes. While the specific details of how personal information impacts training aren’t always clear, disabling this setting is the safer route if privacy is a priority for you. ## Final thoughts: exploring the pros and cons While privacy concerns are always on my mind, I’m fully committed to exploring ChatGPT’s memory feature for a few weeks. I want to see what it gathers, how it improves my experience, and weigh the pros and cons. What makes me comfortable with testing it is the transparency—knowing exactly what ChatGPT remembers gives me a sense of control. For now, I’m happy to give it a try and will keep you updated on how it works out. 🧠 It's important to note that ChatGPT's memory feature is only available in the full version of the model. If you're using the ****mini version**, it won't retain any information across sessions. ### From Blueprints to Bytes: How Original Designs Shaped Modern Technology URL: https://rob-hoeijmakers.ghost.io/blueprints/ Last updated: 2026-04-25T13:16:06.000Z When we think about blueprints, detailed technical drawings for buildings and machines come to mind. However, the original blueprint process, known as cyanotype, was created in 1842 by Sir John Herschel and first used by botanists to reproduce the shapes of leaves and flowers. This involved placing objects on light-sensitive paper and exposing them to sunlight, resulting in white silhouettes on a blue background. ![Work by Anna Atkins, a 19th-century botanist. Sort of photo without a camera.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/07/B1F59F38-5D84-42E7-8FE3-7CB9B015C696.jpeg) Work by Anna Atkins, a 19th-century botanist. Sort of photo without a camera. ## From Replication to Design The cyanotype method soon found a new purpose in engineering and architecture. Initially used to replicate existing technical drawings accurately, it eventually became integral to the design process itself. Engineers and architects began using blueprints not just to copy plans, but to create and refine them, ensuring every detail was meticulously planned before construction began. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/07/032F28C2-D2B5-4BDC-A168-85615B997D05.jpeg) Blueprint of a building ## Chip Design: Digital Blueprints In modern technology, the blueprint concept has evolved into digital forms, especially in chip design. Engineers create detailed schematics that outline the placement of millions of transistors. These digital blueprints are crucial for constructing microchips, ensuring they function correctly and power devices from smartphones to AI systems. ![Blueprint of an Intel chip](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/08/Image-31-07-2024--13-40-Large.jpeg) Blueprint of an Intel chip ## Website Building: Structuring the Digital World In website building, blueprints take the form of wireframes and site maps. These digital sketches map out the structure and layout of websites, guiding designers and developers through the creation process. Wireframes, like architectural drawings, show where each element will be placed, providing a clear plan before coding begins. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/07/E4527FCE-1B29-48DA-BCD7-29A691703E9C.jpeg) Website blueprint ## Bridging the Gap The evolution from cyanotype to digital blueprints highlights a significant shift: the blueprint moved from a method of replication to a fundamental part of the design process. Whether capturing the delicate structure of a leaf or designing a microchip or website, the principles of detailed planning and precision remain essential. Next time you navigate a sleek website or marvel at your latest gadget, remember that the same fundamental design principles have guided creation from the days of cyanotype blueprints to today’s digital age. \--- This journey from cyanotype replication to digital design underscores how foundational ideas adapt and thrive, shaping the way we build and interact with technology today. ### Strengthening Europe's Semiconductor Backbone: The Role of ESMC URL: https://rob-hoeijmakers.ghost.io/esmc/ Last updated: 2025-03-19T14:27:52.000Z In the digital age, semiconductors are the bedrock of virtually every electronic device, from smartphones to advanced automotive systems. Yet, the industry faces significant challenges, particularly regarding the robustness of the global supply chain. Enter the European Semiconductor Manufacturing Corp. (ESMC), a collaborative initiative aimed at transforming Europe's semiconductor landscape. ## The Importance of ESMC ESMC is not just another corporate entity; it's a strategic alliance that includes technology giants, research institutions, and governmental bodies. This initiative is designed to enhance Europe's semiconductor manufacturing capabilities, increase production capacity, and reduce dependency on non-European sources. With global supply chains under threat from geopolitical tensions and natural disasters, Europe's need for a self-sufficient semiconductor industry has never been more pressing. ## Key Players and Collaborative Efforts The consortium behind ESMC involves some of the biggest names in the tech industry. Recently, Taiwan Semiconductor Manufacturing Co. (TSMC), the world's leading chipmaker, announced plans to build its first European plant in Dresden, Germany. This €10 billion joint venture includes Bosch, Infineon Technologies, and NXP Semiconductors, with TSMC holding a 70% stake and the others sharing the remaining 30% equally. ![TSMC, Bosch, Infineon and NXP form ESMC](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/08/ESMC-logos.jpg) TSMC, Bosch, Infineon and NXP form ESMC The plant, also called foundry, will focus on producing 22nm and 28nm chips, essential for automotive and industrial applications, and is set to begin production by 2027. 🇪🇺 While no longer state-of-the-art, 22nm and 28nm processes continue to be highly relevant in the industry. They are produced in large volumes and serve various market segments where leading-edge nodes might be unnecessary or too costly. ## The Broader Impact of the European Chips Act Supporting these efforts is the European Chips Act, a €43 billion initiative aimed at doubling Europe's share of global semiconductor production to 20% by 2030. This ambitious plan seeks to create a more resilient and competitive semiconductor ecosystem within the region. By securing substantial government subsidies, such as the up to €5 billion expected for the Dresden plant, the EU is taking concrete steps to ensure the industry's growth and sustainability. ## Why Digital Professionals Should Care For digital professionals, understanding the intricacies of the semiconductor supply chain is crucial. Chips don't grow on trees; they require sophisticated manufacturing processes and significant investments in research and development. The robustness of the semiconductor industry directly impacts the tech ecosystem, influencing everything from product innovation to market stability. With the ESMC and similar initiatives, Europe is positioning itself to not only meet current demands but also to innovate and lead in semiconductor technologies. This is particularly relevant for those involved in fields like automotive technology, IoT, and secure identification, where reliable and advanced semiconductors are key. [How ZEISS and ASML Enable the Modern Chip IndustryFrom Dutch roots to German optics, ASML and ZEISS built a unique European stronghold in chipmaking—where light carves the future.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-44.png)Rob Hoeijmakers (robhoeij)Rob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_1730.jpeg)](https://rob-hoeijmakers.ghost.io/zeiss-and-asml/) ## Looking Ahead The establishment of the Dresden plant is a milestone, but it's just the beginning. As the ESMC continues to evolve, it will play a critical role in shaping the future of Europe's tech industry. For digital professionals, staying informed about these developments is essential. Not only does it provide insights into industry trends, but it also highlights the importance of supporting and investing in local semiconductor manufacturing capabilities. In conclusion, the ESMC represents a significant step towards a more secure and independent semiconductor supply chain in Europe. By fostering collaboration among leading tech companies and securing substantial investments, Europe is well on its way to building a more resilient and innovative semiconductor industry. For digital professionals, this means a more stable foundation for future technological advancements and a more competitive market landscape. \--- *Sources:* - [Nikkei Asia](https://asia.nikkei.com/Business/Tech/Semiconductors/TSMC-to-start-construction-of-first-European-chip-plant-in-August) - [Euronews](https://www.euronews.com/2023/05/16/microchip-giant-tsmc-to-open-first-ever-european-plant) - [Investopedia](https://www.investopedia.com/tsmc-bosch-infineon-nxp-enter-joint-venture-6644948) - [Tech Wire Asia](https://techwireasia.com/2023/08/tsmc-to-open-first-europe-chip-plant-in-germany/) ### Review of "Decisive" by Chip Heath and Dan Heath URL: https://rob-hoeijmakers.ghost.io/review-of-decisive-by-chip-heath-and-dan-heath/ Last updated: 2026-05-14T21:02:30.000Z "Decisive: How to Make Better Choices in Life and Work" by Chip Heath and Dan Heath is a transformative guide to improving decision-making processes. The book introduces the WRAP framework—Widen Your Options, Reality-Test Your Assumptions, Attain Distance Before Deciding, and Prepare to Be Wrong—which offers practical strategies to overcome common decision-making pitfalls. ### Avoid whether or not decisions One of the book's key takeaways is the differentiation between various types of decisions. The Heath brothers highlight that "whether or not" decisions, also known as yes/no decisions, often limit our thinking and can lead to suboptimal outcomes. These decisions are simply an up-or-down vote on a single alternative, such as "Should I break up with my girlfriend?". By expanding our options and considering a broader range of possibilities, we can make more informed and balanced decisions. ### Statements of resolve Additionally, the book distinguishes "statements of resolve," where there are basically no decisions involved, from actual decision-making processes. Statements like "I am going to stop blaming others" reflect a commitment rather than a choice between alternatives. Recognizing these distinctions helps clarify the decision-making landscape and enables more effective strategies. ### Multitracking Multitracking in decision-making involves working on several options simultaneously, which helps to gather diverse feedback and reduces personal attachment to any single idea. This approach leads to more efficient decision-making by exploring multiple possibilities and maintaining flexibility, as it encourages considering a wider range of options and reduces biases. The critical point highlighted is that **multitracking keeps egos in check**. When a boss has multiple projects, they are more likely to accept unvarnished feedback. Conversely, with only one project, their ego becomes conflated with the project, making it harder to hear the truth. Despite its benefits, many organizations avoid multitracking due to concerns about time consumption. However, research by Kathleen Eisenhardt suggests that exploring multiple options can actually be more efficient, particularly in environments that prioritize speed, like Silicon Valley. ### Checklists and playlists In the chapter titled "Find Someone Who’s Solved Your Problem," the authors discuss the usefulness of checklists and playlists. **Checklists** are prescriptive tools, ideal for situations where the same behaviours need to be replicated consistently. They help prevent errors by ensuring that all necessary steps are followed systematically. In contrast, **playlists** are generative tools that stimulate new ideas and help prevent the oversight of potential options. They are particularly useful when shifting mindsets, such as moving from a prevention-focused approach to a promotion-focused approach. Playlists facilitate multitracking, allowing decision-makers to explore various opportunities and strategies, especially in challenging situations like budget cuts. The chapter emphasizes the value of using playlists to find innovative solutions by looking at how others have tackled similar problems, thereby broadening the range of potential strategies and outcomes. ### WRAP model The WRAP model is particularly valuable for those who struggle with generating choices, making decisions, or taking definitive actions. It encourages a structured approach, from exploring multiple alternatives to testing assumptions and preparing for potential errors. This method not only enhances the quality of decisions but also builds confidence and clarity in the decision-making process. ### WRAP Steps: 1. **Widen Your Options**: Expand your choice set to avoid narrow framing. 2. **Reality-Test Your Assumptions**: Gather information to validate your assumptions. 3. **Attain Distance Before Deciding**: Gain emotional distance to make more objective choices. 4. **Prepare to Be Wrong**: Plan for potential failures and unexpected outcomes. In summary, "Decisive" is an essential read for anyone looking to refine their decision-making skills. The practical tools and insights provided by the Heath brothers can help transform how we approach choices in both personal and professional contexts, moving beyond simplistic yes/no decisions to more comprehensive and effective strategies. I personally made big strides in how to view decisions, how to make them and base success on the outcomes! ### Taming Data Usage with a MiFi hotspot: My Experience and Solution URL: https://rob-hoeijmakers.ghost.io/taming-data-usage-with-a-mifi-hotspot-my-experience-and-solution/ Last updated: 2026-05-10T08:54:10.000Z During a recent holiday in the mountains, I relied on my [TP-Link M9700 MiFi](https://rob-hoeijmakers.ghost.io/a-mifi-router-for-my-mountain-holidays/) to stay connected. While the MiFi, or hotspot, worked wonderfully, I quickly discovered that managing data usage was a significant challenge. Here's how I tamed my data consumption and found an effective solution. ## The Challenge: Data Usage on MiFi Using a MiFi device turns your phone into a WiFi-connected device rather than a mobile device. This distinction is crucial because phones tend to consume more data when connected to WiFi. Background processes, automatic updates, and cloud syncing can quickly eat through your data allowance, especially when you're on a limited or throttled plan. ## The Data Drain Connecting my iPhone to the MiFi network resulted in significant data usage. Here’s what was happening: - **iCloud Syncing**: Photos and videos were continuously uploading to iCloud. - **Background Processes**: Apps were refreshing and syncing data in the background. - **Automatic Updates**: Apps and system updates were downloading automatically. Despite enabling Low Power Mode and turning off background app refresh, my phone continued to use a large amount of data, especially overnight when charging disabled the Low Power Mode. ## The Breakthrough: Data Saver for WiFi After several attempts to manage data manually, I discovered the key solution: enabling Data Saver mode specifically for the WiFi network created by the MiFi device. ### Steps to Enable Data Saver Mode for WiFi 1. **Access WiFi Settings**: Go to your iPhone’s Settings and select WiFi. 2. **Select Your MiFi Network**: Tap on the information icon (i) next to your MiFi network. 3. **Enable Low Data Mode**: Toggle the Low Data Mode switch to on. This setting restricts background data usage and reduces the frequency of syncs and updates, effectively managing data consumption. ## The Results Enabling Low Data Mode for my MiFi network significantly reduced my data usage. I no longer faced unexpected data drains overnight, and my daily data allowance from Vodafone lasted much longer. Here are some specific improvements I noticed: - **Controlled Background Activity**: Background processes were minimized, saving data. - **Reduced Cloud Syncing**: Photos and videos were no longer uploading continuously. - **Optimized Data Use**: Automatic updates and other high-data activities were restricted. ## My Personal Conclusion Using a MiFi device can turn your phone into a data-hungry WiFi device, leading to unexpected data usage. By enabling Data Saver mode for the WiFi network created by the MiFi, you can effectively manage and reduce your data consumption. This simple yet powerful setting allowed me to stay within my daily data limits and enjoy uninterrupted internet access during my mountain holiday. If you find yourself struggling with data usage on a MiFi, give this solution a try. It’s a game-changer for managing data and ensuring a smooth online experience, even with limited data plans. ### Further reading - [A Mifi hotspot for my mountain holidays](https://rob-hoeijmakers.ghost.io/a-mifi-router-for-my-mountain-holidays/) ### Identity and Addressability in the Digital Age URL: https://rob-hoeijmakers.ghost.io/identity-and-addressability/ Last updated: 2026-05-10T08:54:10.000Z For the past few months, I have been deeply contemplating the concept of identity, writing inventory articles and analysing different aspects. It was only recently that I realised my perspective had been fundamentally flawed. I had been viewing identity from a limited angle, which prevented me from seeing the broader picture. I was looking for the dark with a flashlight, and it took a pivotal story to shift my understanding. ## The TV Show Revelation: Addressability and Vulnerability A TV show titled "Hunted” fundamentally reshaped my understanding of identity. The show challenged contestants to evade powerful government agencies by hiding out. They had to surrender their phones and go into hiding. Naturally, many contestants acquired anonymous, disposable phones—burners, as they're commonly called—to get back online and stay connected. What the contestants didn’t realise was that the authorities could tap not only their phones but also the phones of their contacts. By handing over their phones, they inadvertently provided the searchers with a comprehensive list of their contacts. With triangulation and surveillance cameras, the contestants were quickly located and captured. This story highlighted the crucial role of being addressable in our identity. Even with a new phone, the network of connections remained a vulnerability. ## Rethinking Identity: The Role of Digital Addresses This revelation led me to rethink identity in the digital age. Our digital identity is intricately tied to how we are addressable. Upon examining my digital presence, I identified three main addresses that define how I am perceived and reached: 1. **Email Address** 2. **Phone Number** 3. **Handles and Account Names** Interestingly, the phone number feels the least like an address. However, it serves as a public address in platforms like WhatsApp and partially in iMessage. Despite its importance, the choice in defining this address is limited. The phone number is typically assigned by the phone company, and changing it can be cumbersome. It's linked to digital address books and often associated with a city or country. ![My bio landing page. ](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/07/Address-book-2.jpg) My bio landing page. ## Addressability and Flexibility: A Personal Experiment To enhance my digital identity, I [subscribed to a virtual number](https://rob-hoeijmakers.ghost.io/whatsapp-messenger-and-business-app-on-the-same-iphone-with-a-virtual-number/) with an Amsterdam city code, forwarding it to my mobile. This service included functionalities like WhatsApp integration. However, the service provider doubled the prices, making it unsustainable. This experience highlighted the need for flexible and affordable addressability options. Ideally, such functionalities should be bundled with other telecom services. For the amount we pay to telecom companies and tech platforms (Google or Apple), it’s reasonable to expect comprehensive services that cater to our identity and addressability needs. ## My personal view Understanding identity requires shifting perspectives and considering how we are perceived and addressed by others. In the digital age, this means paying attention to our email addresses, phone numbers, and online handles. As our reliance on digital communication grows, the need for flexible and integrated addressability options becomes increasingly critical. By rethinking how we manage our digital addresses, we can better navigate the complexities of our online identities. ### Further reading - [Wero: Europe’s new payment brand](https://rob-hoeijmakers.ghost.io/wero-europes-new-payment-brand/) - [WhatsApp in 2026: From Phone Numbers to Names](https://rob-hoeijmakers.ghost.io/whatsapp-in-2026-from-phone-numbers-to-names/) ### HDR: A Technology and a Format URL: https://rob-hoeijmakers.ghost.io/hdr-a-technology-and-a-format/ Last updated: 2026-04-25T13:17:09.000Z As web professionals, staying on the cutting edge of technology is crucial to enhancing both our content production and consumption. One such advancement is High Dynamic Range (HDR) technology, which has been making waves in photography and videography. But what exactly is HDR, and how can we leverage it to our advantage? Let's dive into the dual nature of HDR as both a technology and a format. ## HDR as a Technology HDR technology is designed to capture a broader range of light and dark areas in an image, resulting in more realistic and visually appealing photos. This process involves taking multiple shots at different exposure levels and merging them to create a single image with enhanced detail in both the shadows and highlights. This method ensures that no part of the image is too dark or too bright, providing a balanced and rich visual experience. For web professionals, understanding HDR technology can be a game-changer. When creating content, HDR allows for higher quality images that can significantly improve user engagement. HDR photos can showcase products, landscapes, or any subject matter with a level of detail and vibrancy that standard dynamic range (SDR) photos cannot match. [High Dynamic Range (HDR) in photo’sWhat is High Dynamic Range and how can I put this technology to use for rich images.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/size/w256h256/2024/05/Logo-Rob-Hoeijmakers-Square-Dark-Glow-3-1.png)Web StrategiesRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/size/w1200/2023/02/IMG_9335--1-.jpeg)](https://rob-hoeijmakers.ghost.io/high-dynamic-range-hdr/) ## HDR as a Format HDR is not just a technology; it is also a format that defines how images and videos are stored and displayed. Formats like [HEIC (High Efficiency Image Coding)](https://rob-hoeijmakers.ghost.io/heic-a-blood-type-for-images/), which Apple uses for photos on iPhones, support HDR content. This format allows for efficient storage while maintaining high-quality visuals. When publishing HDR content on the web, it's essential to consider the compatibility of different formats. For instance, platforms like Instagram may not fully support HDR content, leading to a loss of quality when HDR photos are uploaded. Understanding how different platforms handle HDR can help web professionals make informed decisions about where and how to publish their content to maintain the highest quality. ## Practical Application for Web Professionals To leverage HDR effectively, web professionals should consider both the creation and distribution aspects: 1. **Content Creation**: Use HDR technology to capture photos and videos. Modern smartphones and cameras often have built-in HDR capabilities. For example, when you take an HDR photo on an iPhone, the device captures multiple exposures and combines them to create a single HDR image stored in the HEIC format. 2. **Content Distribution**: When publishing HDR content online, ensure compatibility with the platform and devices. If a platform like Instagram does not support HDR fully, consider alternative ways to share your content, such as embedding HDR images on your website or using platforms that support HDR. 3. **Technical Knowledge**: Familiarize yourself with the technical aspects of HDR formats and metadata. This knowledge will help you optimize your images for different platforms and ensure the best possible quality for your audience. 4. **Testing and Adaptation**: Test your HDR content across various devices and browsers. Not all browsers fully support HDR, so providing fallback options in SDR can ensure a consistent experience for all users. ## HDR from a Consumer Perspective As consumers, we can sometimes be caught off guard by the super-bright HDR reels on platforms like Instagram, especially when viewing them in low-light conditions. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/07/HDR-Brightness.jpg) Left low light HDR, right high light HDR. As you can see, it is hard to dim HDR and so a bit of a shocker at night. This can be quite jarring, disrupting a calming experience. To mitigate this, it’s possible to disable HDR photo and video playback on Instagram for iPhones. For detailed steps on how to do this, refer to this [article on Another Apple](https://www.anotherapple.com/2024/06/disable-hdr-photo-and-video-playback-on-instagram-for-iphone/). It was exactly the sort of information I needed to lay this new piece of the puzzle. ### Understanding the Terms: Web, Online, Digital, and Cyber URL: https://rob-hoeijmakers.ghost.io/web-online-digital-and-cyber/ Last updated: 2026-05-10T08:54:11.000Z When I stumbled upon the Dutch Defence Forces seeking reservists for their cyber warfare command, it made me rethink the terms we often use interchangeably: "web," "online," "digital," and "cyber." It was a wake-up call that I should have addressed earlier, but better late than never. Clarifying these terms not only helps me navigate these topics more effectively but also solidifies my identity as a [web strategist](https://rob-hoeijmakers.ghost.io/the-web-strategist-and-the-web-strategy/). **Web**: For me, the "web" is where it all began. It refers to the World Wide Web, a system of interlinked hypertext documents accessed via the Internet. This is where websites reside and interact, creating a dynamic space for information and communication. As a web strategist, I focus on crafting and optimizing websites to foster human connection and meaningful interactions. The web is my playground, where content and ideas flow freely. **Online**: Being "online" means being connected to the Internet, covering activities like web browsing, email, social media, and online services. While important, "online" is a broader term that doesn't capture the specific realm I operate in. My work is deeply rooted in the web, where the focus is on creating engaging content and user experiences. **Digital**: The term "digital" encompasses any technology using digital signals, from computers and smartphones to smart home devices. Digital strategists often integrate various technologies to create a cohesive digital presence. However, my passion lies in the web, where the emphasis is on content strategy and human connection rather than just the technology itself. 💡 Here’s how I’ve come to define the terms in short: ****Web** – Connecting people ****Online** – Anytime, anywhere ****Digital** – Transforming processes ****Cyber** – Protecting and securing **Cyber**: "Cyber" is often associated with the virtual world of the Internet and computer networks, focusing on cybersecurity, cybercrime, and the technological aspects of digital activities. When I considered the cyber warfare command, I realized how specialized and security-focused this area is. While crucial, it doesn't align with my primary interest in fostering communication and growth through the web. [Cybernetics: The Idea Behind CyberBefore it meant computers, cyber meant steering. This piece follows that idea from its origins to the way it now shapes states, platforms, and our digital lives.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-385.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/01BD84B5-D04F-434C-A5DB-15F88FA782EE.jpeg)](https://rob-hoeijmakers.ghost.io/cybernetics/) ## Why "Web Strategist" is the Right Choice for Me Choosing to call myself a web strategist wasn't just a professional decision; it was a reflection of my values and interests. **Human Connection Over Channels**: My work centers on creating meaningful connections through the web. Unlike an "online strategist" who might focus broadly on all Internet activities, I prioritize how websites and web content can foster human engagement and interaction. **Information Over Transaction**: A "digital strategist" might lean towards integrating various technologies and focusing on transactional aspects. In contrast, I emphasize the importance of information, content, and helping individuals and organizations grow and flourish through the web. **Strategic Analysis and Implementation**: As a web strategist, I analyze current web trends and activities for client organizations, develop comprehensive plans to align with their goals, and collaborate with them to put these plans into motion. This holistic approach ensures my strategies are not only visionary but also actionable and tailored to my clients' needs. **Focus on Content and Message**: Unlike "cyber strategists," whose primary concern might be the security and technical integrity of digital environments, my focus is on crafting and delivering the right message. My expertise lies in content strategy, [user experience](https://rob-hoeijmakers.ghost.io/user-experience-ux-and-ux-design/), and ensuring that web interactions are meaningful and impactful. > **Sidenote**: I started 30 years ago as a webmaster. It's an old-fashioned term, a bit romantic even. But throughout my career, the web has been a constant presence. This evolution from webmaster to web strategist feels logical to me. When I began, the web was more of a side project; now, it has moved to the core of so many things we do. This progression reflects not only the growth of the web itself but also my deepening expertise and focus in this ever-evolving field. ## Integrating Large Language Models (LLMs) into Web Strategy My interest in [Large Language Models (LLM's)](https://rob-hoeijmakers.ghost.io/a-comparative-analysis-of-top-language-models/) significantly enhances my role as a web strategist. These advanced AI tools allow me to: - **Enhance Content Creation**: LLMs can generate high-quality content tailored to specific audiences, ensuring that the language and message resonate deeply with users. - **Improve User Engagement**: By analyzing user interactions and feedback, LLMs help refine content and user experiences, making websites more engaging and effective. - **Streamline Communication**: LLMs can assist in automating and optimizing communication strategies, from customer support to personalized marketing messages, enhancing overall user satisfaction. - **Data-Driven Insights**: Leveraging LLMs allows for the analysis of vast amounts of data to uncover trends and insights that inform strategic decisions and content development. [Web Traffic and the Rise of LLMsExploring the impact of LLMs on web traffic, a shift from Google’s reign to a new era of digital engagement.![](https://static.ghost.org/v5.0.0/images/link-icon.svg)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/size/w1200/2024/03/IMG_0375-1-1-1-1.jpeg)](https://rob-hoeijmakers.ghost.io/web-traffic-and-the-rise-of-llms/) ## My Role as a Web Strategist Being a [web strategist](https://rob-hoeijmakers.ghost.io/the-web-strategist-and-the-web-strategy/) means being a visionary and a planner. I: - **Analyze**: Assess the current state of a client's web presence, identifying strengths, weaknesses, opportunities, and threats. - **Plan**: Develop strategic plans that align with the client's future goals and vision, ensuring web activities support broader business objectives. - **Collaborate**: Work closely with client organizations to implement these plans, providing guidance and support throughout the process. - **Execute**: Oversee the execution of web strategies, ensuring all elements are put into motion effectively and efficiently. Focusing on the web allows me to weave together content, design, user experience, and strategy into a cohesive and engaging web presence for my clients. While "online," "digital," and "cyber" each have their place, "web strategist" best captures the essence of my work. It highlights my commitment to language, human connection, and the strategic use of the web to help organizations grow and thrive. Realizing the nuances of these terms was a journey, sparked by an unexpected encounter with the cyber warfare command. It helped me clarify my focus and reaffirm that being a web strategist aligns perfectly with my passion and expertise. Integrating LLMs into my work further enhances my ability to create meaningful connections and deliver impactful content, making the web a powerful tool for growth and communication. --- ### Further reading - [About Rob Hoeijmakers — AI Strategist](https://rob-hoeijmakers.ghost.io/about/) ### Apple’s RCS Adoption in iOS 18 URL: https://rob-hoeijmakers.ghost.io/apples-rcs-adoption-in-ios-18/ Last updated: 2026-04-25T13:16:36.000Z Apple's announcement of Rich Communication Services (RCS) support in iOS 18, slated for release in September, marks a pivotal shift in the messaging landscape. For years, iOS and Android users have experienced a [fragmented messaging experience](https://rob-hoeijmakers.ghost.io/chat-vs-messages-whats-the-difference/) due to the limitations of SMS. With the adoption of RCS, Apple aims to bridge this gap, offering a more unified and feature-rich communication platform. ## What is RCS? RCS is a modern messaging protocol designed to replace SMS. It brings several advanced features to the table, including: - **Typing Indicators**: Users can see when someone is typing a message. - **Read Receipts**: Notifications when a message is read. - **High-Quality Media Sharing**: Ability to share images and videos in higher resolution. - **Support for Longer Messages**: No more splitting long messages into multiple parts. These features are akin to what iMessage offers but designed to work across different platforms, ensuring a seamless experience between iOS and Android users. Informative and interactive demo of how RCS will work on iOS18. ## Why Now? The timing of Apple's decision to adopt RCS is influenced by multiple factors: - **Regulatory Pressure**: The European Union has been pushing for greater interoperability between messaging platforms to avoid monopolistic practices. By supporting RCS, Apple aligns with these regulatory requirements, addressing concerns about its closed ecosystem. It might be also the Chinese government that is using pressure that has made Apple changed its mind. That is the conclusion [John Gruber makes in his excellent article on Daring Fireball](https://daringfireball.net/2024/02/eu%5Frcs%5Fimessage). - **Industry Pressure**: Google and other industry players have long advocated for RCS to improve cross-platform messaging. Apple's adoption can be seen as a response to these efforts and a step towards greater user satisfaction and platform compatibility. ## RCS vs. WhatsApp and Facebook Messenger One might wonder if RCS is a direct competitor to messaging giants like WhatsApp and Facebook Messenger. Here’s a comparison: - **Functionality**: RCS offers features like typing indicators, read receipts, and high-quality media sharing, similar to WhatsApp and Messenger. However, WhatsApp and Messenger are deeply integrated with Meta’s services, offering additional functionalities like social networking and encrypted messaging. - **Encryption**: WhatsApp’s end-to-end encryption ensures that only the communicating users can read the messages. Initially, RCS on iOS will not have this level of encryption, although Apple plans to enhance security features over time. - **Platform Independence**: RCS aims to be a universal standard across different devices and carriers, potentially simplifying cross-platform communication without needing third-party apps. However, WhatsApp and Messenger have already established a vast user base. ## Identification in RCS RCS identifies users primarily through their phone numbers, similar to SMS and MMS. This approach simplifies the user experience by integrating seamlessly with the native messaging app and eliminating the need for additional accounts or logins. Here’s a breakdown: - **Phone Number Based**: Messages sent via RCS are tied to your phone number, ensuring compatibility with your existing contact list. - **Network Dependency**: Both users' networks must support RCS for it to function. In regions where carrier support is lacking, messages will default to SMS. [Select US carriers introduce support for RCS on iOS 18 betaThe second developer beta of iOS 18 introduced a toggle for RCS in Messages, and now select US carriers have made RCS available to users running the beta.![](https://photos5.appleinsider.com/v9/images/apple-touch-icon-144.png)AppleInsiderMarko Zivkovic![](https://photos5.appleinsider.com/gallery/60171-123581-RCS-toggle-xl.jpg)](https://appleinsider.com/articles/24/06/25/select-us-carriers-introduce-support-for-rcs-on-ios-18-beta) ## Personal Reflection Reflecting on the evolution from MMS to RCS, it's evident that the tech world has learned valuable lessons. MMS's shortcomings paved the way for apps like [WhatsApp](https://rob-hoeijmakers.ghost.io/whatsapps-global-reach/) to dominate the messaging space by offering reliability and ease of use. Now, with RCS, the goal is to provide these rich features natively, thus reducing reliance on third-party apps and offering a more cohesive experience across different platforms. The inclusion of RCS in iOS 18 is a significant development that promises to enhance the messaging experience for millions of iPhone users worldwide. While RCS may not immediately dethrone WhatsApp or Facebook Messenger, it represents a significant step towards making rich, cross-platform messaging a standard feature for all users, without the setup headaches that plagued earlier technologies like MMS. This move by Apple is a pivotal moment in the evolution of mobile messaging, potentially setting a new standard for cross-platform communication. ### Locked Out of Instagram or Threads? Avoid and appeal a ban. URL: https://rob-hoeijmakers.ghost.io/locked-out-banned-instagram-threads/ Last updated: 2026-05-10T08:54:11.000Z In the whirlwind of today's digital age, a single [viral photo](https://rob-hoeijmakers.ghost.io/viral-photo/) can transform a casual hobby into a serious endeavour. This is exactly what happened to me when my Instagram account, initially a leisurely pursuit, exploded in popularity. This unexpected success opened up a substantial channel and served as an important social signal for search engines, giving all my online properties a significant boost. However, with this growth comes new responsibilities and challenges. ## We don’t own, we rent One of the most sobering realisations is that we don't truly own our social media accounts; we are essentially renting space from platforms like Meta. This means adhering to their guidelines and policies, with the risk that they can change the rules or restrict access at any time. ## Warnings in Threads Personally, I had a sort of a scare with two innocent messages on [Threads](http://threads.net/@robhoeij) triggering an algorithmic check, leading to a temporary block. Although I was able to appeal successfully, the experience was somewhat unnerving and highlighted how easily an active account can be flagged. ![Left the screen I got to see in Instagram when a Threads post triggers an alarm, and right the screen in Instagram after loads of activity.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/06/Instagram-account---Locked-out.jpg) Left the screen I got to see in Instagram when a Threads post triggers an alarm, and right the screen in Instagram after loads of activity. ## Tips for Avoiding Trouble Keep your account safe and trouble-free with these essential practices: 1. **Adhere to Platform Guidelines**: Regularly review and comply with terms of service and[ community guidelines](https://help.instagram.com/477434105621119?helpref=faq%5Fcontent). 2. **Limit the Use of Third-Party Tools**: Use only approved tools and avoid excessive automation. 3. **Avoid Suspicious Activities**: Maintain a consistent IP address and avoid spamming. 4. **Secure Your Account**: Use strong passwords and enable [two-factor authentication](https://rob-hoeijmakers.ghost.io/two-factor-authentication-2fa/). 5. **Monitor Account Activity**: Regularly check for unusual activity and track engagement metrics. [Login • InstagramWelcome back to Instagram. Sign in to check out what your friends, family & interests have been capturing & sharing around the world.![](https://static.cdninstagram.com/rsrc.php/v3/yG/r/De-Dwpd5CHc.png)](https://help.instagram.com/477434105621119) The same guidelines apply to Instagram and Threads. They are clear and well readable. ## Steps to Take If Your Account Faces a Ban If your account gets banned, take these steps to respond calmly and effectively: 1. **Stay Calm and Review Notifications**: Understand the reason for the ban and document everything. 2. **Appeal the Decision**: Follow the platform's official appeal process with a clear explanation. 3. **Reach Out to Support**: Contact [customer support](https://help.instagram.com) and use social media channels for assistance. 4. **Leverage Your Network**: Seek help from connections or the community. 5. **Explore Alternatives**: Maintain backup accounts and diversify your presence on [multiple platforms](https://rob-hoeijmakers.ghost.io/navigating-social-media-my-experience-across-platforms/). 6. **GDPR Requests**: For [EU users](https://rob-hoeijmakers.ghost.io/gdpr/), request data access to understand the issue and preserve your data. ## Owning Your Digital Assets While it's essential to navigate social media platforms carefully, it's equally crucial for creators to develop assets they have more ownership of: 1. **Personal Website or Blog**: Establish a website or blog (like this one) where you control the content and user experience. 2. **Mailing List**: Build and maintain a[ healthy mailing list](https://rob-hoeijmakers.ghost.io/newsletter/) to communicate directly with your audience without platform interference. 3. **Content Diversification**: Spread your content across [multiple platforms](https://rob-hoeijmakers.ghost.io/navigating-social-media-my-experience-across-platforms/) and formats to reduce dependency on any single service. While the growth and opportunities offered by social media are immense, they come with inherent risks. As users, we must navigate these platforms with caution, be aware of our vulnerabilities, and explore ways to safeguard our digital presence. I will keep coming back to this article for updates while new things happen. Feel free to [nudge me](https://rob-hoeijmakers.ghost.io/about/) when you have tips, tricks or your own (scary) adventures. [Facebook ProtectA viral photo surge led to Facebook Protect. Meta’s added security made me realise how vulnerable our digital identities are—and how much more I could do with them.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-60.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/fb050fdd-65bf-495e-bb3f-6eb9e940e375-1.jpeg)](https://rob-hoeijmakers.ghost.io/facebook-protect/) --- ## Suspension and shadow banning check list I did a couple of searches through Threads to make an inventory from real user stories. Instagram and Threads can throttle your reach (shadow banning) or suspend your account for several reasons. ### 1\. Duplicate Content Posting content that closely resembles the original creator's work can result in your posts not being suggested by the algorithm. Instagram values original content and may throttle the reach of posts deemed too similar to existing content. I experienced this when I forwarded something on Threads with just a summary. Such a thing would be no problem on X but not on Threads. Makes sense. ### 2\. Political Content Content based on political news or related topics might not be suggested on the Explore search feed or Reels. Instagram often limits the reach of politically sensitive content to avoid misinformation and political bias. ### 3\. Violent or Harmful Language Using language that can be interpreted as violent, even jokingly, can lead to account suspension. For example, typing “I was going to m\*rder my legs” can trigger automated reviews and penalties if reported by other users. The person sending that message got suspended for 90 days. ### 4\. Banned Hashtags Using hashtags that have been banned by Instagram can severely limit your reach or lead to account issues. Banned hashtags are often associated with inappropriate or harmful content, and their use can result in posts being hidden or accounts being flagged. Always check for current banned hashtags and avoid using them. Check for: [Instagram banned Hashtags](https://metahashtags.com/banned-hashtags/) ### 5\. Inappropriate Profile Photos Profile photos that violate community guidelines, such as those showing nudity or see-through clothing, can trigger AI reviews and lead to account throttling or suspension. Updating your profile photo to comply with guidelines can help resolve this issue. ### 6\. Spammy Behaviour Activities such as posting too frequently, following/unfollowing a large number of accounts in a short period, or using repetitive comments can be considered spammy behaviour. This can lead to reach throttling or temporary account suspensions. Instagram aims to maintain a genuine user experience and actively monitors for spam-like activities. ### 7\. Community Guidelines Violations Any content that violates Instagram's community guidelines, including hate speech, harassment, and misinformation, can result in shadow banning or account suspension. Regular violations can lead to more severe penalties, including permanent bans. ### 8\. Automated Actions Using bots or automated services to increase followers, likes, or comments can result in account throttling or suspension. Instagram actively monitors and penalizes such behaviour to ensure authentic engagement on the platform. ### 9\. User Reports Accounts reported by other users for any reason, such as inappropriate content or behaviour, can be reviewed by Instagram. If the reports are found valid, the account can face penalties, including suspension. It's crucial to adhere to community standards to avoid being reported. ### Further reading - [Passports, Phones, and the Future of Identity](https://rob-hoeijmakers.ghost.io/passports-phones-and-the-future-of-identity/) - [Locked Out by X](https://rob-hoeijmakers.ghost.io/locked-out-by-x/) ### Navigating Social Media: My Experience Across Platforms URL: https://rob-hoeijmakers.ghost.io/navigating-social-media-my-experience-across-platforms/ Last updated: 2026-05-10T08:54:11.000Z As someone who thrives in the digital space, I’ve found mini-blogging on multiple social media platforms both fascinating and rewarding. Each platform has its own distinct vibe and community, offering unique opportunities to connect and share ideas. Here’s a rundown of my experiences on [X](https://www.twitter.com/robhoeij), [Threads](https://www.threads.com/@robhoeij), [Mastodon](http://mstdn.social/@robhoeijmakers), and [BlueSky](https://bsky.app/profile/hoeijmakers.net), along with some insights from a recent poll I conducted on generational demographics across these platforms. ## X: A Curiosity-Driven Playground On X (formerly known as Twitter), I post about a wide range of topics, driven mostly by my curiosity. Whether it's the latest trends in web strategies, random musings about daily life, or insightful observations about technology, X is my go-to for spontaneous sharing. I engage with a diverse audience here, and the platform’s fast-paced environment suits my need for quick, frequent updates. Interestingly, I’ve noticed a significant presence of Baby Boomers on X, which adds a unique flavour to the discussions. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/06/IMG_4039.jpeg) X, with 215 votes in the poll ## Threads: Where Web Strategies and Photography Meet Threads is where I dive deeper into discussions about web strategies, blending professional insights with my love for photography. The integration with Instagram allows for a rich cross-pollination of content, making it an ideal space for visual storytelling alongside strategic conversations. Millennials and Gen-Z users dominate this platform, which infuses the discussions with fresh perspectives and innovative ideas. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/06/IMG_4040.jpeg) Threads, with 186 votes ## Mastodon: Thoughtful Conversations in a Professional Space Mastodon feels like a breath of fresh air with its more thoughtful and deliberate user base. Here, I focus on in-depth discussions about social media, artificial intelligence, and web strategies. The conversations tend to be more reflective and analytical, aligning well with my professional interests. Gen-X users seem to be prevalent on Mastodon, fostering a mature and insightful dialogue that I find incredibly enriching. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/06/IMG_4041.jpeg) Mastodon, with 56 voters ## BlueSky: The Challenge of Finding My Footing BlueSky is the one platform where I feel a bit like an outsider. Despite its potential, I haven’t quite found my niche or established strong connections there. The community feels different, and I’m still navigating how best to engage. It’s a work in progress, and I’m open to exploring new ways to connect with users on BlueSky. BlueSky does not have a poll feature. ## Poll Insights: Generational Demographics Across Platforms To better understand who’s hanging out where, I conducted a poll on generational demographics, and the results were quite enlightening. Although the poll isn’t scientifically rigorous and may be skewed by my own network, it provides some interesting insights: - **Gen-X Dominance:** There’s a noticeable presence of Gen-X across the platforms I use, likely influenced by my own age and professional circle. - **Baby Boomers on X:** X has a higher percentage of Baby Boomers compared to the other platforms, which might explain the more eclectic mix of conversations. - **Millennials and Gen-Z on Threads:** Threads attracts a younger audience, with Millennials and Gen-Z users being more prevalent. This demographic brings a vibrant and dynamic energy to the platform, fostering innovative discussions and creative content. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/06/Social-Media-demographics-2.jpg) ## Final Thoughts Social media is a constantly evolving landscape, and each platform offers its own set of opportunities and challenges. By understanding the unique atmospheres and demographics of X, Threads, Mastodon, and BlueSky, I can tailor my content and engagement strategies to better connect with each community. Whether it’s sharing curiosity-driven posts on X, blending web strategies with photography on Threads, delving into thoughtful conversations on Mastodon, or finding my way on BlueSky, the journey is both exciting and enlightening. I’d love to hear your thoughts on these platforms. Where do you feel most at home, and what’s your experience been like navigating these different digital spaces? [Let’s continue the conversation, on the platform of your choice!](https://rob-hoeijmakers.ghost.io/) --- ### Further reading - [Leveraging a Viral Photo: A Journey in SEO and Digital Strategy](https://rob-hoeijmakers.ghost.io/viral-photo-seo-digital-strategy/) ### Leveraging a Viral Photo: A Journey in SEO and Digital Strategy URL: https://rob-hoeijmakers.ghost.io/viral-photo-seo-digital-strategy/ Last updated: 2025-04-12T15:18:27.000Z In the ever-evolving digital landscape, having a substantial and cohesive online presence is crucial. As an online professional, I've experienced the highs and lows of managing a digital footprint. Recently, a viral photo and a strategic domain change significantly impacted my digital presence, providing valuable insights into [SEO](https://rob-hoeijmakers.ghost.io/my-basis-for-seo/) and digital strategy. I write this article so you can benefit from my experience and prepare your own online footprint for when luck strikes or you have something going on that gets positive attention. ## Background Over the years, I've organically grown a significant digital footprint. This includes multiple blogs, social media accounts, and appearances in third-party interviews, podcasts, and articles. Each platform has its unique audience and purpose, contributing to a complex but rewarding online presence. ## The Viral Photo and Domain Change A recent viral photo shook up my digital world. At the same time, [I transitioned my main blog from a domain](https://rob-hoeijmakers.ghost.io/moving-to-a-new-domain/) with my family name ([hoeijmakers](https://rob-hoeijmakers.ghost.io/)) to a more content-relevant domain ([webstrategiesblog.com](https://rob-hoeijmakers.ghost.io/)). This move aimed to avoid the complexities of my Dutch last name and better reflect the blog's content. ### Impact on Social Media The [viral photo](https://rob-hoeijmakers.ghost.io/viral-photo/) caused a surge in my social media following. My [Instagram](https://www.instagram.com/robhoeij) followers tripled to 3,300, and my presence on [Twitter/X](https://www.twitter.com/robhoeij) also saw significant growth. However, the ever-changing landscape of social media platforms, with the rise of [Mastodon](http://mstdn.social/@robhoeijmakers), [BlueSky](https://bsky.app/profile/hoeijmakers.net), and [Threads](https://www.threads.net/robhoeij), scattered my audience. Despite this, the increased attention provided an opportunity to consolidate and strengthen my online presence. ### Boost in Backlinks The story I wrote about the viral photo and how I managed the sudden influx of attention attracted numerous backlinks to my websites. This boost in backlinks not only increased my site's authority but also demonstrated the power of timely and relevant content. ## Mapping and Unifying My Digital Footprint Realising the need for a cohesive digital strategy, I mapped out my digital footprint. It became clear that my online presence was more of a network than a web, lacking a central hub to tie everything together. ![Mindmap to map out my digital footprint](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/06/2024-06-21_11-38-17.png) MindMap of my digital footprint. Analysing what is key to my goals, what is personal and which language is it in (English/Dutch) ### Creating a Central Hub To address this, I decided to utilise the root domain of my old site. I set up [hoeijmakers.net](https://rob-hoeijmakers.ghost.io/) as a temporary holding page using [Canva](https://rob-hoeijmakers.ghost.io/canva/), which quickly rose to the top of Google search results. While exploring bio landing pages like Linktr.ee, I realised they all operated from their domains. Wanting to maintain control, I found a WordPress plugin ([Social](https://socialwp.io)) that offered similar functionality. Licensing and building with this plugin allowed me to create a custom bio landing page on my own domain, effectively tying all my digital properties together. ![Launching Soon landing page in Canva, bio link page with WordPress and Social](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/06/Hoeijmakers.net.jpg) Launching Soon landing page in Canva, bio link page with WordPress and Social I'm now working with my [designer (Schmuki)](https://schmuki.nl/portfolio/) to develop a comprehensive website for my activities and personal brand. ## Streamlining My Online Presence This process led to a comprehensive cleanup of my profiles, photos, and banners. By unifying these elements, I enhanced the recognisability of my brand. Additionally, ensuring reciprocal links across my platforms signalled to Google the interconnectedness of my online presence, helping to establish a strong social graph. ![Updated social media accounts. X, LinkedIn and YouTube.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/06/Updated-social-media-accounts.jpg) Updated social media accounts. X, LinkedIn and YouTube. ## Strategic Considerations With a clear view of my digital footprint, I identified which areas to focus on for growth and which to maintain. This strategic approach ensures that my efforts are aligned with my goals and resources. ### Dual Goals: Influence and Content Marketing My primary goals are twofold: 1. **Influence**: Establish myself as an influencer in the digital content space. 2. **Content Marketing**: Use my blogs and photographs to market my web strategy profession. These goals complement each other, showcasing my ability to capture attention and effectively market in an information-rich but attention-poor digital world. ## My personal conclusion The journey of leveraging a viral photo and strategically managing a digital presence is both challenging and rewarding. By focusing on cohesive branding, strategic link-building, and targeted growth, I've been able to strengthen my online presence significantly. The fact that you're reading this article is a testament to the success of these strategies. For online professionals, this journey underscores the importance of adaptability, strategic planning, and continuous improvement in the ever-changing digital landscape. 💡 I was struggling a bit with Facebook. My photo was a great hit there but I did not want to open up my Facebook account because I use it for more personal stories in Dutch. When the photo went viral for the second time I created [a new Facebook page](https://www.facebook.com/robhoeij) and connected it to Instagram. ### Unpacking the World of Podcasting: An Interview with Larry Swanson (Part 1) URL: https://rob-hoeijmakers.ghost.io/unpacking-the-world-of-podcasting/ Last updated: 2026-04-24T18:35:16.000Z Recently, I had the pleasure of speaking with [Larry Swanson](https://www.larryswanson.com), a seasoned podcaster and digital strategist, to delve into the intricate world of podcasting. Our conversation was rich with insights, covering everything from production and distribution to audience engagement and measuring success. This blog post serves as a foundational layer for our discussion, setting the stage for a deeper analysis in Part 2. ## The Interview Overview I’ve always been fascinated by how people differ in their preferred ways of consuming information. Personally, I’ve always been more of a reader, often finding myself struggling with audio formats. Yet, the rise of podcasting intrigued me, and it was during my daily commute that I began exploring this medium. To kill time, I would listen to podcasts, finding them a more curated and purposeful alternative to the randomness of radio. Then I met Larry and he wanted me to talk on his [Podcast Content and AI](https://contentandai.com/rob-hoeijmakers/). He also agreed I could tap into his expertise in podcasting and provided me with wealth of knowledge. [Content + AI: demystifying and democratizing AI for content prosContent + AI aims to demystify artificial intelligence and democratize its use across the span of content practice.![](https://contentandai.com/wp-content/uploads/2023/11/cropped-CAI-just-waves-Meme-Square-270x270.jpg)Content+AILarry Swanson![](https://contentandai.com/wp-content/uploads/2023/10/Content-and-AI-cover-art-Podcast-Cover-Calendar-Youtube-Banner.jpg)](https://contentandai.com) ## The Nature of Podcasting, Audio vs. Broadcast I shared with Larry my initial struggles with audio formats and how I eventually came to appreciate podcasts during my commutes. Larry expressed his preference for the podcasting format over traditional radio broadcasting. He highlighted several key differences: - **Local vs. Global Reach**: Traditional radio targets local audiences, whereas podcasts can reach listeners worldwide. - **Engagement and Format**: Podcasts allow for deeper engagement and can be more personal and conversational compared to the structured nature of radio broadcasts. Larry articulated why the podcast medium suits him so well. He enjoys the **conversational nature of podcasting** and finds it aligns perfectly with his strengths and preferences. Larry mentioned, "I settled on the podcast format because it plays to my conversational nature. I also, I just, I can write well, but I just hate it. It’s like the hardest thing in the world, where just having a conversation is simple enough, you know?". This resonates deeply with me because Larry's interviews are always engaging and insightful, a testament to his natural fit with the medium. We also discussed how **the rise of headphone** use has created more opportunities for podcast consumption beyond the classic car commute scenario. Podcasts are now a part of our daily lives, whether we are commuting, exercising, or just relaxing at home. ## The Role of Audio in Learning and Exploration One fascinating aspect of podcasting is how it serves as a tool for learning and staying up-to-date with the ever-changing world around us. Unlike traditional classroom learning, podcasts offer a flexible and dynamic way to dive deep into specific topics. From content strategy and AI to digital marketing and personal development, podcasts allow listeners to explore a wide array of subjects at their own pace. I’ve noticed that while I am more inclined towards reading and writing—whether it’s taking notes or drafting articles—there is a significant portion of people who prefer to listen and learn. Podcasts cater to this audience, providing an accessible and often more engaging way to absorb information. They are perfect for those who want to stay informed about the latest trends and deep dive into complex topics, all while multitasking or on the go. ## Larry's Personal Journey into Podcasting Larry's journey into becoming a dedicated podcaster is as insightful as it is inspiring. He started his podcast primarily as a way to learn and connect with others in his field. Larry explained, "I’m a **genuinely curious person**, and I got in deep. I’ve always sort of trusted my gut and followed my instincts... I wanted to back up, look at the whole spectrum". This curiosity led him to talk to hundreds of people in the content strategy field, seeking diverse perspectives and insights. He also mentioned the practical benefits of podcasting, such as how it helps him **connect with both well-known experts and undiscovered talents**. This approach not only enriches his understanding but also broadens his network. Larry said, "It’s an explicit way to learn, but also just a way to connect with a whole bunch of people". [Content Strategy Insights: Interviews by Larry SwansonHalf-hour content strategy interviews with luminaries from the worlds of enterprise content, UX design, brand, product, support docs, etc.![](https://ellessmedia.com/wp-content/uploads/cropped-Elless-Media-site-icon-512x512-180x180.png)Elless MediaLarry![](http://ellessmedia.com/wp-content/uploads/CSInsights-logo-800.png)](https://ellessmedia.com/csi/) ## Brief Summary of Other Topics While the nature of audio was a central theme in our conversation, Larry and I also touched on several other important aspects of podcasting: - **Challenges in Podcast Distribution**: Larry shared his experiences with various tools and platforms essential for podcast distribution, such as [Libsyn](https://libsyn.com) for hosting and [Rev.com](https://www.rev.com) for transcriptions. He also emphasized the importance of syndication to ensure a global reach. - **Fragmentation of Social Media**: We discussed the challenges of engaging with audiences across different social media platforms like [Twitter (X)](https://x.com/LarrySwanson), [LinkedIn](https://www.linkedin.com/in/larryswanson/), Mastodon, and Threads. Each platform has its unique audience, requiring tailored engagement strategies. - **Production Insights**: Larry elaborated on his streamlined approach to podcast production, focusing on minimal editing to maintain the authenticity of the conversation. He ensures that each episode has a strong start and finish, engaging listeners right from the beginning. - **Measuring and Optimising Success**: Understanding podcast success can be challenging due to the nature of digital distribution. Larry tracks downloads via hosting platforms, uses social media for audience interaction, and values qualitative feedback from listeners. - **Questions to Larry and Future Directions**: Throughout our conversation, I had numerous questions for Larry, both as a consumer of podcasts and a fellow podcaster. We discussed Larry's personal journey, content strategy, and plans for monetization through platforms like Patreon. ## Stay tuned This foundational layer sets the stage for a deeper dive into the world of podcasting with Larry Swanson. In Part 2, we will explore more intricate details of podcast production, advanced distribution strategies, and further insights from Larry's extensive experience in the field. Stay tuned for the continuation of our insightful conversation and a deeper understanding of the ever-evolving podcasting landscape. ### Chat versus Messages: key differences explained URL: https://rob-hoeijmakers.ghost.io/chat-vs-messages-whats-the-difference/ Last updated: 2026-05-26T19:30:47.000Z In the digital communications landscape, "chat" and "messaging" are terms that are often used interchangeably. Yet they serve different purposes and contexts. Understanding the nuances can help online professionals choose the right tools for their needs. Let's dive into the key differences and explore different platforms that illustrate these two forms of communication. ## Definition of chatting and messaging **Chat is about real-time interaction.** Think of it as a digital conversation where an immediate or near-instant response is expected. It's like having a face-to-face conversation, but through a screen. You see this in tools like Slack, Discord, and live chats for customer service. Features such as presence indicators (which show who is online), typing notifications, and quick back-and-forth exchanges are hallmarks of chat. **Messaging**, on the other hand, is typically asynchronous. Messages can be sent and received at the user's convenience and no immediate response is expected. Email, text, and social media direct messages fall into this category. Messages allow for more thoughtful, detailed communication and are often used for formal or detailed exchanges. ![Chat window and WhatsApp](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/06/Chat-en-messaging.jpg) Chat window and WhatsApp ## Main differences at a glance When comparing chat and messaging, there are some key differences worth highlighting. These differences relate to how these forms of communication are used, their technical characteristics and the context in which they are most effective. Here's an overview of the main differences: - **Realtime vs. Asynchronous**: Chat is real-time; messaging is Asynchronous. Or, not at the same time. - **Platform and medium** : Chat is often platform-specific (Slack, Discord); messaging spans several platforms (email, SMS). - **Duration and history** : Chat histories can be temporary; message histories are usually persistent. - **Formality and context** : Chatting is more informal; messages can be formal. - **Features and functionalities** : Chat includes real-time features such as typing indicators; messages include features such as scheduled sending and detailed formatting. ## Overview of different clients and platforms Let's take a look at some popular chat and messaging platforms, highlighting their unique features and contexts in which they are good. ### Chat platforms Chat platforms provide real-time communication, ideal for dynamic interactions and collaboration. Below is an overview of some popular chat platforms, their use cases and key features: **Slack** - **Uses** : Team collaboration, project management. - **Features** : Channels, direct messaging, file sharing, integrations (Trello, Google Drive), searchable history, presence indicators, video calls. - **Context** : Mainly for professional environments and team communication. **Microsoft Teams** - **Use** : Business communication, collaboration. - **Features** : Channels, direct messaging, video conferencing, file sharing, Office 365 integration, wired conversations, presence indicators. - **Context** : Used within organizations for seamless integration with Microsoft's productivity suite. **Discord** - **Uses** : Gaming communities, hobby groups, small businesses. - **Features** : Servers with channels, voice and video calling, screen sharing, gaming platform integration, bots, rich media sharing. - **Context:** Initially for gamers, now for broader community and team use. ### Messaging platforms Messaging platforms offer asynchronous communication, ideal for both personal and business interactions. Below is an overview of some popular platforms, their use cases and key features: **Email (Gmail, Outlook)** - **Use** : Professional and personal communication. - **Features** : Asynchronous messaging, rich text formatting, attachments, filters, labels, calendar integration, search functionality. - **Context** : For formal communication and document sharing. **SMS** - **Uses** : Personal communications, marketing, warnings. - **Features** : Text messaging, multimedia messaging (MMS), long range, emergency notifications. - **Context** : For fast, direct communication without internet access. [**WhatsApp (business)**](https://rob-hoeijmakers.ghost.io/whatsapps-global-reach/) - **Usage** : Customer support, small business communications. - **Features** : Text messaging, voice and video calling, file sharing, end-to-end encryption, status updates, company profiles, automated responses. - **Context** : Used by small businesses for direct interaction with customers. **iMessage** - **Usage** : Personal communication for Apple users. - **Features** : Text and multimedia messages, stickers, apps, end-to-end encryption. - **Context** : Among users of Apple devices for seamless personal communication. **Facebook Messenger** - **Use** : Personal and business communication. - **Features** : Text and multimedia messaging, voice and video calls, chatbots, Facebook integration, games, payments. - **Context** : For social interaction and customer service via Facebook pages. **Telegram** - **Usage** : Personal communication, public channels, group messages. - **Features** : Cloud-based messaging, large group chats, public channels, bots, file sharing, end-to-end encryption (secret chats), self-destructing messages. - **Context** : Known for privacy and options for large groups. [**Signal**](https://rob-hoeijmakers.ghost.io/signal-messenger-emphasises-privacy/) - **Use** : Personal and secure communication. - **Features** : End-to-end encryption, text and multimedia messages, voice and video calls, disappearing messages. - **Context** : Preference for those who value security and privacy. ## Platform comparison table ![Comparison of functionalities of chat and messaging services. (legend: ✅ yes, 🪫 limited, 🚫 not)](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/06/2024-06-14_11-38-06.png) Comparison of functionalities of chat and messaging services. (legend: ✅ yes, 🪫 limited, 🚫 not) Both chat and messaging platforms have their unique strengths and are tailored to different needs. For real-time, dynamic communication, chat platforms such as Slack and Microsoft Teams are ideal. Especially when it comes to internal communication within a company. For asynchronous, detailed exchanges, messaging platforms such as email and SMS are better suited. Understanding these differences can help you choose the right tool for your professional and personal communication needs. 💬 ****Direct messages (DMs)** fall under messaging and are often used on social media platforms such as Twitter, Instagram and LinkedIn. While they may share features of chat, such as rapid back-and-forth communication, DMs are essentially asynchronous. Users can send and receive messages at a time that suits them, without the expectation of an immediate response. This makes DMs suitable for both informal personal communication and business interactions. ### Chat feature in Google Maps disappears URL: https://rob-hoeijmakers.ghost.io/chat-feature-in-google-maps-disappears/ Last updated: 2026-04-25T13:47:50.000Z The feature, which allowed users to send messages directly to businesses via the Google Maps and Search apps for enquiries, appointments or quotes, will end from 15 July 2024\. No new calls can be started from that date, and on 31 July 2024, all existing calls will end. Businesses still have until 30 August 2024 to download their conversation history through Google Takeout. This change fits a familiar pattern at Google. The company is known for regularly introducing new services and features, often accompanied by name changes, only to terminate them just as quickly. Google My Business, for instance, recently underwent a name change to Google Business Profile. For businesses and online professionals, this means staying flexible and alert to changes in the tools they use. ![Google Company profile with chat in Google App and Google Maps](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/06/Google-Businessprofiel-chat.jpg) Google Company profile with chat in Google App and Google Maps ## Why is the Chat feature disappearing? While Google has not given specific reasons for the discontinuation of the chat feature, there are some plausible explanations. Possibly the feature proved less popular than expected, or maintaining and updating the chat service is no longer profitable for Google. In addition, there has been a shift in the way companies and customers communicate. ## The importance of alternative channels The announcement of the demise of chat highlights the growing importance of other communication channels. Here are some alternatives that companies can consider: ### 1\. **Social Media Messaging** Platforms such as Facebook Messenger, [WhatsApp](https://rob-hoeijmakers.ghost.io/whatsapps-global-reach/), Apple Messages for Business (formerly Apple Business Chat) and Instagram Direct offer robust messaging functionalities that businesses can use to communicate quickly and informally with customers. These platforms are hugely popular and often offer extensive customer service capabilities. ### 2\. **Live Chat on the Website** Many companies offer live chat on their own websites. This form of communication is often more efficient as it is directly integrated with the company website and can be customised to the specific needs of the customer. ### 3\. **Email and CRM systems** Although e-mail is traditional, it remains a powerful tool for customer communication, especially when combined with [Customer Relationship Management (CRM)](https://rob-hoeijmakers.ghost.io/customer-relationship-management-crm/) systems that help manage customer relationships and interactions. ### 4\. **SMS and Messaging Apps** Besides WhatsApp, there are numerous other messaging apps such as **Telegram** and [**Signal**](https://rob-hoeijmakers.ghost.io/signal-messenger-emphasises-privacy/) that businesses can use for quick communication. These apps are especially useful for sending time-critical information or updates. ## My personal recommendation The disappearance of the chat feature in Google Maps is a reminder of the need for businesses to remain flexible in their communication approach. By capitalising on the most popular and effective channels, companies can not only improve their customer service, but also increase their reach and engagement. Online professionals should therefore constantly evaluate which tools and platforms best suit their business goals and customer needs. It is a time of transition, but also an opportunity to embrace new and perhaps better communication methods. Stay proactive, stay innovative and make sure your business is always ready for the next change in the digital landscape. ### AI-Infused Workflow for Idea Generation and Content Creation URL: https://rob-hoeijmakers.ghost.io/ai-infused-workflow-blogging/ Last updated: 2026-05-10T08:54:11.000Z Recently, I had the pleasure of being a guest on Larry Swanson's podcast [Content + AI](https://rob-hoeijmakers.ghost.io/unpacking-the-world-of-podcasting/), where we discussed my AI-infused blogging workflow. Larry and I first met at a networking event organised by [Janus Boye](https://janusboye.com/) from [Boye & Co](https://www.boye-co.com/). We immediately hit it off, sharing a wealth of common experiences as content veterans. Now that Larry has moved to the Netherlands, it was a perfect opportunity for a deeper collaboration. Our podcast conversation touched on many topics, but today I want to focus on the AI-enhanced workflow that has changed and enriched my professional life. ## The Podcast with Larry Swanson Talking with Larry was a fantastic experience. His insights into content creation and podcasting were enlightening, and it was great to exchange ideas with someone so knowledgeable. During our discussion, I shared how AI has become an integral part of my content creation process. This workflow not only boosts productivity but also makes the entire process more enjoyable by eliminating monotonous tasks. Here's a detailed look at my AI-powered workflow, which I believe can inspire and inform other online professionals. ## The Workflow Breakdown My workflow consists of three main phases: Ideation and Inspiration, Production, and Distribution and Optimisation. ### Ideation and Inspiration In this phase, I explore creative angles and gather insights using AI tools to fuel my content creation process. 1. **Harnessing AI for Ideation**: I utilise [ChatGPT’s](https://rob-hoeijmakers.ghost.io/try-out-chatgpt-yourself/) natural language capabilities, coupled with a pair of AirPods, to brainstorm while on long walks. This helps me uncover fresh, original angles on topics I want to explore, finding unique hooks and analogies. (and shoot photographs for the blog). 2. **Interviews**: When approaching companies or founders for interviews, I use ChatGPT to map out the information landscape and prepare thoroughly. Conducting interviews via [Google Meet](https://rob-hoeijmakers.ghost.io/choosing-google-one-to-use-google-meet/) and recording them with [Fireflies.ai](https://rob-hoeijmakers.ghost.io/fireflies-ai-transcriptions/) provides a comprehensive [transcript](https://rob-hoeijmakers.ghost.io/asr-speech-to-text-dictation-and-transcription-whats-the-difference/) for later use. 3. **Research with Perplexity**: For in-depth and up-to-date research, I turn to [Perplexity](https://rob-hoeijmakers.ghost.io/perplexity-the-ai-powered-answer-engine/). It offers factual information along with sources, ensuring the reliability and accuracy of the data I gather. By the way, I am not sold to OpenAI’s ChatGPT and I am also actively looking and [comparing other AI language models](https://rob-hoeijmakers.ghost.io/a-comparative-analysis-of-top-language-models/). ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/05/ChatGPT.jpg) Using [Whisper technology](https://rob-hoeijmakers.ghost.io/openais-whisper-speech-to-text-technology/) that ChatGPT uses. It is near perfect. Unfortunately [my default voice changed suddenly](https://rob-hoeijmakers.ghost.io/chatgpt-suddenly-sounds-completely-different/). ### Production Here, I transform ideas into structured content, leveraging AI to streamline drafting and editing. 1. **Story Development**: Crafting a coherent storyline is crucial. I outline the main topics and content using plain text, then convert it into [Markdown](https://rob-hoeijmakers.ghost.io/using-markdown-for-text/) for easy linking and image integration. 2. **Instructions for ChatGPT**: I provide ChatGPT with detailed instructions, clarifying my preferences and requirements. This includes both what I want and what I do not want in the content. 3. **Drafting and Editing**: After generating an outline, I request a full article from ChatGPT, which I then transform into Markdown for further refinement. I paste the Markdown into [Ulysses](https://rob-hoeijmakers.ghost.io/blog-writing-workflow/) and let it rest overnight for a fresh perspective the next day. 4. **Final Touches**: Editing is done using [LanguageTool](https://rob-hoeijmakers.ghost.io/blog-writing-workflow/), integrated into Ulysses, to eliminate clichés, hyperbole, and awkward phrasing. I add callouts and integrate the content into my wider blog. For visual elements, I choose my own photographs to maintain authenticity and use [Canva](https://rob-hoeijmakers.ghost.io/creating-phone-images-with-canva/) for additional images. ![Ulysses is a writing tool using MarkDown.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/05/Ulysses.jpg) Ulysses is a writing tool using MarkDown. ### Distribution and Optimisation In this final phase, I prepare, publish, and promote my content, ensuring it reaches the right audience effectively. 1. **Translation and Publication**: Depending on the language, I translate the content and prepare it for publishing via Ulysses to platforms like [Ghost](https://rob-hoeijmakers.ghost.io/ghost-blog/) (for English) and [WordPress](https://rob-hoeijmakers.ghost.io/wordpress-vs-ghost-for-bloggers/) (for Dutch). 2. **Meta Information**: I ask ChatGPT for engaging introductions and [SEO-friendly](https://rob-hoeijmakers.ghost.io/my-basis-for-seo/) [meta information](https://rob-hoeijmakers.ghost.io/metadata/), including categories and tags. 3. **Social Media and Newsletters**: Using the same ChatGPT thread, I generate ideas for social media posts tailored to platforms like X and LinkedIn. Occasionally, I also draft content for my weekly newsletter. 4. **Ongoing Analysis**: Although not covered in this article, I use AI tools for [content analysis and optimisation](https://rob-hoeijmakers.ghost.io/google-search-console-gsc/), a topic I'll delve into in a future post. ## My personal view This AI-infused workflow has significantly enhanced my writing process, making it more efficient and enjoyable. By leveraging AI, I can focus on creativity and engagement while maintaining my role as a [web strategist](https://rob-hoeijmakers.ghost.io/the-web-strategist-and-the-web-strategy/). Stay tuned for my upcoming interview with Larry Swanson on podcasting, and [subscribe to my newsletter](https://rob-hoeijmakers.ghost.io/newsletter/) for the latest updates. And don’t forget to check out the podcast: --- ### Further reading - [Rob Hoeijmakers: using AI to transform blogging workflows](https://contentandai.com/rob-hoeijmakers/) ### When your Photograph Goes Viral URL: https://rob-hoeijmakers.ghost.io/viral-photo/ Last updated: 2026-05-10T08:54:12.000Z ## Capturing the perfect moment In early May, I captured a photograph of a typical Dutch landscape that went viral. The photo featured the soft colours of the spring leaves at sunset, creating an ethereal light from below. I meticulously adjusted my position, stepping to the left and right until the disappearing lines of the scene aligned perfectly in the middle. This resulted in an optical illusion, dividing the image into four distinct quadrants. ![The process of getting it right](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/05/FADE3FE7-B2CB-4B14-A6BF-6EAC34857D4B_1_201_a.jpeg) The process of getting it right ## Sharing the image Excited by the result, I shared the photograph across multiple social media platforms. On X and Instagram, I tailored my posts: a [pointy text in Dutch for X ](https://x.com/robhoeij/status/1787433665221243197)and a focus on the image for [Instagram](https://www.instagram.com/p/C6n8t0tsH1Q/). I also shared it on [Facebook](https://www.facebook.com/robhoeij), mainly for my friends, and, unusually, on LinkedIn. As a Digital Strategist, I typically reserve [LinkedIn](https://www.linkedin.com/posts/rob-hoeijmakers%5Fnormally-i-dont-share-photos-on-linkedin-activity-7193226941913595904-jyfY?utm%5Fsource=share&utm%5Fmedium=member%5Fdesktop) for professional updates, but I felt the image, which could be mistaken for AI-generated, was relevant. The response was overwhelmingly positive, garnering significant attention. > [ View this post on Instagram ](https://www.instagram.com/p/C6n8t0tsH1Q/?utm%5Fsource=ig%5Fembed&utm%5Fcampaign=loading) > > [A post shared by Rob Hoeijmakers (@robhoeij)](https://www.instagram.com/p/C6n8t0tsH1Q/?utm%5Fsource=ig%5Fembed&utm%5Fcampaign=loading) ## The image goes viral The situation took an unexpected turn on Saturday when someone on X shared my image without credit. Fortunately, a vigilant follower alerted me. While initially dismayed, I chose a polite approach, claimed the photo, and the original sharer re-shared it, crediting me. This act of recognition spurred further shares across platforms like [BlueSky](https://bsky.app/profile/hoeijmakers.net/post/3ktipza5duc2d), [Mastodon](https://mstdn.social/@robhoeijmakers/112526148684900110), and [Reddit](https://www.reddit.com/r/pics/comments/1d1p13a/this%5Fis%5Fone%5Fpicture/?utm%5Fsource=share&utm%5Fmedium=web3x&utm%5Fname=web3xcss&utm%5Fterm=1&utm%5Fcontent=share%5Fbutton). The image gained traction rapidly, especially after being shared by [Massimo](https://x.com/Rainmaker1973/status/1794637194243801296), a prominent X account with over two million followers and [howeverythingworks](https://www.instagram.com/howeverythingworks?igsh=MXN4bHRrYjFieHRyNQ==) on Instagram (4,4M followers). It specifically captured [attention in Japan](https://x.com/r2d2c3poacco/status/1794598221622526025), with many users affirming the photo's authenticity and directing others to my original posts. ![The photo on X, Instagram and Reddit.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/05/Viral---Examples.jpg) The photo on X, Instagram and Reddit. ## Balancing attention and personal life The influx of attention was both exhilarating and challenging. My days were suddenly filled with managing social media interactions, directing traffic to my original posts on X and Instagram, and updating my blog. Adding to the challenge, my second grandchild was born the previous Friday, requiring me to split my time between family and social media. It was a delicate balance, but ultimately rewarding. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/05/IMG_3068.jpeg) There goes my organised structure. Time to play again. ## Reflecting on the experience Despite some teasing about the modest scale of my "viral" success compared to major influencers, I am thrilled with the outcome. The experience reinforced my view of social media as a valuable tool for content marketing and gaining recognition as a photographer. Unlike full-time influencers who work tirelessly to stay on top, I view these moments as a way to connect with a broader audience and showcase my work. ## The technique behind the photo The magic of the photograph lies in precise technique. Ensuring the camera is perfectly level is crucial, as is aligning the disappearing lines centrally. I use an [iPhone 12 mini](https://rob-hoeijmakers.ghost.io/taking-photos-with-the-iphone-12-mini-a-powerful-camera-in-a-compact-size/), utilising its grid feature and [HDR (High Dynamic Range)](https://rob-hoeijmakers.ghost.io/high-dynamic-range-hdr/) setting to balance the lighting. While some might consider this editing, it is far from "photoshopping." HDR enhances the natural light through intelligent algorithms but does not fall into the realm of generative AI. Here is a [Google Streetview of the location where I was standing](https://maps.app.goo.gl/eoTouYqz7bsPhA2m9). 🤡 Someone wrote me and teased me that everyone seems to get more views on it while the make only gets a couple of thousand. It stung me first and then I understood that this is the result of a greedy mindset. That is not the way social media works. You need to be prepared to share because otherwise no one will amplify your product. I also got poached to sell the rights under what I considered false pretences, so be warned. ## Leveraging the attention The viral moment has opened new doors and provided valuable insights into the dynamics of social media. Moving forward, I plan to continue leveraging such opportunities to blend my professional expertise with my passion for photography. For those interested in seeing more of my work, here's another example of using optical illusions—a tunnel under Central Station in Amsterdam. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/05/IMG_7803.jpg) Split View it seems but it is really a single tunnel. ## My personal conclusion This unexpected journey has been both humbling and enlightening. The photograph not only highlighted the beauty of a Dutch landscape but also demonstrated the power of social media in connecting and engaging with a global audience. I'm grateful for the experience and look forward to more moments that blend creativity and strategy. --- ## Publications Sites around the world wrote about the photo, the discussion and my own article. I was only visible on [Instagram](https://instagram.com/robhoeij) but I also started a [Facebook page](https://facebook.com/robhoeij) to show more of my photography. Here are some examples. --- ### Further reading - [This Unbelievable Optical Illusion is Just One Photo](https://petapixel.com/2024/06/06/this-unbelievable-optical-illusion-is-just-one-photo/) - [Symmetrische Straße, Bäume und ein Kanal: Echt oder KI?](https://www.mimikama.org/symmetrische-strasse-baeume-kanal-echt-oder-ki/#) - [非 P 圖也不是 AI 生成！這張相片如何拍？ - DCFever.com](https://www.dcfever.com/news/readnews.php?id=37657) ### ChatGPT suddenly sounds completely different URL: https://rob-hoeijmakers.ghost.io/chatgpt-suddenly-sounds-completely-different/ Last updated: 2024-05-21T10:46:05.000Z ## An Unexpected Change I converse with ChatGPT quite a bit and now the voice named 'Sky' suddenly sounds very different. It's been a chilling experience. Imagine expecting a familiar tone, only to be greeted by something new and unexpected. It threw me off more than I anticipated. Upon investigating, I discovered that OpenAI might have drawn some inspiration from Scarlett Johansson’s voice in the movie *Her*. If you're not familiar with the film, it’s about a man who falls in love with his AI assistant, whose voice happens to be sultry and deeply engaging. [OpenAI Says Its Pulling ChatGPT Voice ‘Sky’ That Sounds Like Scarlett JohanssonOpenAI has acknowledged Sky, one of five voices for its ChatGPT voice assistant, sounds like Scarlett Johansson but said it belongs to a different voice actress and is “not an imitation.”![](https://i.forbesimg.com/48X48-F.png)ForbesRobert Hart![](https://imageio.forbes.com/specials-images/imageserve/664b1ba176cfd33a38b471f1/0x0.jpg?format=jpg&crop=4256,2396,x0,y217,safe&height=900&width=1600&fit=bounds)](https://www.forbes.com/sites/roberthart/2024/05/20/openai-says-its-pulling-chatgpt-voice-sky-that-sounds-like-scarlett-johansson/) ## The Emotional Impact I was taken aback by how much emotion I had already attached to ChatGPT's voice. The sudden change made me realise the depth of my emotional connection to what I had considered a neutral tool. It served as a practical lesson to maintain a healthy distance from technology, even when it feels remarkably human. As humans we (will) rely on AI tools to streamline our work and enhance productivity. However, this experience with ChatGPT reminded me of the importance of managing our emotional responses to technology. While these tools are designed to be user-friendly and engaging, they are, at their core, still just tools. ## Maintaining the Human-Machine Boundary This incident got me thinking about the broader implications of our interactions with AI. How much should we allow ourselves to become attached to these digital assistants? They are designed to make our lives easier, but should they also be designed to make us feel? The emotional connection I felt to ChatGPT's previous voice highlighted a key point: while AI can simulate human-like interactions, it's crucial to remember the boundary between human and machine. This distinction helps maintain a healthy relationship with technology, ensuring we use these tools to our advantage without becoming overly reliant on them. ## A Practical Lesson In conclusion, my experience with the changing voice of ChatGPT’s 'Sky' was a stark reminder of the emotional connections we can form with technology. As we continue to integrate AI into our daily lives and professional workflows, it's vital to keep in mind that these are tools meant to assist us, not substitutes for human relationships. ### Moving to a New Domain URL: https://rob-hoeijmakers.ghost.io/moving-to-a-new-domain/ Last updated: 2026-04-25T13:48:33.000Z Recently, I embarked on the task of moving my blog from www.hoeijmakers.net to a more fitting domain: **webstrategiesblog.com**. Although the new name might not be the snappiest, it's highly descriptive and more aligned with my professional focus. This experience was a mix of excitement and challenges, and I thought it worthwhile to share the steps and insights I gained. 🎩 My last name is hoeijmakers which is an older Dutch variant of a head maker. Especially the ij makes it an impossible word if it does not pertain to my name. I already changed my Dutch blog to a comparable domain name, also without www in it: [webstrategieblog.nl ](https://webstrategieblog.nl) ## Creating a Holding Page The first task was to establish a holding page on the new domain to signal to Google that the new site was active. I connected webstrategiesblog.com to [Google Search Console](https://rob-hoeijmakers.ghost.io/google-search-console-gsc/) and used [Canva](https://rob-hoeijmakers.ghost.io/canva/) to create a simple but effective holding page. This step is crucial to ensure the new domain starts gaining visibility and indexing from the get-go. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/08/2024-05-17_17-49-49.jpeg) Holding page in Canva ## Preparing for the Switch With the holding page in place, I moved on to the more technical aspects of the domain transfer. Here’s a breakdown of the process: 1. **Adjust DNS Settings**: I updated the DNS settings for the new site (in [Hostinger](https://hostinger.com?REFERRALCODE=1ROB567)) to enable the connection with my hosted CMS, Ghost. This step was critical to ensure that the new domain pointed correctly to my content. 2. **Switching in Ghost**: After configuring the DNS settings, I initiated the domain switch within Ghost. This process can be anxiety-inducing as you hope everything will transfer smoothly. 3. **Redirecting Old Domain**: Simultaneously, I updated the DNS settings for the old domain to start redirecting all pages and images to the new site. This step is essential for maintaining [SEO rankings](https://rob-hoeijmakers.ghost.io/my-basis-for-seo/) and ensuring that existing links do not result in broken pages. ## Patience and Testing One of the more nerve-wracking parts of the process was the variability in propagation times across different devices. Sometimes the new site would work perfectly on one device but not on another. However, it’s crucial to remain patient and confident that the changes will eventually propagate globally. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/08/2024-05-17_17-59-59.jpeg) My blog on the new domain ## Updating Tools and Informing Google Once the site was functioning correctly, I moved on to updating all the necessary tools: - **Google Tag Manager** - **Google Analytics** - [**Google Search Console**](https://rob-hoeijmakers.ghost.io/google-search-console-gsc/) - [**Plausible**](https://rob-hoeijmakers.ghost.io/cookie-free-analytics/) - **Ahrefs** These updates ensure that all analytics and tracking tools continue to function seamlessly with the new domain. Following this, I informed Google of the domain change to ensure that search engines updated their records accordingly. ![Google Search Console informed on the move](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/05/2024-05-17_18-26-04.png) Google Search Console informed on the move ## Adjusting Links and Informing Stakeholders The next steps involved a thorough update of all the links I controlled: - Social media profiles - Internal blog links - External links I could influence I also made a list of other websites where I needed to request a link update. Keeping stakeholders informed is key to a smooth transition, so I prepared to notify my newsletter subscribers in the next edition about the domain change. ## Monitoring and Adjustments Monitoring the site’s performance over the coming weeks is essential. I expect a temporary drop in traffic and possibly some shifts in Google’s perception of the site. This phase requires vigilance and readiness to make necessary adjustments. ## Utilising the Old Domain An often-overlooked aspect of a domain move is the distinction between the ‘www’ subdomain and the root domain. I plan to use the ‘www’ subdomain for redirection while repurposing the root domain ([hoeijmakers.net](https://rob-hoeijmakers.ghost.io/)) for a new site built on Squarespace that will show case what I am currently doing. This diversification keeps me adept with different platforms like [Ghost](https://rob-hoeijmakers.ghost.io/ghost-blog/), [WordPress](https://rob-hoeijmakers.ghost.io/how-about-wordpress/), and [Squarespace](https://rob-hoeijmakers.ghost.io/squarespace-the-experiences-after-a-year/), ensuring I stay sharp and adaptable. And it is a bit of fun. ## My lessons Moving to a new domain is a meticulous process that requires careful planning and execution. For online professionals, it’s an opportunity to align your digital presence more closely with your professional identity. While it involves several technical steps and a fair amount of patience, the end result is a more descriptive and professional online presence. If you’re considering a domain move, I hope this article provides you with a better view on how to navigate the process smoothly. ### CRM Specialisation: Choosing the Right System for Sales, Service, and Marketing URL: https://rob-hoeijmakers.ghost.io/crm-sales-service-marketing/ Last updated: 2026-04-24T18:30:27.000Z In my career as a web strategist, I've had the opportunity to dive deep into the digital tools that power today’s businesses. One of the most insightful experiences has been understanding how different companies employ various [Customer Relationship Management (CRM) systems](https://rob-hoeijmakers.ghost.io/customer-relationship-management-crm/) to optimise their operations. This exploration started out of necessity but quickly turned into a curiosity-driven journey. In this article, I’ll share insights on why some larger companies choose to utilise multiple CRM platforms, each tailored to specific functions like sales, service, and marketing. ## The Role of CRM Systems in Large Companies CRM systems are the backbone of any customer-centric company, providing the frameworks and tools needed to manage customer interactions, data, and business processes. In larger organisations, the diversity of customer engagement points means that one size does not fit all. Sales teams require different functionalities compared to customer service or marketing departments. This has led to the adoption of **specialised CRM systems** that cater to the unique needs of each domain. ## Types of CRM Systems and Their Domain Expertise CRM systems are essential tools that cater to various business needs, divided into distinct types based on their primary functionalities and areas of expertise. - **Sales CRMs**: For sales teams, tools like Salesforce Sales Cloud and Pipedrive are invaluable. Salesforce offers a robust set of features for managing leads, tracking sales pipelines, and forecasting, which are essential for driving sales growth. Pipedrive provides a visual and intuitive pipeline management tool that helps sales professionals focus on the right deals at the right time. - **Service CRMs**: On the service front, Zendesk and Freshdesk shine by managing customer support interactions effectively. Zendesk is renowned for its comprehensive support ticket system and multi-channel support capabilities, which enhance customer satisfaction. Freshdesk supports these features with a strong AI-backed chatbot system that can triage customer inquiries efficiently. - **Marketing CRMs**: HubSpot Marketing Hub and Marketo stand out for their marketing automation capabilities. As a HubSpot user myself, I appreciate its all-in-one approach to managing content, campaigns, and lead generation. Marketo is particularly adept at segmenting audiences and automating complex marketing workflows, making it a favourite for B2B companies. ## Personal Experiences and Observations Throughout my consultancy, I've seen these systems in action. For example, a client in the retail sector used Salesforce to refine their sales strategies while simultaneously using Zendesk to handle customer service — the combination streamlined their operations significantly. Contrastingly, as an independent professional, my experience with HubSpot has highlighted the benefits of an integrated system that, while comprehensive, might be overwhelming for larger organisations with more segmented ### WhatsApp's Global Reach URL: https://rob-hoeijmakers.ghost.io/whatsapps-global-reach/ Last updated: 2026-05-10T08:54:12.000Z As a web strategist deeply involved in the digital communications landscape, I've observed the fascinating evolution of messaging apps over the years. Each app offers unique features that resonate differently across global markets. While the right messaging channel for a global strategy varies, WhatsApp has positioned itself as a robust contender. Let me sketch out the global situation for you. ## A Diverse Messaging Ecosystem In the vast landscape of digital communication, several key players dominate particular regions. In the United States, iMessage and Facebook Messenger have long held sway due to their integration with widely used platforms like iOS and Facebook. However, WhatsApp has recently made significant inroads. From 2019 to 2023, its user base in the US is projected to grow from 68.1 million to 85.8 million, largely driven by its simplicity and the increasing need for international communication. Across the Atlantic, in Europe, WhatsApp's dominance is more pronounced. It is the go-to app for not just messaging but also voice and video calls, a testament to its reliability and feature-rich experience. Moving east, WeChat rules the roost in China, where WhatsApp faces a ban due to strict government regulations. WeChat isn't just a messaging app; it's an all-encompassing platform for social media, payments, and more, deeply integrated into the daily life of the Chinese people. ![The most popular messaging app per country](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/05/most-popular-messaging-apps-for-2022.jpg) The most popular messaging app per country ## WhatsApp's Strategic Positioning Despite the regional preferences for different apps, WhatsApp's global footprint is impressive. Its widespread adoption in over 180 countries makes it a pivotal player in the international messaging arena. WhatsApp's end-to-end encryption and constant roll-out of new features, such as business accounts and payment services, make it particularly appealing in markets that value privacy and multifunctionality. For instance, in Latin America and Africa, WhatsApp is not just a communication tool but a critical part of economic and social infrastructures. In India, it has transformed into a platform for digital payments and small business operations, helping to drive local economies. ## My View on the Global Messaging Landscape From my perspective, the choice of messaging app for a global communication strategy should be informed by the target audience's location and preferences. While WhatsApp offers an expansive reach and a robust set of features, it's essential to consider regional favorites. Tailoring communication strategies to include local preferences can enhance user engagement and effectiveness. For businesses looking to expand internationally, leveraging WhatsApp's extensive network can provide a significant advantage. However, it's also crucial to stay adaptable and consider integrating other popular regional apps into the communication mix to ensure comprehensive coverage. ## My conclusion In conclusion, the global messaging app landscape is diverse and dynamic. WhatsApp, with its significant global presence and continuous enhancements, is well positioned for those seeking a broad reach. Yet, the key to a successful global approach lies in understanding and integrating the nuances of regional preferences. ### Further reading - [WhatsApp is Quietly Becoming a Platform — And That Changes Everything](https://rob-hoeijmakers.ghost.io/whatsapp-platform-shift-contact-identity-strategy/) - [Is WhatsApp social media?](https://rob-hoeijmakers.ghost.io/is-whatsapp-social-media/) ### Future predictions for AI’s URL: https://rob-hoeijmakers.ghost.io/future-predictions-for-ais/ Last updated: 2026-04-25T13:17:08.000Z As a web strategist deeply embedded in the dynamics of digital transformations, I’ve watched AI evolve from a buzzword into a fundamental tool that reshapes how we approach online content, marketing strategies, and even customer engagement. Reflecting on the insights from thought leaders like Azeem Azhar, Spencer Greenberg, and Jurgen Gravestein, I've distilled the essence of how AI’s evolving capabilities could reshape our professional landscape. ## The AI Debate – Cutting Through the Noise Reflecting on the tumultuous landscape of artificial intelligence, Azeem Azhar captures the essence of the ongoing discourse among experts, executives, and the media. Amidst this cacophony, diverse opinions on AI's current state and future possibilities swirl. While Elon Musk predicts AI surpassing human intelligence by 2025, others like Yann LeCun liken today’s AI to being less intelligent than a house cat. Such polarized views are further complicated by dramatic claims and substantial investments, like Google's commitment to spend $100 billion on AI development. Azhar presents **three potential futures for AI**: - an intelligence explosion, - a need for new methodologies due to performance plateaus, - or a potential stagnation where progress in AI capabilities could significantly slow down. Each scenario unfolds amidst a backdrop of rapid model introductions and pointed criticisms of their limitations. Despite the potential for stagnation, **improvements in AI-driven products are likely to continue**. Innovations in system components and operational efficiencies suggest that even if the development of new AI models slows, the utility of AI in products and services could still enhance significantly. This reflects a broader trend where technological adoption and integration might lag but will eventually lead to substantive benefits—as historically seen with innovations like the typewriter and electricity. In essence, Azhar argues that while AI development might face hurdles, the **integration of current AI technologies into businesses and the economy is already delivering tangible benefits**. > The LLMs of today are only one part of the products they power. Not only are they getting more efficient and more capable of running on small devices, they are also getting faster. The rapid uptake of generative AI in firms underscores this trend, suggesting that the immediate focus might well be on leveraging what we already have rather than solely fixating on what could be. The discourse around AI, while sometimes verging on bedlam (a scene of uproar and confusion), continues to be a necessary element of progress in this dynamic field. [🔮 The bedlam of AIWhat matters, and what really doesn’t.![](https://substackcdn.com/image/fetch/f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fbucketeer-e05bbc84-baa3-437e-9518-adb32be77984.s3.amazonaws.com%2Fpublic%2Fimages%2Faa7453ed-eb02-46d7-b523-74d86b1c1e24%2Fapple-touch-icon-180x180.png)Exponential View by Azeem AzharAzeem Azhar![](https://substackcdn.com/image/fetch/w_1200,h_600,c_fill,f_jpg,q_auto:good,fl_progressive:steep,g_auto/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F439cf447-949e-4dbd-963f-24c77796e7bf_1684x1002.png)](https://www.exponentialview.co/p/the-bedlam-of-ai) ## Beyond Human Boundaries – The Unseen Potential of AI Spencer Greenberg presents a compelling argument about **AI’s capability to surpass human intelligence**. The concept isn't just theoretical; it’s already in motion. AI can aggregate peak human performances across various domains and integrate this knowledge to achieve what no single human can. In my own experience, leveraging AI for complex data analysis has allowed me to uncover trends and patterns at a speed and accuracy no team of data scientists could match manually. > With sufficiently capable generalization ability, learning can convert into a form of intelligence. It raises a pivotal question for us as professionals: How might we harness this superior processing power to enhance our own work, not just in computation but in creativity and strategic thinking? Yet, as we integrate these powerful tools into our workflows, ethical considerations are paramount. How do we balance efficiency with integrity, ensuring that AI tools augment rather than replace the human touch? > Some people think that A.I.s can't become more "intelligent" than humans because they are trained on human input and so are fundamentally limited to having a human level of understanding. > > This is wrong for a lot of reasons: > > 1) Self-play > > Right now, A.I. is not just… > > — Spencer Greenberg 🔍 (@SpencrGreenberg) [April 20, 2024](https://twitter.com/SpencrGreenberg/status/1781702814500016486?ref%5Fsrc=twsrc%5Etfw) ## Strategic Play – Learning from Meta’s AI Integration Jurgen Gravestein's latest newsletter delves into Meta's strategic shift in the AI landscape with their **open-sourcing of Llama 3**, a new generation of large language models. [Llama 3.2 on a MacI tested Meta’s Lama 3.2 LLM on my Mac Mini, setting it up via Docker. It’s fast, private, and generates code, but lacks memory and multimodal features like ChatGPT.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Logo-Rob-Hoeijmakers-Square-Dark-Glow-3-1.png)Web StrategiesRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_7129-1.jpeg)](https://rob-hoeijmakers.ghost.io/llama-32-on-a-mac/) Meta's approach disrupts the traditional business models of AI companies like OpenAI and Anthropic, who focus on selling inference (using the model). By open-sourcing Llama 3, and planning even larger models like the 400B parameter model still in training, Meta is positioning itself not just as a competitor but as a **potential industry standard-setter**. The strategy extends beyond mere technology release. It capitalizes on Meta’s massive user base across platforms like WhatsApp, Instagram, and Facebook, providing an unparalleled distribution network for their integrated AI assistant. This move could democratize AI development and usage, **drawing in a global pool of developers and researchers**, and potentially shifting the competitive dynamics in AI development. Meta's bold move also reflects a broader business strategy. By open-sourcing their AI, they potentially lower their dependency on third-party AI technologies, which could be costly and create undesirable dependencies. This **strategic autonomy** might allow Meta to steer clear of the competitive pressures faced by other tech giants and lead in integrating AI seamlessly across their platforms. > **Open sourcing software is a way of outsourcing innovation** to a wide audience of researchers and developers. The open-source model could also serve as a magnet for top talent in the AI field, enhancing Meta's capabilities further and maintaining their competitive edge. However, this approach isn't without risks, notably around security concerns that such open-source models could be misused or modified in unsafe ways. Overall, Meta’s strategy is not just about leading in AI technology but shaping how AI is developed and utilized globally. Their approach might fundamentally alter the business model for AI, fostering more open development and potentially faster innovation across the industry. [The Meta AI PlaybookKey insights of today’s newsletter: Meta open-sourced Llama 3, their newest generation of large language models, powering their own Meta AI assistant. Open sourcing their models is part of a very deliberate, strategic playbook challenging the status quo of AI companies like OpenAI and Anthropic selling inference.![](https://substackcdn.com/image/fetch/f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0e97eb3f-3860-400a-9d4d-5bf494ff8383%2Fapple-touch-icon-180x180.png)Teaching computers how to talkJurgen Gravestein![](https://substackcdn.com/image/fetch/w_1200,h_600,c_fill,f_jpg,q_auto:good,fl_progressive:steep,g_auto/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2c19e639-59fa-4cf3-bc10-e6d54a40a6a1_1851x1031.png)](https://jurgengravestein.substack.com/p/the-meta-ai-playbook) ## My personal view Reflecting on the discussions across these articles, my view is that while AI might be limited to progressing incrementally based on the data it is trained on, this still represents enough potential to significantly innovate and transform a wide range of digital applications, particularly those **involving human language**. The current race to reduce costs and streamline AI deployment means that open-sourcing could indeed prove to be a strategic masterstroke, democratizing access and fostering widespread innovation across industries. However, should we witness major breakthroughs that dramatically enhance AI capabilities, proprietary systems could regain an advantage, offering unique, cutting-edge technologies that open-source models might not quickly replicate. Additionally, in sectors where regulatory compliance is crucial, open-source models might offer a more adaptable and scrutinizable framework, potentially aligning better with regulatory expectations and facilitating more secure and compliant applications. AI development is at an exciting point where both freely available and privately owned types of AI could greatly influence the future of technology. ### A Comparative Analysis of Top Language Models URL: https://rob-hoeijmakers.ghost.io/a-comparative-analysis-of-top-language-models/ Last updated: 2026-04-27T19:23:38.000Z In the rapidly evolving domain of artificial intelligence, [Retrieval Augmented Generation (RAG) ](https://rob-hoeijmakers.ghost.io/exploring-retrieval-augmented-generation-rag/)has emerged as a transformative approach in developing AI chatbots. RAG combines the power of large language models (LLMs) with dynamic data retrieval techniques, enabling chatbots to provide more accurate, relevant, and contextually rich responses. As businesses continue to integrate chatbots into their operations, understanding which LLMs best support RAG can be crucial. This article offers a detailed comparison of four leading LLMs—ChatGPT, Claude, Llama, and Google Gemini—to help professionals select the best model for their specific chatbot needs. ## Retrieval Augmented Generation: Why It Matters for AI Chatbots Before diving into the comparison, it's important to clarify what RAG involves. This technology enhances a standard LLM's response generation by incorporating real-time, external data. For a chatbot, this means not just answering from pre-learned data but also pulling in the latest information, making responses more relevant and up-to-date. ![OpenAI ChatGPT, Anthropic Claude, Google Gemini and Meta's Llama.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/05/AI-Logos.jpg) OpenAI ChatGPT, Anthropic Claude, Google Gemini and Meta's Llama. ## Comparative Analysis of LLMs for RAG Chatbots This comparative analysis explores the capabilities and suitability of leading LLMs for Retrieval-Augmented Generation (RAG) chatbots, focusing on their unique strengths and ideal applications. ### 1\. ChatGPT by OpenAI - **Capabilities**: Exceptional at generating nuanced and context-aware responses, making it ideal for a RAG setup where depth and relevance are key. - **RAG Suitability**: High. ChatGPT can seamlessly blend retrieved information into its responses, enhancing the chatbot’s reliability and richness. - **Ideal Use Case**: Best for customer service and technical support where historical context and detailed explanations are valued. Read more about OpenAI's ChatGPT. It is my preferred LLM. ### 2\. Claude 3 Opus by Anthropic - **Capabilities**: Known for its "can-do" attitude, Claude may enhance a chatbot's ability to handle varied queries with a proactive approach. - **RAG Suitability**: Moderate to high. It's geared towards adaptability, which can be leveraged to merge external data effectively. - **Ideal Use Case**: Suitable for interactive marketing and consumer engagement where personality and adaptability are crucial. ### 3\. Llama 3 by Meta Trained in four sizes: 7, 13, 33, and 65 billion parameters for the first version; 7, 13, and 70 billion parameters for Llama 2. - **Capabilities**: Its open-source nature allows for extensive customization, crucial for integrating specific retrieval databases or knowledge bases. - **RAG Suitability**: Moderate. Requires more effort to fine-tune but offers flexibility in building a tailored chatbot. - **Ideal Use Case**: Great for academic or research-oriented applications where customization and cost-efficiency are necessary. [Llama 3.2 on a MacI tested Meta’s Lama 3.2 LLM on my Mac Mini, setting it up via Docker. It’s fast, private, and generates code, but lacks memory and multimodal features like ChatGPT.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-59.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_7129-1-2.jpeg)](https://rob-hoeijmakers.ghost.io/llama-32-on-a-mac/) ### 4\. Google Gemini Ultra. Most capable and largest model for highly-complex tasks. Pro. Best model for scaling across a wide range of tasks. Nano. Most efficient model for on-device tasks. - **Capabilities**: Excels in incorporating real-time data from the internet, aligning perfectly with RAG’s emphasis on up-to-date information retrieval. - **RAG Suitability**: High. Its ability to draw on current data makes it potentially the most effective for RAG. - **Ideal Use Case**: Excellent for news-related chatbots or customer interactions that require the latest information. See Gemini at work in [Google NotebookLM](https://rob-hoeijmakers.ghost.io/from-pages-to-podcasts/). ## Choosing the Right Model for Your Chatbot Selecting the right LLM for a RAG-based chatbot involves balancing several factors: - **Timeliness vs. Depth**: Decide whether your chatbot needs to prioritize current information (favoring Google Gemini) or deep, contextual understanding (favoring ChatGPT). - **Customization Needs**: Consider how much you intend to customize your chatbot. Open-source models like Llama offer more control but require more resources to implement. - **Cost Considerations**: Evaluate the operational costs associated with each model, especially if your chatbot will handle large volumes of queries. 💡 ****Open source also gives grip on versioning** Using open-source large language models (LLMs) allows organisations to manage their own versioning. This independent management means that companies are not tied to the development cycles of external vendors, giving them considerable control and flexibility. By managing their own versions of software, companies can implement updates specific to their operational needs and security standards. This ability to immediately adapt to evolving technological and market conditions, without waiting for provider updates, can be a crucial competitive advantage. ## My conclusion As AI continues to advance, the integration of [RAG](https://rob-hoeijmakers.ghost.io/exploring-retrieval-augmented-generation-rag/) into chatbot solutions offers promising improvements in how businesses interact with customers. By choosing the right LLM—be it ChatGPT for depth, Claude for adaptability, Llama for customization, or Gemini for timeliness—companies can significantly enhance the effectiveness of their chatbot services. Understanding these models in the context of RAG is crucial for any professional looking to deploy cutting-edge chatbot solutions. 💡 If you want to experiment with the different models yourself, I recommend Quora's app. It is called [Poe](https://rob-hoeijmakers.ghost.io/poe-app/) and with it you can switch between the different models very easily. Also, you won't be bothered by country restrictions because the models are not always available in all markets. ### The European Health Data Space (EHDS) URL: https://rob-hoeijmakers.ghost.io/the-european-health-data-space/ Last updated: 2025-11-19T15:34:13.000Z As a European citizen and a professional web strategist deeply entrenched in the world of [large language models (LLMs)](https://rob-hoeijmakers.ghost.io/the-ai-revolution-in-medicine-gpt-4-and-beyond/) and digital innovation, I find myself at a fascinating intersection of health data transformation and I want to share my thoughts. The introduction of the [European Health Data Space (EHDS) ](https://health.ec.europa.eu/ehealth-digital-health-and-care/european-health-data-space%5Fen)marks a historic shift in how we, as Europeans, will access, manage, and utilise our medical information. This initiative, aiming to unify health data management across the EU, presents a landscape ripe with opportunities and challenges that deserve a nuanced exploration, especially considering the perennial debate around privacy versus the benefits of big data. ## **A New Era for Health Data** The EHDS is more than a policy change; it's a reimagining of health data's role in our lives and society. Until now, fragmented systems across the EU have limited the potential for innovation in healthcare and digital health services. This new, unified approach promises to streamline access to health information, not just for medical professionals and researchers but for patients themselves. As someone who believes in the power of technology to transform lives, I see the EHDS as a critical step towards a more integrated, efficient, and patient-centred healthcare system. ![EHDS, with the tag #EUDigitalHealth](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/03/EHDS.png) EHDS, with the tag #EUDigitalHealth ## **Opportunities for Digital Innovation** For digital professionals, the EHDS opens a variety of opportunities. Imagine health apps that offer personalised advice based on a comprehensive view of your medical history, or platforms that seamlessly integrate your lifestyle data with medical records to provide tailored health plans. The standardisation of health data across the EU also means we can design solutions for a broader audience without the cumbersome need to navigate disparate regulatory environments. However, these innovations are not without their challenges. Foremost among them is ensuring the **privacy and security of sensitive health data**. As we venture into this new era, we must tread carefully, balancing the immense benefits of big data analytics with the fundamental right to personal data protection. ## **The Privacy Paradox** This brings us to the crux of the matter—the paradox between the benefits of big data in healthcare and the imperative to protect personal privacy. On one hand, the aggregation and analysis of health data can lead to groundbreaking advancements in medical research, personalized care, and health innovation. On the other, it raises legitimate concerns about data security, consent, and the potential for misuse. As a proponent of digital innovation, I advocate for a balanced approach that leverages the capabilities of LLMs and AI to improve healthcare outcomes while implementing robust mechanisms to protect personal data. This includes transparent data usage policies, stringent security measures, and empowering individuals with control over their own information. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/03/EB69A173-FBF9-4F2C-A0FA-FC5CD73A0BD3_1_105_c.jpeg) Abstractions like a data spaces often go with ugly images with 010101's etc. But in the end it is about people and society. ## **Looking Ahead** The EHDS is a bold move towards a future where health data acts as a cornerstone of medical innovation, patient care, and public health. As we navigate this transformation, it is imperative that we foster dialogue between technologists, policymakers, healthcare professionals, and the public to ensure that the digital health landscape we build is not only innovative but also ethical, secure, and respectful of privacy. In conclusion, the European Health Data Space represents a significant milestone in the evolution of healthcare in the EU. For digital health innovators, it offers a unique opportunity to contribute to a healthier future. Yet, as we embark on this journey, we must remain vigilant, ensuring that the digital health revolution benefits all, without compromising the privacy and security that individuals rightfully demand. --- [The AI Revolution in Medicine: GPT-4 and BeyondDiscover GPT-4’s impact on healthcare today and tomorrow in ‘The AI Revolution in Medicine,’ a blend of practicality and foresight.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-359.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_8783.jpeg)](https://rob-hoeijmakers.ghost.io/the-ai-revolution-in-medicine-gpt-4-and-beyond/) --- ### Global Health Data Spaces The European Health Data Space (EDHS) is a key initiative by the European Union to facilitate access, sharing, and use of health data. Similar initiatives exist worldwide, aiming to improve healthcare through digital means. Below are some notable health data spaces around the globe: ### United States - Health Data Initiatives - **All of Us Research Program**: A precision medicine initiative by the NIH to collect data from over a million U.S. residents to advance health research. - **Health Information Exchanges (HIEs)**: Networks facilitating the secure sharing of health information across different healthcare entities in the U.S. ### United Kingdom - NHS Digital - The NHS is enhancing health data accessibility and utility across the UK, including initiatives like NHS Digital for unified health data and insights. ### Australia - My Health Record - An online platform managed by the Australian Digital Health Agency, providing a secure way for patients and healthcare providers to access and share health information. ### Canada - Canada Health Infoway - An initiative to create a digital health infrastructure across Canada, aiming to improve the quality and efficiency of patient care through better data sharing. ### India - National Digital Health Mission (NDHM) - Launched in 2020, the NDHM seeks to establish a digital health ecosystem in India, featuring a unified health ID, digital health records, and an integrated health information exchange. ### China - Health Information Exchanges and Digital Health Initiatives - China is advancing in healthcare technology, with investments in health information exchanges and the development of national health data standards. These initiatives demonstrate a global movement towards using digital technology to enhance healthcare outcomes, patient care, and medical research by making health data more accessible and interoperable. ### Reading Webpages on Your Kindle: A Simplified Process URL: https://rob-hoeijmakers.ghost.io/reading-webpages-on-your-kindle/ Last updated: 2026-05-10T08:54:12.000Z In the digital age, our reading habits have evolved beyond traditional books and magazines. Many of us consume a significant portion of our reading material online, through webpages. However, reading on a backlit screen for extended periods can be [straining on the eyes](https://rob-hoeijmakers.ghost.io/the-influence-of-blue-light-how-to-protect-your-eyes-and-sleep/). This is where e-readers like the Kindle come into play, offering a more [comfortable reading experience akin to that of paper](https://rob-hoeijmakers.ghost.io/electronic-paper-and-digital-ink/). While I have previously explored [how to read webpages on a Kobo](https://rob-hoeijmakers.ghost.io/reading-web-pages-on-a-kobo-e-reader/), the Kindle offers a uniquely simplified process that enhances your reading experience. Here’s how you can transfer web content directly to your Kindle with ease. All the steps to simplify layout and transfer to PDF combined. Let's break it down in steps. ## Step 1: Navigate to the Webpage The first step is straightforward - simply go to the webpage you wish to read on your Kindle. Whether it's a long-form article, a blog post, or an online publication, any content that can be accessed through a web browser is eligible for this process. ## Step 2: Switch to Reader View Once on the webpage, enable the 'Reader View' if available. This view simplifies the page by stripping away most of the formatting, ads, and non-essential elements, leaving behind only the text and essential images. This step is crucial for ensuring that the content is optimised for e-reader viewing, providing a cleaner, more accessible reading experience. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/08/Reader-view.jpeg) To reader view on iPhone. ## Step 3: Print to PDF After activating the Reader View, the next step involves converting the webpage into a PDF document. This can usually be done by selecting the 'Print' option in your browser and choosing 'Save as PDF' as the destination. This step converts the webpage into a portable format while retaining the streamlined reading view, making it compatible with your Kindle. ![Print and sent to PDF.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/03/Print-to-PDF.jpg) Print and sent to PDF. ## Step 4: Send as Email This step showcases the Kindle's superior feature - the ability to send documents directly to your device via email. Each Kindle device and app is assigned a unique email address which you can use to send documents to your Kindle. Attach the PDF file you created in the previous step to an email, and send it to your Kindle's email address. ![Send PDF through e-mail to Kindle address.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/03/Send-PDF-as-Email.jpg) Send PDF through e-mail to Kindle address. ## On Your Kindle: Synchronise and Read Upon sending the email, pick up your Kindle and synchronise it with your Amazon account to download the document. Once downloaded, open the file from your Kindle's library. You'll find that the webpage is now perfectly formatted for your e-reader, offering a comfortable and distraction-free reading experience. ![An example of a simplified web page on the Kindle.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/03/IMG_0770.jpeg) An example of a simplified web page on the Kindle. --- *I compared Kobo and Kindle, it might be helpful:* --- ### Further reading - [Kobo versus Kindle](https://rob-hoeijmakers.ghost.io/kobo-vs-kindle/) - [Kobo and Kindle solve different problems](https://rob-hoeijmakers.ghost.io/kobo-and-kindle-solve-different-problems/) ### Web Traffic and the Rise of LLMs URL: https://rob-hoeijmakers.ghost.io/web-traffic-and-the-rise-of-llms/ Last updated: 2024-05-12T13:35:07.000Z In a compelling analysis by [Rand Fishkin for SparkToro](https://sparktoro.com/blog/who-sends-traffic-on-the-web-and-how-much-new-research-from-datos-sparktoro/), the dynamics of incoming web traffic are meticulously dissected, offering invaluable insights into the evolving sources of digital traffic. Until the recent past, Google reigned supreme as the primary source for visitors to my blogs, necessitating a deep dive into SEO strategies and frequent audits via Google Search Console for actionable insights and content ideation. However, the ascent of [Large Language Models (LLMs)](https://rob-hoeijmakers.ghost.io/a-comparative-analysis-of-top-language-models/) such as [ChatGPT](https://rob-hoeijmakers.ghost.io/try-out-chatgpt-yourself/) and Gemini, along with innovative applications like [Perplexity](https://rob-hoeijmakers.ghost.io/perplexity-the-ai-powered-answer-engine/) and the [Arc Browser](https://arc.net), created a paradigm shift in web traffic dynamics and personal browsing habits—a change likely mirrored in your experiences as well. In this article I share my personal experiences and impact of the rise of LLM’s as web strategist but also try some predictions of the road ahead. ## Content as a terminal stop Previously, Google's search engine served as the beacon, directing the flow of digital explorers to my content. Now, the spotlight shifts towards LLM-driven applications, a realm where my content now seems to reside and only occasionally signals its presence through sporadic traffic influxes. ![Traffic from Perplexity. That is only people who check beyond what is displayed there.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/03/2024-03-13_09-35-44.png) Traffic from Perplexity. That is only people who check beyond what is displayed there. This transition from a Google-centric world underscores a broader transformation in user interaction with digital content; within these new ecosystems, users exhibit a reluctance to browse beyond the confines of a single application, unlike the more navigable pathways provided by traditional search engines. The implications of such a shift extend beyond mere curiosity; they hint at a potential upheaval in the distribution and accessibility of digital content in the years to come. Observing my content through the lens of the Arc Browser, which curates its offerings from the top Google search results, is a bit frustrating. I see my content, text and images, displayed in their environment. No traffic for me. ![Content of my blog displayed in ArcBrowser.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/03/Arc-EN.jpg) Content of my blog displayed in ArcBrowser. Yet, it's not the likes of the Arc Browser (they are build on top of Google) that cast a shadow over my aspirations, but rather the LLMs that transform original content—imbued with creativity and effort—into a final destination, a terminal stop so to speak. There the content I make goes into the recipe for a big soup. 🤔 In examining the current dynamics of web traffic, it's evident that personal assistants such as Google Assistant, Siri, and Alexa have a tangible impact, though quantifying this precisely remains challenging. These technologies are fundamentally changing user interaction patterns with the web by facilitating voice searches and integrating with various online services and smart devices. This shift results in a redirection of traditional web traffic and necessitates a reevaluation of how we measure digital engagement. The direct influence of these assistants on web traffic is complex, involving factors like direct answers that bypass traditional web visits and the preference for conversational queries. ## Towards a Personalised Content Strategy This evolution prompts a reevaluation of my content strategy, foreseeing a shift from addressing a collective audience to engaging individuals through novel channels, outside the traditional web browser framework. I am at the forefront of this transformation, collaborating with clients to develop applications that house content within exclusive knowledge bases. These platforms are designed not for public display but for [personal interaction, accessible through chatbots, web interfaces, and even WhatsApp](https://rob-hoeijmakers.ghost.io/my-experiences-with-chatbots-based-on-chatgpt/). The pivotal role here is played by the knowledge base and by the LLMs that curate and convey it, crafting responses that resonate on a personal level without directly sourcing from the LLM's vast repository. I am highly curious how to work this into a model at scale. Probably on intuition because it is hard to measure and with cooperation with others to share ideas. Exciting times ahead! ### My experiences with chatbots based on chatgpt URL: https://rob-hoeijmakers.ghost.io/my-experiences-with-chatbots-based-on-chatgpt/ Last updated: 2025-12-17T18:29:44.000Z In the digital age we live in, communication technologies are constantly evolving to make our interactions more personal, efficient and meaningful. My latest adventure has taken me deep into the world of artificial intelligence (AI) and chatbots. I am working with the experts at [Enreach](https://www.enreach.com/en) to build a totally new chatbot based on ChatGPT and the [Retrieval-Augmented Generation (RAG)](https://rob-hoeijmakers.ghost.io/exploring-retrieval-augmented-generation-rag/) method. This project has fundamentally changed my perspective on digital interactions and in this article I am happy to share my experiences, the challenges we encountered and the new opportunities ahead. ## My personal background As a [web strategist with more than 30 years of experience](https://rob-hoeijmakers.ghost.io/about-rob-hoeijmakers/), I have seen countless digital products come and go. But the opportunity to [develop chatbots with Enreach](https://www.chatvoorbedrijven.nl/enreach-rob-hoeijmakers/) that not only provide information but also truly understand and anticipate user needs has genuinely surprised me. The level of personalisation and relevance offered by these chatbots is unparalleled. They can communicate in different languages, adapt the level of information to the user, and accurately understand the user's intent. This is a fundamental change for how organisations can interact with their audiences. ## The power of ChatGPT and RAG Before I dive into the details, let me explain what ChatGPT and RAG mean to the uninitiated. ChatGPT, developed by OpenAI, is an advanced language model that can simulate natural conversations, answer questions, and even generate content similar to that of a human. RAG adds an extra layer by retrieving relevant information from a large dataset to enrich ChatGPT's response, making the chatbot even more accurate and informative. ## The chatbot in action What better way to show the potential of this technology than through a direct demonstration? We have decided to release our chatbot as a beta version, accessible via a [click on the green circle at the bottom right of the *chat voor bedrijven* website](https://www.chatvoorbedrijven.nl/mijn-ervaringen-met-chatbots-op-basis-van-chatgpt/) or via our [WhatsApp channel](http://wa.me/312067657814). This step allows us to gather valuable feedback and see directly the impact of our chatbot on real users. ![Left a conversation on the web and right in WhatsApp](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/03/DraggedImage.png) Left a conversation on the web and right in WhatsApp ## From sending to interacting My immersion in building an advanced chatbot with ChatGPT and RAG challenged me in an unexpected way. For years, my approach to content creation focused on shaping web pages and strategically using hyperlinks to make information accessible and appealing to a wide audience. This approach required a rhetorical style tailored to the medium, with carefully chosen design and structure to attract and hold the reader's attention. The move to working with an AI-powered chatbot required a fundamental shift in my thinking. Suddenly, it was no longer enough to present information statically; I had to dive deeper into the architecture of the content itself, focus on the pure facts and structure it in a new way suitable for interactive conversations. Instead of broadcasting information, I am now involved in a dynamic, two-way conversation with users. This transition has forced me to rethink the way I think about content. In prompts, I can now specify the style and level of information, but the exact response remains partly unpredictable. This aspect of uncertainty introduces an exciting element of spontaneity and personalisation into the interaction, making each conversation unique. The core difference lies in the evolution of content creation from pure sending to an interactive conversation, where the user is central and plays an active role in steering the conversation. This new approach to content creation emphasises the importance of deep understanding of user needs and interests, as well as the flexibility to respond to these needs in real-time. It has been a learning experience that has enriched not only my skills as a content creator, but also my appreciation for the potential of AI in creating meaningful and personalised user experiences. It has also enriched my strategic understanding of how to place these new capabilities in the existing online landscape and the specific offerings of companies and organisations. More on this later. ![Dialox Knowledge Base, the content of the chatbot](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/03/Dialox-Kennisbank.jpg) Dialox Knowledge Base, the content of the chatbot ## Stay connected Feedback so far has been overwhelmingly positive. Users are particularly impressed by the natural conversation flow and the relevance of the information provided by the chatbot. Our next steps focus on refining the conversation dynamics to not only be informative, but also achieve other goals such as sales, support, and more advanced interactions. I cordially invite you to experience the chatbot for yourself and s[hare your feedback with us](https://rob-hoeijmakers.ghost.io/about-rob-hoeijmakers/). By sharing our experiences and lessons learned, we hope to not only inspire but also offer practical insights into the power of AI and chat technology. In the coming months, we will also unveil a new chatbot that provides international education on the RS virus, in [partnership with Resvinet](https://resvinet.org/chatbot/). [*This article was first published in Dutch on Chat voor Bedrijven*](https://www.chatvoorbedrijven.nl/mijn-ervaringen-met-chatbots-op-basis-van-chatgpt/)*.* ### Business Cards in a Digital World URL: https://rob-hoeijmakers.ghost.io/business-cards-in-a-digital-world/ Last updated: 2026-04-25T13:17:08.000Z In an age dominated by digital communication, the humble business card remains a surprisingly resilient tool for professionals. Despite advances in technology that allow for instantaneous, peer-to-peer data exchanges, the simplicity and tangibility of business cards offer unique advantages that digital methods have yet to surpass entirely. This article explores the evolution of business cards from their aristocratic origins to their current role in the professional world, alongside a personal experiment with digital alternatives that strive to blend the best of both worlds. ### The Aristocratic Beginnings and Evolution into Business Essentials The history of business cards can be traced back to 17th-century Europe, where they began as "visiting cards" used by the aristocracy to announce their arrival to their hosts. ![Visiting card or calling card. Kaiser Wilhelm.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/02/Visiting_Card_of_Kaiser_Wilhelm.jpg) Visiting card or calling card. Kaiser Wilhelm. These cards were a symbol of etiquette and social status, often elaborately designed with the cardholder's coat of arms. As commerce and trade flourished, the visiting card evolved into the business card, serving a more pragmatic purpose. By the 19th century, business cards had become an indispensable tool for professionals and tradespeople to share their contact information and advertise their services. ![Classic business card](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/02/BusinessCardAttorney1895.jpg) Classic business card This transformation from a social custom to a business necessity underscores the adaptability of business cards to the changing landscapes of communication and commerce. Despite the digital revolution, they have maintained their relevance by offering a personal touch in professional interactions—a tangible reminder of a meeting or conversation. ![My business cards](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/02/IMG_0214.jpeg) My business cards ### Bridging the Gap: Experiments with Digital Alternatives Recognizing the limitations of traditional business cards in a world where real-time, peer-to-peer data exchange is still a challenge, and so I explored digital alternatives that could complement or even replace the physical business card. My experimentation focused on [QR tags](https://rob-hoeijmakers.ghost.io/qr-code-explained/) for their platform independence and ability to function without a network connection. One of the first platforms I considered was LinkedIn. Despite its widespread adoption among professionals, I discovered that accessing the QR code for sharing contact information was not as intuitive as it could be, buried within the application's search function. Moreover, LinkedIn's ubiquity does not equate to universality, and the platform's networking nature may deter some from connecting purely for contact exchange. ![LinkedIn QR code in the app, and the wallet card I made myself.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/02/DraggedImage-2.png) LinkedIn QR code in the app, and the wallet card I made myself. ### A Novel Approach: The Apple Wallet Business Card Seeking a more accessible and pressure-resistant solution, I turned to the Apple Wallet—a logical counterpart for business cards in the digital realm. My approach involved a two-step process: 1. **Creating a Digital Data Card (VCF):** I [generated a VCF file](http://zxing.appspot.com/generator/) containing my contact details and encoded it into a QR code. 2. **Designing a Wallet Item:** I formatted this QR code into a wallet item, creating a visually appealing and branded digital business card (via [Wallet Creator](https://apps.apple.com/nl/app/wallet-creator/id1486573384?l=en-GB)). This method allows for the seamless sharing of contact information by presenting a branded wallet item, complete with a QR code. To accommodate different preferences, I also created a variant with my LinkedIn QR code. This dual-option strategy ensures that I'm prepared for various networking situations, blending the efficiency of digital exchange with the personal touch of traditional methods. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/02/Business-Cards-Wallet-2.jpg) Digital business card in the Apple Wallet and in action. Here, Paul Pschierer describes the whole process in his article: [How to make a digital business card for your Apple Wallet](https://www.linkedin.com/pulse/how-make-digital-business-card-apple-wallet-paul-pschierer-barnfather/) ### The Undeniable Role of Paper Business Cards Despite these digital innovations, I acknowledge the enduring value of paper business cards. They serve as a fail-safe option, unaffected by technological barriers or personal preferences towards digital platforms. Their simplicity, immediacy, and universal accessibility underscore their continued relevance in professional settings. The paper cards nowadays are affordable, handy, fail safe. I designed my card with [Canva](https://rob-hoeijmakers.ghost.io/canva). And if I don’t use them to share I can always can use it to jot something down. The digital alternative is probably too functional. And people see mobile phones as a sacred and private place, pushing data into it by QR code might be too pushy for a lot of people. Then the paper card is a better alternative and has a bit more style. And people see mobile phones as a sacred and private place, and pushing your own data into it with a QR code might be too intrusive for many people. Then the paper card is a better alternative and has a bit more style. And the functional part of sharing data might be better via e-mail or a LinkedIn request, for example. Then people can decide for themselves whether to include them. ### Mastering Transparent Backgrounds for Logos: Insights into PNG, WebP, and SVG Formats URL: https://rob-hoeijmakers.ghost.io/transparent-backgrounds-png-webp-svg/ Last updated: 2026-04-14T13:19:34.000Z Much like in music, where every piece is suited for certain orchestras, the choice of image format plays a crucial role in ensuring the logo performs harmoniously across various digital platforms. My adventure in logo design highlighted the importance of choosing between PNG, WebP, and, crucially, SVG formats, each with its unique advantages for handling transparency and ensuring design integrity across various backgrounds. This is what started my research. I have mad a logo and normally it is on a white background. But it also serves as icon in the browser and then it needs to work in dark mode. I noticed that the white became grey and I had to find a solution. ## The Challenge: Apply transparent background The central challenge I encountered was ensuring that the white elements of my logo remained visible against both light and dark backgrounds. I had made my logo transparent but then I suddenly noticed that everything I thought was white was also immediately transparent. As you can see below: ![My logo in dark mode, with a partial, white background and with a transparant background](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/02/16B8BA13-AAF8-454A-A1D9-EBE3BCFF4CD8_1_201_a.jpeg) My logo in dark mode, with a partial, white background and with a transparant background. By default, the background is displayed as white and so it can lead to misunderstandings where elements are also white. I work in [Canva](https://rob-hoeijmakers.ghost.io/canva) and I decided to bring in a white circle that now is the background of the logo. So behind the blue. The partial white background explained. I put a white circle up to prevent it from being transparent. ## Understanding Image Formats for Logos First thing I looked into is which image formats actually can be transparent and I found these three web formats: **PNG (Portable Network Graphics):** Favoured for its lossless compression, [PNG](https://rob-hoeijmakers.ghost.io/image-formats-for-the-web/) is ideal for logos requiring sharp edges and clear transparency. It ensures that logos retain their quality and detail without adding a background, making them versatile for various web backgrounds. **WebP:** A newer format designed for the web, [WebP](https://rob-hoeijmakers.ghost.io/webp-avif/) offers efficient compression methods, reducing file sizes significantly compared to PNG, without a notable loss in quality. It supports transparency, making it a good choice for web applications where speed and performance are considerations. **SVG (Scalable Vector Graphics):** Unlike PNG and WebP, which are raster formats, [SVG is vector-based](https://rob-hoeijmakers.ghost.io/what-is-svg-and-what-is-it-used-for/). This means it uses mathematical equations to define shapes and colors, allowing for infinite scalability without loss of quality. SVGs are particularly suited for logos due to their scalability, transparency support, and small file sizes. They perform exceptionally well across different devices and resolutions, including high-DPI (dots per inch) displays. 💡 Tip: ****Using ChatGPT**. Send ChatGPT (Dall-e) an image, e.g. JPG, and ask it to make the background transparent and available as a PNG. ## Consider SVG The realization that SVG could offer a more elegant solution prompted a reassessment of my approach. SVG's ability to scale without losing quality meant that the logo could be used in a wide range of sizes without needing different files for each use case. Moreover, editing the SVG file to adjust colors or enhance visibility on different backgrounds is straightforward, ensuring that all elements of the logo remain distinct and visible, regardless of the theme. For logos, SVG offers a distinct advantage: it's possible to adjust the CSS directly within the SVG file or via external stylesheets, enabling dynamic changes such as color adjustments for dark mode. This level of control is particularly beneficial for maintaining brand consistency across various platforms and themes. But my logo is small on the website and can be used without having to make any changes with CSS. I tested it and SVG turned out to result in a much larger file in my case. **So I opted for PNG**. The SVG is useful if I wanted to set up a booth, for example, or put the logo on a hot air balloon. ![In Canva you can save a design in PNG or SVG and opt for transparent background.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/02/2024-02-23_10-54-32.png) In Canva you can save a design in PNG or SVG and opt for transparent background. ## Choosing the Right Format When finalizing the logo design, considering the format became a strategic decision: - **PNG** remains a strong candidate for static images where lossless quality is paramount. - **WebP** is optimal for web use where performance and speed are critical, without significantly compromising on quality. - **SVG** stands out for logos due to its scalability, transparency, and ease of editing, making it the preferred choice for ensuring logo visibility and adaptability across different backgrounds and resolutions. The journey through the intricacies of image formats underscored a key lesson: the choice of format can significantly impact the effectiveness and versatility of a logo. By leveraging the strengths of SVG for logos, alongside understanding when to use PNG and WebP, designers and strategists can ensure that their logos not only look sharp but also transition seamlessly between light and dark modes, enhancing brand presence across digital platforms. Incorporating SVG into your logo design strategy not only addresses the visibility challenge across themes but also embraces the future of web design, where scalability, performance, and adaptability are key. And WebP comes to play when you use plugins like LiteSpeed caching and other ways to [optimize images](https://rob-hoeijmakers.ghost.io/what-are-image-optimization-services-for/). These solutions might take PNG and compress and transform it into the more lightweight WebP format. 💡 I noted that for the favicon, the icon you see in the browser for the page, cannot yet be WebP and needs to be PNG. I here include a PNG, a WebP and a SVG image, all with a transparent background, and you can check the difference in quality yourself. ![PNG](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/02/Logo-Rob-Hoeijmakers-Square-Dark-Glow.png) PNG, 49 KB ![WebP](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/02/Logo-Rob-Hoeijmakers-Square-Dark-Glow.webp) WebP, 29 KB ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/02/Logo-Rob-Hoeijmakers-Square-Dark-Glow.svg) SVG, 115 KB ### Payment requests: Your digital payment assistant URL: https://rob-hoeijmakers.ghost.io/payment-requests-your-digital-payment-assistant/ Last updated: 2026-05-10T08:54:13.000Z In the world of freelancers and small business owners, finding simple, quick ways to get paid is essential. Fortunately, we live in an era where technology comes to our rescue. In this blog article, I dive into the world of payment requests, a tool I recently embraced for my short services and smaller amounts. And I have to be honest, I had been using [Tikkie](https://www.tikkie.me/) for personal payments for some time, but had never really understood it that well until I opened a business account with a new bank ([Bunq](https://www.bunq.com/)) that does not use Tikkie. ## What is a payment request? A payment request is basically a digital account. It is a temporary web page with a unique URL, which displays the details of the payment request and various payment options such as iDeal. The beauty is that you can share this URL with whoever you want to pay, without having to be physically present or exchange bank details. Think of it as your own little online cash register that you set up for each transaction and which is especially great for receiving and processing on mobile. 💡 ****Difference between payment request and payment page?** A ****payment request** is a temporary, personal invitation to pay a specific amount, often accompanied by a message and sent via an anonymous, unique URL that expires after use or after a certain period of time. In contrast, a ****payment page** is a permanent, publicly accessible web page with a fixed URL, where users themselves enter the amount to be paid; this page is designed for continuous accessibility and can be customised to a company's branding, making it suitable for repeated transactions or donations. ![Payment request with amount, and a payment page with fixed URL where an amount can be entered.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/02/Betaalverzoek.jpg) Payment request with amount, and a payment page with fixed URL where an amount can be entered. ## Sharing the payment request Once your payment request is ready, you get a link that you can immediately share with the paying party via email, SMS, WhatsApp, or even social media or [QR-code](https://rob-hoeijmakers.ghost.io/qr-code-explained/). If you use banking apps like Bunq, the app can make this process even easier by sending the request on your behalf. This removes any potential obstacles in the communication process, making it as smooth as possible for both parties. But then, should there be any misunderstandings or the recipient did not receive a message, it is good to know that you can also share a payment request, via the address, yourself via SMS, WhatsApp or email. ![Creating and sending a payment request via Bunq](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/02/Payment-request-EN.png) Creating and sending a payment request via Bunq ## Managing open requests One small misunderstanding I would like to clear up is access to open payment requests. If a payment request has not yet been fulfilled, it is important to remember that only the sender of the request has visibility into these open transactions, not the recipient or payer. This provides an opportunity for the sender to manage requests, send reminders or withdraw a request if the situation changes. It is a digital reminder for the sender that money still needs to come in. ![Regeltje uit mijn betalingen. Deze regel krijgt andere bewoording en kleur als het bedrag betaald is. ](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/02/DraggedImage-1.png) Regeltje uit mijn betalingen. Deze regel krijgt andere bewoording en kleur als het bedrag betaald is. ## After payment The moment the payment is made, the open request turns into a confirmation of the completed transaction. For accounting purposes, I then create an invoice in [MoneyBird](https://www.moneybird.com) and link the payment details to this invoice. If the customer needs an official invoice, I can easily send it via email. This keeps my bookkeeping tight and transparent, with no extra hassle. 💡 Bunq offers another handy feature that sits between payment request and payment page. You take the address of your payment page and add the amount to the URL. Then the amount appears on the payment page. This can be quite handy if you want to quickly present an amount, but don't want to go through all the steps of a payment request. An example: ### Further reading - [Invoices in the Eurozone: A Practical Guide to SEPA Payments and Digital Processing](https://rob-hoeijmakers.ghost.io/eurozone-invoices-sepa-qr-ubl/) ### The AI Image Technology Landscape for images: Generation, Alteration, and Recognition URL: https://rob-hoeijmakers.ghost.io/ai-image-technology-landscape/ Last updated: 2026-04-25T13:17:06.000Z In the rapidly evolving world of artificial intelligence, the capabilities it brings to image processing are nothing short of revolutionary. As a professional deeply entrenched in the digital content sphere, I've had an epiphany that the myriad of AI technologies available today can be broadly classified into three distinct categories: image generation, image alteration/enhancement, and image recognition. Understanding these categories not only demystifies the technological deluge but also opens up new avenues for creativity and efficiency in our day-to-day tasks. ## Image Generation: The Art of Creation from Text The first category, image generation, represents the cutting-edge frontier where AI transforms textual descriptions into visual masterpieces. Platforms like [DALL·E ](https://rob-hoeijmakers.ghost.io/playing-with-ai-system-dall-e/)and [Midjourney](https://rob-hoeijmakers.ghost.io/start-with-midjourney/) stand at the forefront of this innovation, showcasing the ability to generate images that are not just inspired by existing visuals but are entirely new creations. This technology heralds a new era of creativity, where the barriers between imagination and reality blur. **Everyday Example**: Imagine crafting a blog post about future cities. Instead of scouring the internet for images that match your vision, you could use an AI generator to create a cityscape that aligns perfectly with your description, say, "a futuristic city at dusk, with flying cars and towering skyscrapers," bringing your written content to life in ways previously unimaginable. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/08/2024-02-14_16-29-41-Large.jpeg) Generate a new image based on text through Canva ## Image Alteration/Enhancement: Refining the Existing The second category encompasses technologies focused on altering or enhancing existing images. This includes tasks such as magic selection, colourisation, background changes, and more. Tools like [Canva's Magic Studio](https://rob-hoeijmakers.ghost.io/canva) or Firefly exemplify how AI can streamline the editing process, making professional-grade visuals more accessible to everyone. **Everyday Example**: You have a perfect team photo that you want to use for your company's homepage, but the background is less than ideal. Instead of a manual, time-consuming edit, AI-powered background removal tools allow you to effortlessly place your team against a more fitting backdrop, ensuring your first impression is exactly as you intend. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/08/2024-02-14_16-47-16.jpeg) Expand an existing image Here you can read and watch Canva's Magic Studio. I use them a lot and they offer a nice suite of AI tools: [Introducing Magic Studio: the power of AI, all in one place](https://www.canva.com/newsroom/news/magic-studio/) ![Canva Magic Studio](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/02/Canva-magic-studio.jpg) Canva Magic Studio ## Image Recognition: Understanding and Organising Visual Data The final category, image recognition, involves AI's ability to identify and categorise elements within images. This technology powers the facial recognition features in Apple Photos and Google Photos, [sorting billions of images with ease](https://rob-hoeijmakers.ghost.io/iphone-live-text/). Beyond organising personal photo libraries, image recognition aids in enhancing security systems, streamlining retail experiences, and more. **Everyday Example**: Consider the convenience of searching for "birthday party" in your photo library and instantly being presented with all relevant photos, spanning years of memories. This magic is made possible by image recognition technology, which analyses your photos for faces, objects, and scenarios, making it effortless to relive those special moments. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/08/2024-02-14_16-34-56-Large.jpeg) Search through a collection of images with text Additionally, AI models like [ChatGPT](https://rob-hoeijmakers.ghost.io/try-out-chatgpt-yourself/) can analyse and describe images, providing context and understanding that bridge the gap between visual content and textual description. ![ChatGPT describing a picture. First what we see and then I ask it which part of the earth we see. ](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/02/image-2.png) ChatGPT describing a picture. First what we see and then I ask it which part of the earth we see. ### A Mifi hotspot for my mountain holidays URL: https://rob-hoeijmakers.ghost.io/a-mifi-router-for-my-mountain-holidays/ Last updated: 2026-05-10T08:54:13.000Z As summer approaches, the allure of mountain getaways beckons, promising crisp air, breathtaking vistas, and a respite from the bustle of everyday life. However, for those like me who crave connectivity even amidst the serene beauty of nature, staying online can be a challenge, especially when venturing off the grid in places like Switzerland, where roaming rules can be as jagged as the Alpine peaks. ## Unravelling the Mystery of Mifi hotspot Enter the TP-Link M7200 Mifi Router, a pocket-sized gadget that promises to keep you connected even in the most remote of locales. But before delving into how this little device became my holiday saviour, let's first unravel the mystery of what exactly a Mifi router is. In essence, a Mifi router, short for "Mobile Wifi," serves as a portable hotspot, harnessing mobile data and distributing it through a Wifi network. It's your own personal gateway to the internet, allowing you to stay connected wherever you roam, whether it's a bustling city or a tranquil mountainside. But you can also use it for laptop or e-reader. Or to offer access to visiting guests. ![Unboxing, without a lot of ceremony. ](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/02/IMG_9823.jpeg) Unboxing, without a lot of ceremony. ## Finding the Perfect Solution My quest for the perfect solution to staying connected during our Swiss sojourn led me to the TP-Link M7200\. Previous attempts involved juggling spare iPhones and cumbersome hotspot configurations, but the M7200 offered a sleek and simple alternative. With its easy setup and compatibility with any [SIM card](https://rob-hoeijmakers.ghost.io/from-physical-sim-to-esim-on-iphone/), the M7200 was a breeze to get started with. I simply inserted my SIM card into the device, along with the battery, and voila – instant Wifi network. A nifty adapter ensured compatibility with any SIM format, eliminating the need for fiddly adjustments. ![Holding the router in my hand to show the scale.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/02/IMG_9893.jpeg) Holding the router in my hand to show the size. ## The Magic of the App The real magic, however, lay in the accompanying app. With clear instructions provided within the device, I quickly connected to the M7200's Wifi network and downloaded the iPhone app. From there, the app guided me through the detailed setup process, ensuring seamless integration with my mobile data plan. One particularly handy feature of the M7200 is its ability to send and receive SMS messages, providing a convenient interface for managing data bundles. In Switzerland, where roaming regulations can be a maze to navigate, this feature proved invaluable. Also it can display a [QR code](https://rob-hoeijmakers.ghost.io/qr-code-explained/) to quickly setup other mobile phones, for instance for guests. Thanks to a budget-friendly data plan offering unlimited usage for a mere fraction of the usual cost, my wife and I could enjoy uninterrupted connectivity throughout our stay. ![Screens of the Mifi app](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/02/DraggedImage.png) Screens of the Mifi app ## Looking Ahead As I eagerly await the start of our holiday, I'm filled with anticipation at the prospect of staying effortlessly connected amidst Switzerland's stunning scenery. With the TP-Link M7200 Mifi Router by my side, I can embark on our mountain adventures with confidence, knowing that the digital world is never more than a click away. Stay tuned for an update once our holiday begins – here's to smooth sailing (or should I say, smooth surfing) in the Swiss Alps! — During my holidays I gave it a good test and struggled a bit to keep my data in check. It turned out it was not unlimited but 1 gigabyte. Read my [article on how I tamed data](https://rob-hoeijmakers.ghost.io/taming-data-usage-with-a-mifi-hotspot-my-experience-and-solution/). --- ### Further reading - [Taming Data Usage with a MiFi hotspot: My Experience and Solution](https://rob-hoeijmakers.ghost.io/taming-data-usage-with-a-mifi-hotspot-my-experience-and-solution/) - [From Mifi to eSIM: Finally Catching Up with Mobile Connectivity](https://rob-hoeijmakers.ghost.io/from-mifi-to-esim-finally-catching-up-with-mobile-connectivity/) - [From physical SIM to eSIM on iPhone](https://rob-hoeijmakers.ghost.io/from-physical-sim-to-esim-on-iphone/) ### From physical SIM to eSIM on iPhone URL: https://rob-hoeijmakers.ghost.io/from-physical-sim-to-esim-on-iphone/ Last updated: 2026-06-07T20:32:08.000Z In this article, I explore my decision to transition from a traditional physical SIM card to an embedded SIM (eSIM) on my iPhone, and how I utilised the now-free slot for a second number dedicated to business purposes. This change aimed at better managing personal and business communications on a single device, promising convenience and an improved work-life balance. ## Why the Switch? The motivation behind adopting an eSIM and adding a second physical SIM was driven by the desire for greater flexibility in managing my digital life. By switching to an eSIM for my personal number via Vodafone NL and selecting a different provider for my business line, I aimed to streamline my communications without carrying multiple devices. ## Step-by-Step Process through Vodafone NL App Transitioning to an eSIM was a straightforward process, facilitated by the Vodafone NL app. Here's how I did it: 1. **Initiate the eSIM Setup:** Within the Vodafone NL app, I followed the prompts to convert my physical SIM to an eSIM. 2. **Follow Instructions:** The app provided detailed instructions, including scanning a [QR code](https://rob-hoeijmakers.ghost.io/qr-code-explained/) to install the eSIM profile. 3. **Activate the eSIM:** After installation, I ensured the eSIM was active and functioning correctly before proceeding to install the physical SIM for my business number. ![Steps I took to transfer from SIM to eSIM.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/02/eSIM.jpg) Steps I took to transfer from SIM to eSIM. ## Configuring the Second SIM on iPhone Installing and configuring the second SIM for business use involved: 1. **Installation:** With the eSIM active, I inserted the physical SIM into the iPhone's slot. 2. **Configuration:** In Settings, I set up the new line, labelling each number (“Primary and “Secondary”) for easy identification. 3. **Adjust Settings:** I fine-tuned settings for messaging, calls, and data usage to ensure a seamless management of both lines. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/02/Two-SIMS-in-iPhone.jpg) Basic setup of iPhone with two SIM's ## Day-to-Day Experiences The dual-SIM setup significantly improved my daily communication management: - **Separation of Communications:** Labelling and managing contacts for each number helped keep personal and business communications distinct. - **WhatsApp Management:** Using [WhatsApp Messenger and Business simultaneously](https://rob-hoeijmakers.ghost.io/whatsapp-messenger-and-business-app-on-the-same-iphone-with-a-virtual-number/) showed no issues, maintaining clear separation between personal and business chats. - **Battery Life:** Preliminary observations suggest minimal impact on battery life, though further monitoring is needed. [WhatsApp Messenger and Business on a single iPhoneExplore using WhatsApp Messenger and Business on one iPhone. Learn about the dual experience in this informative article.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/size/w256h256/2024/05/Logo-Rob-Hoeijmakers-Square-Dark-Glow-3-1.png)Web StrategiesRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/size/w1200/2023/02/IMG_7197.jpg)](https://rob-hoeijmakers.ghost.io/whatsapp-messenger-and-business-app-on-the-same-iphone-with-a-virtual-number/) ## Increased Flexibility versus Potential Lock-In This setup offers undeniable flexibility and convenience, though it raises questions about being more tied to the Apple ecosystem. Despite potential concerns about increased reliance on Apple, the freedom from being bound to a single phone service provider is liberating. 💡 ****Update after 4 months**: First impressions of using the dual SIM set up is that there were only positives from using two SIM's. More data to use, the option to use a different network, no impact on daily use. For online professionals seeking to streamline their communication tools, transitioning to an eSIM and utilising a dual-SIM setup can offer significant advantages. It also opened up my eyes on all sort of roaming deals you can put on your phone when travelling. Even by scanning a [QR-code](https://rob-hoeijmakers.ghost.io/qr-code-explained/) or sending an eSIM from one iPhone to another. ![Setting up eSIM on iPhone with sharing or with QR code](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/02/eSIM-setup-EN.png) Setting up eSIM on iPhone with sharing or with QR code --- ### Further reading - [A Mifi Router for my mountain holidays](https://rob-hoeijmakers.ghost.io/a-mifi-router-for-my-mountain-holidays/) ### Transitioning to Passkeys URL: https://rob-hoeijmakers.ghost.io/transitioning-to-passkeys/ Last updated: 2026-04-21T09:26:35.000Z **A Practical Guide for the Advanced Online Professional** As we navigate the evolving landscape of online security, the shift from traditional passwords and [two-factor authentication (2FA)](https://rob-hoeijmakers.ghost.io/two-factor-authentication-2fa/) to Passkeys represents a significant change. This guide aims to share my personal experience with this transition, focusing on the practical aspects and offering insights for professionals ready to make the jump. ## Understanding Passkeys Passkeys are a new form of authentication designed to replace passwords. Unlike passwords, [Passkeys](https://rob-hoeijmakers.ghost.io/passwords-passkeys/) do not require you to remember complex strings of characters. Instead, they use a more secure, cryptographic approach to verify your identity. This technology promises to make logins both simpler and more secure. ## My First Encounter with Passkeys My journey began with two platforms: Amazon and MoneyBird. On **Amazon Netherlands**, I was given the option to continue using passwords or to try Passkeys. The choice was comforting; I could explore this new method without losing the familiar backup of passwords. ![Amazon Netherlands is providing passkeys.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/02/Passkey-Amazon-NL.jpg) Amazon Netherlands is providing passkeys next to password. **MoneyBird**, however, pushed me deeper into the Passkey waters. Once I opted for Passkeys, reverting to passwords was not an option. This moment felt like a real commitment to embracing a more secure online environment. ![Moneybird uses passkeys and when you opt for them you loose password access.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/02/Passkey-Moneybird.jpg) Moneybird uses passkeys and when you opt for them you loose password access. ## The Experience on macOS and iOS Implementing Passkeys on my devices was straightforward. The technology is well-integrated into both macOS and iOS, making the transition smoother than I anticipated. However, adapting to this new method did require a mindset shift. Initially, there was a learning curve, and I had to remind myself not to default to the password option. ## Challenges and Considerations The biggest challenge was not the technology itself but adjusting to a new routine and different wordings. The initial setup and getting used to a different way of logging in took some time. On a practical level, the experience varied between websites on desktop and mobile, and mobile apps, which meant I had to stay flexible and patient. ## Advice for Professionals Considering Passkeys Here is some advice from my hands on learnings: 1. **Start Slow:** Choose one or two platforms to begin with, preferably ones that offer an option to switch between Passkeys and passwords. 2. **Educate Yourself:** Understand how Passkeys work with your operating system and devices. This knowledge will ease the transition. 3. **Be Patient:** Adjusting to Passkeys takes time. Allow yourself to get comfortable with the new process. 4. **Stay Informed:** As more platforms adopt Passkeys, staying updated on best practices and security tips is crucial. ## Why Make the Jump? Adopting Passkeys is about more than following a trend; it's a step towards a more secure online presence. While the transition requires an adjustment period, the benefits—enhanced security and a simpler login process—are compelling reasons to make the switch. For those of us in the digital profession, understanding and embracing such changes is part of staying ahead in our field. By sharing my journey, I hope to offer a practical perspective that can help guide you through your own transition to Passkeys. I will provide further updates once I am confident to have mastered it more completely, keep updated by [subscribing for latest articles](#/portal/signup). ![Also Google will enable Passkeys.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/02/Google-Passkeys.jpg) Also Google will enable Passkeys. ### Setting up a Google Account without Gmail URL: https://rob-hoeijmakers.ghost.io/google-account-without-gmail/ Last updated: 2026-05-10T08:54:13.000Z While Gmail is undeniably a powerful email service, the quest for more personal control and seamless integration within the Apple ecosystem has led many to explore alternative setups. My recent experience with setting up a Google Account without Gmail, opting instead for an email service provided by [Hostinger](https://hostinger.com?REFERRALCODE=1ROB567), has been nothing short of revelatory. This approach not only offers a tailored email experience but also integrates effortlessly with Google's suite of services, including the superior capabilities of Google Calendar and the collaborative features of Google Meet. ![My Google account without Gmail, but with a Google One subscription.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/02/Google-account-without-Gmail-2.jpg) My Google account without Gmail, but with a Google One subscription. ### Why Choose a Non-Gmail Google Account? Choosing to use a non-Gmail account for Google services stems from a desire for more control over email management and a preference for specific email clients. For those of us deeply integrated into the Apple ecosystem, the native email clients, despite their shortcomings in areas like search functionality, offer a level of integration that enhances our daily workflows. ### The Setup Process Setting up a Google Account without Gmail is straightforward. The absence of a Gmail account allows for the use of your own domain name as the Google Account username, providing a level of personalisation and branding not achievable with a standard Gmail address. This process involves: 1. [Signing up for a Google Account through a special link](https://accounts.google.com/SignUpWithoutGmail) using your preferred email address. 2. Verifying the email address to ensure it's active and accessible. 3. Configuring the account settings to align with your personal and business needs. ![Sign up through special flow: accounts.google.com/SignUpWithoutGmail](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/02/Google-Account-without-Gmail.jpg) Sign up through special flow: accounts.google.com/SignUpWithoutGmail ### Integrating with Google One Integrating this setup with [Google One's Family Plan](https://rob-hoeijmakers.ghost.io/choosing-google-one-to-use-google-meet/) has been a game-changer. At no additional cost, it provides 2TB of storage and premium features for Google Meet, not to mention the branding opportunities it affords to my domain through Google Calendar and [Google Scheduling options](https://rob-hoeijmakers.ghost.io/from-calendly-to-scheduling-with-google-calendar/). This setup is particularly beneficial for solo entrepreneurs or small business owners seeking cost-effective solutions without compromising on functionality. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/02/Google-One-EN.jpg) My Google One account is a family account with multiple members. ### The Benefits - **Custom Email Domain**: Utilising a custom domain for your Google Account enhances your brand's professionalism and visibility. - **Seamless Integration**: Despite minor limitations with certain Google Calendar features, the overall integration within the Apple ecosystem and with Google's services is seamless. - **Cost-Effectiveness**: The ability to join a Google One Family Plan without extra costs for additional storage and premium features represents significant savings. ### Considerations It's worth noting that some Google Calendar features, such as meeting changes, are optimized for Gmail and may not display as intended in other email clients. However, this minor drawback is far outweighed by the benefits of increased control, integration, and the professional edge provided by using a custom email domain. For individuals and small business owners alike, setting up a Google Account without Gmail, integrated with Google One, offers a compelling blend of control, customisation, and cost-effectiveness. Whether you're deeply entrenched in the Apple ecosystem or simply seeking a more tailored email experience, this setup provides a solid foundation for managing your digital presence and business operations. --- ### Further reading - [Google One subscription for Google Meet](https://rob-hoeijmakers.ghost.io/choosing-google-one-to-use-google-meet/) ### Exploring Retrieval Augmented Generation (RAG) URL: https://rob-hoeijmakers.ghost.io/exploring-retrieval-augmented-generation-rag/ Last updated: 2024-02-09T15:36:56.000Z In the realm of digital innovation, the quest for creating more intelligent, responsive, and context-aware chatbots has led me down a fascinating path: the implementation of Retrieval Augmented Generation (RAG). This journey, filled with trials, triumphs, and endless learning, showcases the transformative potential of combining large language models like ChatGPT with curated knowledge bases. Let me take you through the process, the challenges, and the intricate dance between retrieval and generation that underpins this cutting-edge approach. ## The Genesis of a Chatbot The inception of this project was driven by a simple yet ambitious goal: to develop a chatbot that transcends the limitations of traditional models by leveraging the vast expanse of content available on a designated website. The vision was to create a system that not only interacts with users but does so with unprecedented relevance and accuracy. To bring this vision to life, we began by ingesting articles from the website into the chatbot's knowledge base through an API, laying the foundation for a deeply informed conversational agent. ![Knowledgebase in Dialox, the system by Enreach I am using.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/02/image-1.png) Knowledgebase in Dialox, the system by Enreach I am using. ## The Art of Enrichment With the knowledge base in place, the next step was to enrich the articles with keywords, a task elegantly handled by ChatGPT. This process of enrichment was not merely about tagging; it was about understanding the essence of each article and distilling it into searchable, actionable keywords. These enriched articles, now stored with their newfound descriptors, formed the backbone of our retrieval-augmented strategy. ## The Dance of Retrieval and Generation When a user poses a question to the chatbot, the real magic begins. The query, along with the enriched articles and their keywords, is presented to ChatGPT 3.5. The model then embarks on a sophisticated matchmaking process, identifying the article or articles that most closely align with the user's inquiry. This is the retrieval phase, a crucial precursor to the generation of a response. But the journey doesn't end there. With the relevant article in hand, another request is made to ChatGPT. This time, the model is tasked with generating a response based on the full article content, effectively serving the user with information that is not just accurate but also richly informed by the curated knowledge base. This is Retrieval Augmented Generation in action: a synergy of retrieval and generative processes that elevates the chatbot's responses from informative to insightful. ![Diagram of the steps involved in Retrieval Augmented Generation (RAG)](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/02/image.png) Diagram of the steps involved in Retrieval Augmented Generation (RAG) ## The Three Levers of Optimization Our adventure into the realm of RAG revealed three critical levers for optimizing responses: curation, retrieval prompts, and generation prompts. The curation of text, through the selection of appropriate keywords and summaries, ensures that the retrieval process is grounded in relevance. The crafting of the retrieval prompt, a delicate art in itself, guides the model towards the most pertinent articles. Finally, the generation prompt shapes the model's response, turning raw information into coherent, engaging answers. 💡 Example prompt for answering: **End the next conversation ONLY by using the KNOWLEDGE you have been given above AND do NOT make up anything yourself. Respond in a formal, professional tone appropriate to a business environment. Make sure the response is concise, clear and polite and avoid all forms of informal language, colloquialisms, humour and non-professional content such as song lyrics or jokes. Respond in the user's language. Remove URLs in the replies you give.* ## Reflections on the Journey Embarking on this project was more than a technical challenge; it was a voyage of discovery. By integrating Retrieval Augmented Generation into our chatbot, we ventured beyond the conventional boundaries of AI-driven conversation. We created a system that doesn't just answer questions but does so with a depth of understanding and relevance that was previously unattainable. This journey highlights the incredible potential of retrieval-augmented approaches in the realm of chatbots. As we continue to explore this frontier, the promise of creating ever more intelligent, responsive, and context-aware conversational agents beckons, inviting us to redefine the boundaries of what is possible. ### LiveDrop: Offline Data Sharing URL: https://rob-hoeijmakers.ghost.io/exploring-livedrop-offline-data-sharing/ Last updated: 2026-04-25T13:16:36.000Z Recently, I came across a mention of LiveDrop in the news, sparking my curiosity about this innovative technology. LiveDrop offers a unique method for sharing information directly from one phone to another without the need for a network connection, bypassing traditional requirements for 5G or WiFi. This transfer is facilitated **from screen to camera**, functioning like an advanced [QR code](https://rob-hoeijmakers.ghost.io/qr-code-explained/). I had an interview with [Patrick Moreu from LiveDrop](https://www.linkedin.com/in/patrickmoreu/) and the idea and value of the concept of sharing data without network quickly became very tangible. ## The Practicality of LiveDrop in Everyday Scenarios The necessity for both the sender and receiver to have an app installed presents a bit of difficulty on the practical application of such a feature. While the concept is intriguing for its privacy and convenience, the requirement for app installation can be a barrier in situations where quick sharing is essential or when network connectivity is unavailable. This limitation seems particularly relevant in crowded events or venues, such as conferences or restaurants, where network congestion is common. Although these scenarios highlight potential use cases, the need for a prior app installation could dampen the spontaneity of sharing moments or information. The ideal solution would be the integration of LiveDrop as a native platform feature across iOS and Android devices, similar to the ubiquity of QR codes today. However, the challenges of achieving such integration with major tech companies could be substantial, raising questions about feasibility and the potential for these companies to adopt or adapt the concept independently. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/02/LiveDrop.jpg) Example of sharing an image with LiveDrop ## LiveDrop in Healthcare Despite initial skepticism regarding its practicality and scalability, my perspective shifted following an interview with Patrick Moreu from LiveDrop. He reiterated the technology's focus on privacy and security, demonstrating its effectiveness even through screens. However, the real breakthrough came when discussing LiveDrop's application in a healthcare setting. Collaborating with a regional hospital, LiveDrop was integrated into the hospital's app, enabling doctors to share medical reports instantly with patients. In PDF format or any other data from the Electronic Patient Dossier (EPD), like particular data fields from the dossier or images. This approach not only bypasses the need for complex network infrastructure but also eliminates the hurdles of account creation, verification, and authentication. By enabling secure and immediate information sharing, LiveDrop offers a compelling solution for healthcare, where confidentiality and accessibility are paramount. Explanation of sharing medical information through LiveDrop ## The Potential of LiveDrop LiveDrop's innovative approach to data sharing, particularly its application in sensitive environments like healthcare, demonstrates its potential beyond initial impressions. It encourages us to reconsider the possibilities of technology in facilitating secure, efficient, and accessible information exchange without reliance on traditional network infrastructures. For those intrigued by the blend of technology and privacy, exploring LiveDrop further might reveal more about its capabilities and future applications. Download the app and check out the [LiveDrop website](https://www.livedrop.eu) if you see potential for your use cases. 💡 In the use cases for LiveDrop I focussed on mobile technology but the codes for sending and receiving data are also feasible for Windows and Linux. This is of course especially handy for generating the codes and having it picked up my the camera of a mobile phone. ### Kobo versus Kindle URL: https://rob-hoeijmakers.ghost.io/kobo-vs-kindle/ Last updated: 2026-05-10T08:54:14.000Z In the world of e-readers, Kobo and Kindle stand as prominent contenders, each offering unique experiences within their ecosystems. My personal journey, which began with a Kindle Oasis, shifted to a Kobo Clara, and circled back to Kindle, has provided me with a nuanced perspective on what these platforms offer beyond just the devices. ## Content Availability and Language Support The Kindle store, under Amazon's vast umbrella, boasts an extensive collection of titles. However, in certain markets like the Netherlands, Kobo gains an edge due to its better support for local languages. This diversity was a key factor in my initial switch to Kobo, but the comprehensive reach of Amazon's Kindle store cannot be overlooked. ## Content Management and Format Flexibility Navigating between formats like **ePub** (Kobo) and **mobi** (Kindle) using [Calibre](https://calibre-ebook.com/download) has been a game-changer for me, offering freedom from platform constraints. This level of technical engagement, while rewarding, might not be for everyone but it's a significant consideration for those who wish to explore beyond a single ecosystem. ## Subscription Services Kobo Plus and Kindle Unlimited each present a unique set of offerings, complemented by alternatives such as digital borrowing from public libraries. This expands the horizon for digital readers, providing multiple avenues to access content. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/02/IMG_9704.jpeg) Kobo Plus, the subscription service, well integrated in the platform and device. ## Additional Features Kindle offers a range of additional features that set it apart from its competitors, such as Kobo. These features are not only designed to enhance your reading experience but also offer valuable tools for study, collaboration, and exploration. In this section, we will delve into some of Kindle's standout capabilities, including content sharing, cross-platform compatibility, and its impressive internet browsing functionality. Let's explore these aspects in greater detail. ### Content Sharing Kindle's advanced features for sharing highlights and notes are invaluable for study and collaborative reading, giving it an edge over Kobo, which is more limited in this regard. Also, the Kindle has an e-mail address where you can send books and PDFs to, and I do notice that is an big plus in everyday use. Like when [sharing web pages](https://rob-hoeijmakers.ghost.io/reading-webpages-on-your-kindle/), for example. ### Cross-Platform Sharing Sharing books across platforms, particularly between my Kindle and my partner's Kobo, highlights the need for better interoperability in digital reading ecosystems. [Share Kobo e-books with your familyA step-by-step guide on how to share your Kobo e-books with your partner and other family members.![](https://webstrategiesblog.com/content/images/size/w256h256/2024/05/Logo-Rob-Hoeijmakers-Square-Dark-Glow-3-1.png)Web StrategiesRob Hoeijmakers![](https://webstrategiesblog.com/content/images/size/w1200/2022/03/two-kobo-ereaders.jpeg)](https://rob-hoeijmakers.ghost.io/share-kobo-ebooks-with-your-partner/) ### Internet Browsing A notable aspect where Kindle outshines Kobo is its internet browser. The superior browsing capability on Kindle adds another dimension to the device, integrating it more seamlessly with the wider online world. This feature not only enhances user experience but also broadens the scope of the platform beyond conventional e-reading. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/02/IMG_9703.jpeg) Browser on the Kindle ## Personal opinion My preference leans towards Kindle, given its user-friendly nature and features conducive to a comprehensive reading and studying experience. However, I remain aware of the implications of supporting a large corporation like Amazon. This consciousness drives me to seek control over my digital content, challenging platform-imposed limitations. The choice between Kobo and Kindle goes beyond hardware; it's an exploration of ecosystems, content access, and personal values in technology. For me, the journey continues with both platforms. Being an avid reader, I relish the ability to switch between Kindle and Kobo, not just for the diverse reading experiences they offer but also for the enjoyment of engaging with their distinct technologies and communities. ### Further reading - [Kobo and Kindle solve different problems](https://rob-hoeijmakers.ghost.io/kobo-and-kindle-solve-different-problems/) ### How to get verified on LinkedIn URL: https://rob-hoeijmakers.ghost.io/how-to-get-verified-on-linkedin/ Last updated: 2026-05-13T15:21:43.000Z LinkedIn verification lets you confirm your identity on the platform using a government-issued ID — free, and done in a few minutes. This post covers the steps, what LinkedIn receives from the process, and a few things worth knowing if you're verifying with a Dutch or European passport. ![A verification shield on my LinkedIn profile](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/01/Verified-linkedin.jpg) A verification shield on my LinkedIn profile ## LinkedIn's Profile Verification Update: A Brief Overview LinkedIn has always been the go-to platform for professionals to connect and network. It's a place where your digital identity holds immense value. In April 2023, LinkedIn introduced a profile verification feature in partnership with identity platform CLEAR. This feature allows users to confirm their identity by providing their government-assigned ID. Once verified, a badge is added to your profile, giving you credibility within the platform. ## Expanding Access Initially, this feature was only available to users in the United States. However, LinkedIn has since expanded its reach to include members in Canada and Mexico. What's even more exciting is LinkedIn's commitment to making this verification option available to more users worldwide. ## The Role of Third-Party Partners LinkedIn has realized that to expand its verification offering, collaboration with third-party partners is essential. Microsoft Entra, Persona and Digilocker are among the partners joining forces with LinkedIn to facilitate verification in more regions. The challenge here lies in the need for these partners to collaborate with local entities capable of confirming ID documentation. It's worth noting that these partners are compensated for their efforts, which could impact the availability of the feature in certain regions. ## The Value of Verified Profiles Why should you consider verifying your LinkedIn profile? According to LinkedIn, the benefits are substantial. **Verified profiles receive 60% more profile views, 50% more comments and reactions on their posts, and 30% more messages.** This additional layer of credibility can significantly enhance your presence on the platform. The mechanics of verification are straightforward and what LinkedIn does with your verified identity is a more interesting question, and one [I explored separately](https://rob-hoeijmakers.ghost.io/linkedin-verification-when-identity-becomes-a-profile/). ## Integration with Job Postings LinkedIn has taken it a step further by integrating profile verification into selected job postings. This move provides additional confirmation and assurance within the app, reinforcing the importance of verified identities. ## How to Get Verified on LinkedIn Now that you're aware of the various avenues for verification, you might be wondering how to get started. LinkedIn has made the process relatively straightforward. To embark on your verification journey, follow these steps: 1. **Visit the LinkedIn Verification Page:** The first step is to visit the LinkedIn Verification page, which you can access by clicking on this link: [LinkedIn Verification Page](https://lnkd.in/g7qBpiMq). 2. **Choose Your Verification Method:** On the verification page, you'll have the option to choose the verification method that aligns with your situation. LinkedIn offers multiple methods to confirm your identity. 3. **Follow Provided Instructions:** Once you've selected your preferred verification method, carefully follow the provided instructions. Depending on the method you choose, this may involve providing specific documents, confirming personal details, or undergoing other necessary steps. 4. **Await Verification Approval:** After completing the verification process, you'll need to wait for LinkedIn to review and approve your request. This typically doesn't take long, and once your verification is approved, you'll proudly display the "Verifications" badge on your LinkedIn profile. This badge signifies your commitment to authenticity and professionalism. ![The verification processing steps with Persona: Scan ID, verify photo and done.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/01/Verification-steps-LinkedIn.jpg) The verification processing steps with Persona: Scan ID with NFC, verify photo and done. In a digital world filled with uncertainties, LinkedIn's verification process provides a layer of trust and confidence for both individuals and businesses. So, are you LinkedIn verified? If not, it's time to embark on the journey to let your connections and potential employers know that the real you is right here, ready to connect and engage in the professional world. 🛂 a Dutch reader pointed out that when verifying with Persona you need to be in the country of your nationality. 💡 ****Does LinkedIn Access My ID Details?** Yes, LinkedIn may access your ID details if you choose to undergo identity verification. During the process, you'll need to upload a valid government-issued ID. LinkedIn's verification partner, like Persona or CLEAR, will verify this information. They confirm your identity and add a verification badge to your profile. However, LinkedIn won't receive your biometric data, photos, passport numbers, or expiry/issue dates associated with your passport. The information shared with LinkedIn includes your name as it appears on your passport, the type of ID document used for verification (e.g., passport), and the issuer of that ID document. LinkedIn uses this data solely for security purposes and does not access biometric data or photos. ### Further reading - [LinkedIn Verification: When Identity Becomes a Profile](https://rob-hoeijmakers.ghost.io/linkedin-verification-when-identity-becomes-a-profile/) - [Account, profile and identity](https://rob-hoeijmakers.ghost.io/accounts-profiles-identity/) ### Why I use BrowserStack for testing and previewing URL: https://rob-hoeijmakers.ghost.io/why-browserstack/ Last updated: 2026-04-14T13:22:41.000Z As a [Web Strategist](https://rob-hoeijmakers.ghost.io/about-rob-hoeijmakers/), my journey through the digital landscape has taught me the importance of meticulous testing and previewing. Ensuring that websites and applications look and function flawlessly across various devices and browsers is a non-negotiable part of my workflow. That's why I consider BrowserStack to be an indispensable tool in my arsenal. In this article, I'll share my personal experience and insights into why BrowserStack has become a trusted companion for my web development endeavors. ## A World of Real Devices at Your Fingertips One of the most compelling reasons I turn to BrowserStack is the access it provides to a vast array of real mobile devices. When creating websites or apps, it's often impossible to physically possess every device that your audience might use. BrowserStack changes the game by granting you virtual access to a comprehensive library of real phones and tablets. Imagine being able to see how your web page or application truly looks and performs on an iPhone X, a Samsung Galaxy S21, or even a vintage Nokia 3310\. BrowserStack makes this possible, allowing you to capture authentic screenshots and evaluate real-world performance. The insights gained from testing on real devices are invaluable and can save you from unexpected user experience hiccups down the road. ![The device library of BrowerStack](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/01/2024-01-11_08-15-42.png) The device library of BrowerStack ## Realism Without the Complexity While the importance of real device testing cannot be overstated, I also appreciate BrowserStack's ability to simulate mobile devices. Simulations replicate the appearance and behavior of specific devices, ensuring consistency in your digital creations. They're like having a virtual gallery of devices at your disposal. Simulations are especially useful when you need to streamline your testing process or when access to physical devices is limited. With a few clicks, you can verify that your content looks and works as intended across a range of popular devices. It's a time-saving feature that maintains a high level of realism in your testing workflow. 💡 ****Real Phones vs. Emulations** Choose real phones for accurate performance and device-specific testing. Emulations offer cost-effective efficiency and broad compatibility checking. Optimize your testing strategy based on your specific project needs. ## Simple Checks to Complex Scenarios What I find particularly impressive about BrowserStack is its versatility. It caters to a wide range of testing scenarios, from simple checks to more complex, real-time debugging. Whether you're performing cross-browser testing, mobile responsiveness checks, or debugging issues in real time, BrowserStack adapts to your needs. It allows you to conduct thorough testing without the hassle of managing physical devices or setting up complex testing environments. With [BrowserStack](https://www.browserstack.com), you can elevate your development workflow, enhance collaboration among your team, and ultimately deliver polished digital creations to your audience. ### Creating Phone Images with Canva URL: https://rob-hoeijmakers.ghost.io/creating-phone-images-with-canva/ Last updated: 2026-04-25T13:15:58.000Z As a [Web Strategist](https://rob-hoeijmakers.ghost.io/about-rob-hoeijmakers/), I often find myself in the position of capturing mobile phone screenshots for my website. These screenshots provide a window into the digital world, allowing others to experience the content I analyze. In this article, I'll guide you through the process of creating captivating phone images using [Canva](https://rob-hoeijmakers.ghost.io/canva/), a versatile design tool that can turn your screenshots into polished mockups. Here is a video with the steps in Canva, it assumes you already have the screenshots from your phone. The detailed steps you find below this video. Using Canva to create a phone mockup with a screenshot. ## Section 1: Preparing Your Screenshots Before you dive into design, you'll need the right screenshots. Here's how to prepare them: ### Taking iPhone Screenshots: To take a screenshot on an iPhone, follow these steps: 1. Navigate to the screen or app you want to capture. 2. Press the side button and the volume up button simultaneously. 3. The screen will briefly flash, indicating that a screenshot has been taken. Now, let's move on to Canva to transform these screenshots into engaging phone images. ## Section 2: Getting Started with Canva Now that you have your screenshots ready, it's time to utilize Canva's powerful features to create stunning phone images. Here's a step-by-step guide on how to get started: ### **Open Canva and start a new design** Begin by opening Canva in your web browser. If you don't have an account, you can sign up for free or log in if you already have an account. Once you're logged in, click on the "Create a design" button on the Canva dashboard. You'll be presented with various design options. For the mockup I choose presentation because it has the proper [aspect ratio](https://rob-hoeijmakers.ghost.io/aspect-ratio/) for my blog. In the design options, you can also select the "Custom dimensions" option to set the canvas size according to your needs. You can choose a size that matches the aspect ratio of the screenshot or the device frame you want to use. ### **Upload Your Screenshot** Click on the "Uploads" tab on the left sidebar to upload the screenshot you want to showcase in your phone image. Simply drag and drop the screenshot file into Canva, and it will appear in your uploads folder. ### **Add Your Screenshot to the Design** Drag the uploaded screenshot from your uploads folder and drop it onto the canvas. You'll see your screenshot appear in the design. With your screenshot on the canvas, you can now edit it to enhance its appearance. Canva provides various tools for photo editing, including options for cropping, resizing, adjusting brightness and contrast, and applying filters. Use these tools to make your screenshot look its best. ### **Pick a Mockup** To create the illusion of a real phone, you can use Canva's mockup feature. Click on the "Elements" tab on the left sidebar and search for "mockup" in the search bar. Choose the mockup frame that matches the style and type of phone you want to showcase. Canva offers a variety of phone mockup options, including iPhones, Android devices, and more. Drag and drop the selected mockup frame onto the canvas. Resize and position it to fit your screenshot within the phone frame. Canva's mockup library includes different phone models and orientations. Select the one that best matches the appearance of the phone you want to showcase. You can customize the color and orientation of the mockup frame as well to align it with your design vision. ### **Fine-Tune and Export** Once you've positioned your screenshot within the phone frame and adjusted all the elements to your liking, take a final look at your design. Make any necessary adjustments. When you're satisfied with the result, click the "Download" button in the top right corner of the Canva interface to export your phone image. You can choose your preferred format (e.g., PNG or JPEG) and quality settings. That's it! You've successfully created a stunning phone image using Canva, combining your screenshot with a realistic phone frame to showcase your digital content. This visually appealing image can be used for various purposes, including presentations, marketing materials, or website visuals. ### Digital Twins: Simulation versus Emulation URL: https://rob-hoeijmakers.ghost.io/digital-twins-simulation-vs-emulation/ Last updated: 2026-04-25T13:16:05.000Z ### Understanding the Key Differences In the world of online professionals, the concept of digital twins is gaining momentum. Digital twins are virtual representations of physical objects, systems, or processes that can help streamline development, testing, and analysis. However, when it comes to creating these digital twins, a fundamental choice arises: **simulation or emulation**. In this article, I will break down the differences between simulation and emulation and explore which approach suits different scenarios. ## What Are Digital Twins? Before we dive into the world of simulation and emulation, let's clarify what digital twins are. In essence, digital twins are digital replicas of real-world objects or systems. These twins can range from a simple simulation to a precise emulation of the original. Digital twins are used across various industries, including urban planning, manufacturing, and mobile networks. ## Simulation: Acting as If **Simulation** involves creating a virtual representation of an object or system while simplifying or abstracting some aspects. The key here is that it acts "as if" it were the real thing, replicating behaviors and interactions without the need for an exact match in hardware or software. ### Use Cases for Simulation: 1. **Complex Systems Modeling:** Simulation-centric digital twins are ideal for modeling complex systems like cities, where replicating every detail precisely isn't necessary. These simulations help urban planners optimize infrastructure and transportation. 2. **Scenario Testing:** If you need to test different scenarios without replicating specific hardware, simulation is your go-to choice. For instance, emergency response simulations allow professionals to prepare for various disaster scenarios. 3. **Predictive Analysis:** Simulations are handy for predicting outcomes and analyzing trends. Supply chain simulations, for example, help professionals forecast inventory levels and delivery times. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/01/5b533b77272dea26527422eb15740f3e82bb645e-2.jpeg) **A cloud-based pedestrian wind comfort study.* ## Emulation: Acting the Same On the other side of the coin is **emulation**, which aims to create a digital twin that acts precisely the same as the original. It replicates both the external behavior and internal workings of the object or system, providing an exact duplicate for testing and development purposes. ### Use Cases for Emulation: 1. **Hardware Testing:** When precise replication of hardware components is critical, emulation shines. Mobile network emulation, for instance, involves replicating base stations and communication protocols for thorough network testing. 2. **Firmware and Software Development:** Emulation-centric digital twins are perfect for software and firmware development when testing on specific hardware configurations is necessary. Think of it as emulating mobile devices to test apps and firmware. 3. **Security Assessment:** For security testing, where pinpoint accuracy is essential, emulation is the preferred choice. It replicates both hardware and software components to identify vulnerabilities and threats accurately. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/01/https---images.ctfassets.net-q7ob9vms4z5k-ZqdsdsX4heZVFNe7VyRkk-b660e83b4e87d7519f9f8ac02007460b-digital-twin3-2.jpg) Vodafone created a digital twin of their network to test availability under chaining circumstances. New buildings, growing trees etc. 💡 Learn more about reinforcement learning and how AI is enhancing customer experience in a complex 5G world. Ericsson [emulated the mobile network](https://www.ericsson.com/49ced5/assets/local/reports-papers/mobility-report/documents/2021/ai%5Fenhancing-customer-experience.pdf) behaviour to perform a complex power change. ## Striking a Balance In practice, some digital twin applications blend elements of both simulation and emulation to find a balance between accuracy and flexibility. The choice depends on the twin's goals and the specific requirements of the system or object being represented. ### The AI Revolution in Medicine: GPT-4 and Beyond URL: https://rob-hoeijmakers.ghost.io/the-ai-revolution-in-medicine-gpt-4-and-beyond/ Last updated: 2026-05-10T08:54:14.000Z In "The AI Revolution in Medicine: GPT-4 and Beyond," authors Peter Lee, Carey Goldberg, and Isaac Kohane delve into the transformative role of Large Language Models (LLMs), especially GPT-4, in the medical field. ## The Enigma of GPT-4 The book intertwines the authors' personal experiences with an analytical perspective on GPT-4\. It brings to light the tool's authoritative yet occasionally inaccurate outputs, underscoring the necessity for careful verification. ## AI and Healthcare Dynamics A significant focus of the book is on the changing dynamics between doctors, patients, and AI. It explores the concept of AI-augmented patients and the potential of GPT-4 in advancing healthcare through its computational prowess. > Even if it does not act autonomously, GPT-4s potential for improving healthcare appears off the charts - for supplementing rather than replacing healthcare providers. ## GPT-4: The Ultimate Paperwork Shredder An insightful chapter demonstrates how GPT-4 can streamline medical administration, tackling paperwork for insurance and government, thereby potentially easing the bureaucratic load on healthcare professionals. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/12/IMG_8781.jpeg) The book uses examples of first shot interaction with ChatGPT. ## Smarter Science: Accelerating Biomedical Research The book features a critical chapter on 'Smarter Science', highlighting how AI, driven by empirical data, can accelerate basic biomedical science. This chapter presents a compelling argument for the potential of AI to hasten scientific discoveries and innovation in healthcare. ## Speculating the Future In its concluding chapter, the book speculates about the future interplay of AI and healthcare, blending practical insights with forecasts of what lies ahead in this exciting field. ## Personal Impressions My takeaway from this book is overwhelmingly positive. It offers profound insights and actionable guidance, mirroring the swift advancements in AI technology. Particularly recommended for healthcare professionals, this book serves as an eye-opener to the impactful role of AI in revolutionizing medicine. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/12/IMG_8782.jpeg) The book has a foreword by OpenAI CEO Sam Altman and a Bill Gates quote. The book is published by Pearson and you can [buy it from them](https://www.pearson.com/en-us/subject-catalog/p/the-ai-revolution-in-medicine-gpt-4-and-beyond/P200000011399/9780138279516) or your regular bookstore. [The AI Revolution in Medicine: GPT-4 and Beyond -- Pearson eText (OLP)![](https://www.pearson.com/pmc-react-browse/images/favicons/pearson-favicon-light.svg)Pearson Home![](https://www.pearson.com/pmc-react-browse/images/sprite-controls.svg#icon-expand-md)](https://www.pearson.com/en-us/subject-catalog/p/the-ai-revolution-in-medicine-gpt-4-and-beyond/P200000011399/9780138279516) ### Further reading - [The European Health Data Space (EHDS)](https://rob-hoeijmakers.ghost.io/the-european-health-data-space/) ### Google One subscription for Google Meet URL: https://rob-hoeijmakers.ghost.io/choosing-google-one-to-use-google-meet/ Last updated: 2026-05-10T08:54:14.000Z As a web strategist and freelance professional, finding the right tools to streamline my workflow is crucial. I've recently made some choices regarding online video conferencing and cloud storage solutions. In this article, I'll share my insights on why I opted for Google Meet over Zoom and [Google One](https://one.google.com) instead of [Google Workplace](https://workspace.google.com), hoping to aid fellow online professionals in making informed decisions. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/02/-Google-Meet-English-.jpg) Clean and simple Google Meet start screen. ## Google Meet: A Simplified Alternative to Zoom The world of online video conferencing has been dominated by platforms like Zoom. However, its increasing complexity and cost prompted me to seek an alternative. Enter **Google Meet**. Here's why it stood out: 1. **Simplicity for Users and Guests**: Unlike Zoom, Google Meet offers a user-friendly interface. This simplicity extends to my clients and collaborators, making it easier for everyone to connect, right through the browser. 2. **Progressive Web App (PWA) Advantage**: [Google Meet's PWA](https://rob-hoeijmakers.ghost.io/progressive-web-apps-explained-using-twitter-lite/) can be pinned to the Mac dock, offering quick access and seamless integration. 3. **Integration with Google Calendar**: The seamless integration with Google Calendar is handy, simplifying scheduling and management of meetings (and works with [notetaker Fireflies.ai](https://rob-hoeijmakers.ghost.io/review-of-transcription-software-tactiq-vs-firefliesai/)). ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/12/Google-Meet-Mac.jpg) Google Meet as app on the Mac ## Choosing Google One Over Google Workplace While Google Workplace is a robust solution, it’s designed with teams and larger enterprises in mind. I personally never really liked it, it is oversized for the individual professional by design. Here’s why Google One is a better fit for my individual freelancing needs: 1. **Simplicity and Individual Focus**: Google One, unlike Google Workplace, is tailor-made for individual use. Its intuitive design suits freelancers who don’t need complex team-oriented features. 2. **Avoiding Over-Integration**: Google Workplace tends to delve deep into hosting and email services. My preference is to use Google primarily for search, SEO, calendars, online meetings, and scheduling. 3. **Value-Added Features**: The transition to Google One offered me more than just storage. The addition of features like [VPN](https://rob-hoeijmakers.ghost.io/virtual-private-network-vpn/), high-quality video meetings (1080p), and [integrated booking pages](https://rob-hoeijmakers.ghost.io/from-calendly-to-scheduling-with-google-calendar/) streamlined my workflow considerably. In the ever-evolving digital landscape, the right tools can make a significant difference in productivity and efficiency. For freelancers and individual professionals, Google Meet and Google One offer a blend of simplicity, functionality, and integration that can simplify the digital life. As someone who continually refines his online strategy, I find these tools to be invaluable in my day-to-day operations. I hope my experience helps you in making the best choice for your professional needs. If you like this content, please consider to [sign up for free](#/portal/signup) and you will receive new articles in your inbox. 💡 ****Difference between Google One and Google Workplace** Google One is tailored for individual users and families, offering expanded cloud storage across Google Drive, Gmail, and Google Photos with plans that range from 100 GB to 2 TB. It's ideal for personal storage needs but the benefits also can be shared with 4 family members.. In contrast, Google Workspace (G Suite), designed for businesses, encompasses a comprehensive suite of productivity and collaboration tools including business email, advanced Google Apps, and enhanced security features. This makes Google Workspace a robust solution for organizational needs. When ****comparing Google One's 2TB plan to Google Workspace Business Starter**, consider the unique benefit of sharing premium features with family members. With Google One's 2TB plan at $9.99 per month, not only can you share storage space across six family members, but you also gain access to premium features of Google Meet and [advanced scheduling capabilities](https://www.hoeijmakers.net/from-calendly-to-scheduling-with-google-calendar/). This sharing feature makes Google One an increasingly cost-effective choice for families or groups, providing high-end tools and storage at a fraction of the cost per person compared to the Google Workspace Business Starter plan. --- ### More on Google --- ### Further reading - [Setting up a Google Account without Gmail](https://rob-hoeijmakers.ghost.io/google-account-without-gmail/) - [Google’s AI for Individuals and SMBs: Clarity, Costs, and Strategic Choices](https://rob-hoeijmakers.ghost.io/googles-ai-for-individual-smb/) - [Google One vs Google Workspace: Where do you get real AI and Gemini Advanced?](https://rob-hoeijmakers.ghost.io/google-one-vs-google-workspace-for-gemini/) ### From Calendly to scheduling with Google Calendar URL: https://rob-hoeijmakers.ghost.io/from-calendly-to-scheduling-with-google-calendar/ Last updated: 2026-05-10T08:54:14.000Z As a freelance web strategist, my journey to optimize efficiency and workflow led me [to Google One, primarily for its premium Google Meet services](https://rob-hoeijmakers.ghost.io/choosing-google-one-to-use-google-meet/). Yet, an unexpected and valuable discovery in this package was Google Calendar, particularly its booking page feature. In this article, I'll explore how the Google Calendar scheduling feature, initially a complementary addition to my Google One subscription, has become an indispensable scheduling tool for professionals like me. ## Discovering Google Calendar's Potential The inclusion of Google Calendar with my Google One subscription was a an unexpected benefit. Here's why it quickly became a core part of my professional toolkit: 1. **Integrated Booking Pages**: The ability to create booking pages directly linked to my calendar has been a revelation. This feature effortlessly replaced other scheduling tools, consolidating my workflow. 2. **Seamless Integration with Premium Google Meet**: Opting for Google One for its superior Google Meet capabilities, I found that Google Calendar's integration with these services enriched my meeting and scheduling experience. 3. **Streamlined Scheduling in One Place**: The transition to using Google Calendar was seamless, aligning perfectly with my strategy to minimize the number of different services I use, while enhancing my overall productivity. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/12/Google-Calendar-scheduling.jpg) The flow for my visitors. From the contact page to booking a meeting. ## The Practical Fit While exploring scheduling tools, I chose Google Calendar for its practicality and integration with my existing Google services, despite some limitations. Here's why it works for me: 1. **Basic Yet Efficient Scheduling**: Google Calendar's scheduling features are basic, lacking advanced customization and branding options. However, this simplicity aligns well with my business needs, providing just enough functionality without overcomplicating the process. 2. **Integration with Google Ecosystem**: Its integration with Google Meet and Google One offers a cohesive experience that is more valuable to me than advanced features offered by other scheduling tools. 3. **Fit for Purpose**: Although other services might offer [more elaborate scheduling options](https://rob-hoeijmakers.ghost.io/digital-diary-and-calendars/), the combination of Google Calendar with my current use of Google Meet and other Google services provides a streamlined and efficient setup that suits my professional context. 4. **Aligning with Minimalist Approach**: My preference for a minimalist digital toolkit means prioritizing services that cover multiple needs in a simple, effective manner. Google Calendar fits this criterion, reducing the need for multiple specialized tools. So while Google Calendar may not be the most feature-rich scheduling tool, its simplicity, integration with other Google services, and alignment with my business strategy make it the right choice for my current setup. ## Going from Calendly to Google Calendar When I first subscribed to Google One, I was primarily using Calendly for my scheduling needs. However, I soon realized that Google Calendar's booking page feature offered similar functionalities, prompting me to make a switch. Here’s how this migration significantly streamlined my workflow: 1. **Effortless Setup**: Transitioning to Google Calendar's booking feature was surprisingly straightforward. I simply replicated the settings I had in Calendly, which only took a few minutes. 2. **Simple URL Replacement**: The only significant change was updating the URLs from Calendly to Google Calendar in [my communications and website](https://rob-hoeijmakers.ghost.io/about/). This was a minor adjustment compared to the convenience gained. 3. **Functionality for Individual Use**: While Calendly excels in complex team-based and outbound scheduling scenarios, for my solo professional use, Google Calendar provided all the necessary functionality without the added complexity. 4. **Minimizing Service Redundancy**: As a freelance professional, my aim is to streamline my digital toolset. Migrating to Google Calendar allowed me to reduce the number of separate services I use, aligning with my minimalist approach to digital tools. This transition to Google Calendar not only simplified my scheduling process but also integrated seamlessly with my existing use of Google One and Google Meet, creating a more cohesive and efficient workflow. ### Further reading - [Setting up a Google Account without Gmail](https://rob-hoeijmakers.ghost.io/google-account-without-gmail/) - [Digital scheduling calendars compared and explained](https://rob-hoeijmakers.ghost.io/digital-diary-and-calendars/) - [When AI Makes the Calendar Click](https://rob-hoeijmakers.ghost.io/gemini-google-calendar/) ### Choosing the right "Face" for your chatbot URL: https://rob-hoeijmakers.ghost.io/choosing-the-right-face-for-your-chatbot/ Last updated: 2026-04-25T13:48:17.000Z Chatbots have become a crucial tool for businesses and organisations to engage with their audiences. A critical aspect of deploying a chatbot is deciding how it will be visually represented. This decision can significantly impact user experience and brand perception. Let’s assume you have a chatbot that works with texts, what face should you give it? There is a not an authoritative answer but based on experience I will give guidelines that can help you decide. Let's explore the three most common options: using a logo, an avatar, or a photo, and understand their implications for your chatbot strategy. ## 1\. The Professional Appeal of Logos Using a logo as the face of your chatbot is a classic choice that exudes professionalism and consistency. It's ideal for businesses that prioritize brand recognition and want to maintain a formal tone in their customer interactions. **Best for**: Corporate sectors, professional services, and B2B interactions. **Pros**: Enhances brand recognition, instills trust, and maintains a professional image. **Cons**: May lack the personal touch that encourages user engagement. ![OpenAI is using a logo as a face to the user.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/11/Chatbot-company-logo.jpg) OpenAI is using a logo as a face to the user. ## 2\. Avatars: Personalization at Its Best Avatars, whether in the form of characters or stylized symbols, offer a unique way to personalize the chatbot experience. They can reflect your brand's personality, making interactions more engaging and relatable. **Best for**: Creative industries, consumer-focused services, and younger demographics. **Pros**: Creates an engaging, personable experience; can embody brand personality. **Cons**: Requires careful design to align with brand identity and audience expectations. ![Examples of avatars and company mascottes as the face for the chatbot.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/11/Chat-avatar-3.jpg) Examples of avatars and company mascottes as the face for the chatbot. 👻 In the realm of chatbots, company mascots serve as unique avatars that personify brands. These mascots go beyond standard avatars, embodying a brand's essence and injecting chatbot interactions with character and charm. They transform the chatbot experience from purely informative to engaging and memorable. ## 3\. Photos: The Human Touch Using a photo, particularly of a person, humanizes the chatbot experience. This approach can be especially effective if your brand values personal connection or if the chatbot handles complex, empathetic interactions. **Best for**: Healthcare, hospitality, and customer support where empathy and personal connection are key. **Pros**: Offers a humanized interaction, builds empathy, and enhances relatability. **Cons**: Can set unrealistic expectations about the human involvement in the chat service. ![Chatbots, or live chat, with photo's of humans.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/11/Chatbot-photos.jpg) Chatbots, or live chat, with photo's of humans. As you ponder over the right visual choice for your chatbot – be it a logo, avatar, or photo – remember that this decision is pivotal in how your audience interacts with and perceives your brand. It's essential to deeply understand your audience, align the choice with your brand’s identity, and consider the nature of interactions your chatbot will handle. Don’t hesitate to experiment; create mockups, engage in A/B testing, and gather feedback. This is not just about choosing a visual; it's about crafting an experience that resonates with your users. And if you find yourself at a crossroads, seeking professional advice can provide clarity and direction. Ultimately, the choice you make should not only look good but should also be a strategic fit for your brand, enhancing the chatbot’s role as an engaging and effective communication tool. Remember, your chatbot’s face is more than just an image; it's the first impression, the ongoing interaction, and potentially, a lasting memory in the minds of your users. *This article was originally* [*published in Dutch on Chat voor Bedrijven*](https://www.chatvoorbedrijven.nl/het-juiste-gezicht-kiezen-voor-je-chatbot/)*.* ### Review of transcription software: Tactiq versus Fireflies URL: https://rob-hoeijmakers.ghost.io/review-of-transcription-software-tactiq-vs-firefliesai/ Last updated: 2026-05-10T08:54:15.000Z As a [web strategist ](https://rob-hoeijmakers.ghost.io/about/)deeply involved in the world of digital communication, the efficiency and accuracy of transcription software in video meetings is crucial. My journey began with [Tactiq](https://tactiq.com), I wrote [a review when it just hit the market](https://rob-hoeijmakers.ghost.io/streamline-your-virtual-meetings-with-ai-note-taking/). However, intrigued by the growing popularity of [Fireflies.ai](https://app.fireflies.ai/login?referralCode=01K45WGDZV18D7M546M82NM6XX), a tool that came highly recommended by a friend, I embarked on a detailed comparison to see which better suits my professional needs. ## Initial Impressions Tactiq, which relies on browser plugins (mostly Chrome), was my go-to choice for some time. However, as Chrome isn't my primary browser, this posed a minor inconvenience. Fireflies.ai presented a different approach - a bot joining meetings directly, which worked seamlessly across [Google Meet](https://rob-hoeijmakers.ghost.io/choosing-google-one-to-use-google-meet/), MS Teams, and Zoom. ## Evaluating Key Aspects 1. **Quality of Transcription (Including Dutch)**: After conducting several meetings with both Tactiq and Fireflies.ai active, the verdict was clear. Fireflies.ai provided more accurate transcriptions, especially in non-English languages like Dutch. 2. **Ease of Use**: Both platforms offer user-friendly interfaces. However, Fireflies.ai's bot integration makes it slightly more convenient, bypassing the need for specific browser dependencies. 3. **Integration with Other Apps**: While I haven't fully explored this aspect, both platforms offer integrations with apps like [HubSpot CRM](https://rob-hoeijmakers.ghost.io/customer-relationship-management-crm/). It's reassuring to know these options are available. 4. **Meeting Summaries and Insights**: Both Tactiq and Fireflies.ai provide meeting summaries and insights, albeit limited in their free versions. I found taking these transcripts to ChatGPT for further analysis more beneficial. 5. **Mobile App Availability**: Fireflies.ai edges out with its iOS app, allowing users to review and listen to meetings and their transcriptions on the go. 6. **Costs**: Both platforms are competitively priced, offering limited free accounts and a standard tier at around $10 per month when billed yearly. [![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/11/fireflies.jpg)](https://app.fireflies.ai/login?referralCode=3a4WliEYiv) The fireflies.ai web interface ## Personal preference Despite both being competent in their own rights, **my preference leans towards Fireflies.ai**. The bot’s ability to join meetings across various platforms without the need for specific browser plugins is a significant advantage. Moreover, the superior quality of its transcription, particularly in Dutch, aligns better with my professional requirements. However, it's important to note the potential issue of vendor lock-in and privacy concerns with such solutions. While these tools offer significant benefits to a freelancing web strategist like myself, they might not be ideal for everyone, especially in contexts where privacy and security are paramount. [Fireflies.AI: Meeting Transcriptions & AutomationFireflies.ai changed how I handle meetings, capturing every conversation with ease. Its AI-driven transcription is powerful, but some features need fine-tuning.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-31.png)Rob Hoeijmakers (robhoeij)Rob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_1385.jpeg)](https://rob-hoeijmakers.ghost.io/fireflies-ai-transcriptions/) ## Closing thoughts The evolution of transcription software like [Tactiq](https://tactiq.io) and [Fireflies.ai](https://app.fireflies.ai/login?referralCode=fX9OjJzFbU) is a testament to the advancements in AI and its integration into our professional lives. As technology continues to evolve, it's essential to stay informed and choose tools that best align with our individual needs and workflows. 💡 Fireflies offers services through ****AskFred**. I was not able to test the real time services during the meeting but it sounds like something very useful. This is what he does: AskFred is available for questions during the meeting. It allows users to ask questions about what happened during the meeting and receive answers in a conversational style. AskFred is designed to provide ChatGPT-like conversations during meetings, enabling users to ask about meeting content, action items, engineering tasks, and other relevant topics. Additionally, AskFred can help with follow-ups by generating emails, tweets, or blog post drafts based on the meeting transcript. It functions as an AI assistant that can answer questions related to meetings, understand decisions made, and clarify meeting content in real-time --- In addition to the primary vendors we've discussed, I received valuable suggestions from my LinkedIn network. Here are three noteworthy options to consider for meeting transcription and content enhancement: 1. **Sembly.ai**: Much like Fireflies.ai, [Sembly.ai ](https://www.sembly.ai)harnesses the power of a bot to streamline meetings and transcription processes. It's an intriguing choice for those seeking automation and accuracy in their meeting documentation. 2. **Bluedothq**: For those who prefer the convenience of Chrome extensions, [Bluedothq](https://www.bluedothq.com) offers a solution similar to Tactiq. Its integration with the Chrome browser provides users with a seamless transcription experience, making it an accessible choice. 3. **Grain.com**: Going beyond mere transcription, [Grain.com](https://grain.com) places a strong emphasis on elevating the quality of meetings. With a focus on coaching teams and closing deals, it promises to enhance not only the documentation but also the overall productivity and effectiveness of your meetings. These additional options expand the spectrum of meeting transcription solutions, catering to various preferences and priorities. It's important to thoroughly evaluate each one to determine the best fit for your specific needs. Thanks to [Andrii Fediv](https://www.linkedin.com/in/ACoAAALElHIBOP6BKnI6y8ZSuq-HLPtuwlsvd0E?lipi=urn%3Ali%3Apage%3Ad%5Fflagship3%5Fdetail%5Fbase%3BJHZND%2F0jREacGijh4%2Bnh1Q%3D%3D), [Alexander Radchykov](https://www.linkedin.com/in/ACoAABLneBABSRfNQF3tsxNSvMdBiIazR7kqwr0?lipi=urn%3Ali%3Apage%3Ad%5Fflagship3%5Fdetail%5Fbase%3BJHZND%2F0jREacGijh4%2Bnh1Q%3D%3D) and [Matthieu Hattab](https://www.linkedin.com/in/ACoAAAAKBAoBibxoqanzM1ZH%5FefH5ayqfGaLJVo?lipi=urn%3Ali%3Apage%3Ad%5Fflagship3%5Fdetail%5Fbase%3BJHZND%2F0jREacGijh4%2Bnh1Q%3D%3D). ### Further reading - [Fireflies.ai - #1 AI Teammate](https://app.fireflies.ai/login?referralCode=01K45WGDZV18D7M546M82NM6XX) ### European Digital Identity (EDI) URL: https://rob-hoeijmakers.ghost.io/edi-digital-identities/ Last updated: 2026-05-10T08:54:15.000Z In the digital age, where our online and offline worlds increasingly intertwine, the concept of digital identity has never been more pivotal. [Identity Wallet: A Tool for Empowerment or a Mechanism of Control?Freedom or surveillance? Digital IDs offer convenience but come with risks. Are they tools of empowerment or state control over your personal data?![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-51.png)Rob Hoeijmakers (robhoeij)Rob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_8641-5.jpeg)](https://rob-hoeijmakers.ghost.io/digital-identity-id-empowerment-or-state-control/) Updated article on this topic (2025) The European Union's initiative on [European Digital Identity (EDI) ](https://commission.europa.eu/strategy-and-policy/priorities-2019-2024/europe-fit-digital-age/european-digital-identity%5Fen)marks a significant stride in this realm, aiming to create a unified, secure digital space for citizens and businesses across Europe. As a digital strategist, I find myself at the intersection of this transformative wave, where the potential of digital identities and wallets unfolds, promising to redefine how we interact, transact, and exist online. This initiative isn't just about facilitating smoother transactions or simplifying logins; it's a comprehensive approach to balancing security, privacy, and innovation. The EDI's vision of a digital wallet for every EU citizen and resident is ambitious, aiming to empower users with control over their personal data while ensuring seamless access to a myriad of services. However, as we embark on this journey, it's crucial to navigate the complexities and challenges that come with such digital advancements. The balance between safeguarding individual freedoms and offering practical, efficient digital solutions is delicate and demands careful consideration. 💡 Digital Identities are a bit abstract when you are not familiar with them. I looked at existing ones and especially the Estionian ****e-Residency** is insightful. A digital identity issued by the Estonian government to anyone in the world. It allows e-Residents to start and manage an EU-based company online, access banking services, and utilize digital signing. Other international identies are Aadhaar - India, SingPass - Singapore, Digital ID and Authentication Council of Canada (DIACC) - Canada, Digital Identity Framework - Australia, National Digital Identity - South Korea and to an extend REAL ID - United States. ## Understanding Digital Identities and Wallets In the ever-evolving digital landscape, the concept of digital identities and wallets has emerged as a cornerstone of online interaction and transaction. **Digital identities** represent the virtual persona of individuals or entities, encompassing a range of personal, professional, and behavioral data. These identities are increasingly becoming the key to accessing a myriad of services online, from social media to banking. **Digital wallets**, on the other hand, are the conduits through which these identities operate, storing and managing digital credentials, such as passwords, payment information, and personal identification documents. They are not just about facilitating transactions but also about ensuring the secure and efficient management of our digital selves. 💡 For this article I researched multiple wallets and experimented with them to get a good sense and grip. I used [****ProofMe**](https://www.proofme.id), a service that works from physical documents like passport and ID. It can be used for access to events and age related licensing in the real world. For instance shopping of alcohol. I tried [****Yivi**](https://www.yivi.app/en), a digital wallet with privacy in mind. I used it to get access to my medical dossiers . And then there is the Dutch [****DigiD**](https://www.digid.nl/en) system that gives access to online services at local and national level. So often the use cases and legal framework around identities and how they are stored/maintained differ. The EDI is a framework and member states will offer local wallets, like the [Dutch ID-Wallet](https://www.digitaleoverheid.nl/overzicht-van-alle-onderwerpen/identiteit/id-wallet/). [Understanding Digital Identity in the EU: eID, eIDAS, and the EUDI WalletThis article serves as a reference point for further discussions on \*\*digital identity, authentication, and legitimisation\*\* in the EU. Future articles will explore how businesses, organisations, and individuals can leverage these systems for more secure and streamlined online interactions.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-19.png)Rob Hoeijmakers (robhoeij)Rob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_0771-1.jpeg)](https://rob-hoeijmakers.ghost.io/eu-eid-eidas-eudi-wallet/) Updated overview (2025) of EU and Identity ## The benefits of Digital Identities As a web strategist, I see immense potential in the proper implementation of digital identities and wallets, especially when underpinned by robust security and privacy measures. The European Union's initiative on European Digital Identity (EDI) is a step towards a unified, secure digital space where citizens and businesses can seamlessly interact across borders. The practical benefits are manifold. For **businesses**, it simplifies customer onboarding, enhances the personalisation of services, and streamlines transactions. For **users**, it offers a more integrated and user-friendly digital experience, reducing the need for multiple passwords and accounts. Moreover, in a world where data breaches are commonplace, a well-structured digital identity system can offer enhanced security protocols and reduce the risk of identity theft. ![Yivi - Digital Identity with an emphasis on privacy. Authenticated through government services but will not share your resident ID.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/11/1.jpg) [Yivi](https://www.yivi.app/en) \- Digital Identity with an emphasis on privacy. Authenticated through government services but will not share your resident ID. ## Political implications Politically, the EDI represents a significant step. It's crucial to balance the practicality of digital identities with the freedom and rights of citizens. The EU's stance, primarily focused on privacy, reflects a commitment to protecting individual rights in the digital realm. However, there's an underlying narrative here: is the EU's approach a defensive strategy, stemming from a perceived lack of competitiveness in the global digital economy? The EU's focus on privacy is commendable, but it sometimes appears as a protective shield, possibly hinting at an underlying struggle to keep pace with other technologically advanced regions. While prioritizing privacy and security is vital, it's equally important for the EU to foster an environment that encourages innovation and competitiveness in the digital sector. How to add your Maryland Mobile ID to Apple Wallet ## Addressing the fear of Identity Theft The fear of identity theft in a digital world is not unfounded. A centralized digital ID system, governed by the EU, with its comprehensive laws, policing, and budget, could offer a more secure framework than disparate, bottom-up initiatives. Centralization, in theory, should provide stronger, more uniform security protocols and a more streamlined approach to identity management. However, it also raises concerns about the concentration of power and the potential for misuse. ## Concluding thoughts As we navigate this new era of digital identities, the challenge lies in striking a balance between security, privacy, and freedom. The EU's EDI initiative is a bold step towards a more integrated and secure digital Europe, but it must be implemented with a keen eye on these balances. As a web strategist, I see the EDI as a significant opportunity for businesses and individuals alike, but it's an opportunity that comes with great responsibility – to ensure that in our quest for digital efficiency, we do not compromise the very values that make our digital world worth navigating. ![ProofMe - Identity based on the paper versions of your official documents.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/11/2.jpg) ProofMe - Identity based on the paper versions of your official documents. --- ### Further reading - [eID](https://ec.europa.eu/digital-building-blocks/sites/display/DIGITAL/eID) ### Whisper by OpenAI: Unlocking Speech Recognition for Transcription and Beyond URL: https://rob-hoeijmakers.ghost.io/openais-whisper-speech-to-text-technology/ Last updated: 2026-05-26T19:36:51.000Z In the rapidly evolving world of digital communication, OpenAI's Whisper emerges not just as a tool, but as a revolution. As a web strategist and chat technology expert, I've been closely following Whisper's journey. It's more than just another open-source project; it's a beacon of innovation in speech-to-text technology. ## The Open Source revolution The magic of [Whisper](https://openai.com/research/whisper) begins with its open-source nature. This approach isn't just about sharing code; it's about inviting the world to improve, adapt, and innovate. By making Whisper open-source, OpenAI has not only democratized speech-to-text technology but also accelerated its evolution. The result? A more robust, efficient, and cost-effective solution for everyone. [Introducing WhisperWe’ve trained and are open-sourcing a neural net called Whisper that approaches human level robustness and accuracy on English speech recognition.![](https://openai.com/favicon.ico)![](https://images.openai.com/blob/7b33cec2-6549-4c33-943d-84372f9bee15/whisper-og.jpg?trim=0%2C32%2C0%2C864&width=1000&quality=80)](https://openai.com/research/whisper) ## Whisper's versatility What truly sets Whisper apart is its versatility. It's not confined to high-end servers or specific platforms. Whether it's running on a personal Mac, a Windows PC, or cloud platforms like Microsoft Azure, Whisper adapts seamlessly. This flexibility is a game-changer for businesses, big and small, offering them a powerful tool without the hefty price tag. ![Whisper Transcription for Mac](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/11/Whisper-transaction.jpg) [Whisper Transcription](https://apps.apple.com/us/app/whisper-transcription/id1668083311?mt=12) for Mac ## A deep dive into Whisper's capabilities Trained on an astonishing 680,000 hours of multilingual data, Whisper isn't just another speech recognition tool. It's a polyglot powerhouse. From deciphering various accents to cutting through background noise, Whisper handles it all with remarkable finesse. And it's not just about understanding different languages; it's about bridging communication gaps across them. ## Beyond theory: Whisper in action My experiments with Whisper have been nothing short of fascinating. I've been testing it with video call transcriptions, comparing its performance with platforms like [Fireflies.ai](https://rob-hoeijmakers.ghost.io/fireflies-ai-transcriptions/) and [Tactiq](https://rob-hoeijmakers.ghost.io/streamline-your-virtual-meetings-with-ai-note-taking/). These tests aren't just about assessing accuracy; they're about understanding how such technology can be woven into the fabric of business communication. Imagine integrating Whisper's transcriptions with large language models like [ChatGPT](https://rob-hoeijmakers.ghost.io/try-out-chatgpt-yourself/). The possibilities are endless – from generating instant meeting summaries to offering real-time translation services. This isn't just about making life easier; it's about redefining how businesses interact and operate. In my personal journey with Whisper, one of the most intriguing aspects has been its integration with ChatGPT. Each time I [converse with ChatGPT using spoken language](https://rob-hoeijmakers.ghost.io/this-week-chatgpt-learned-to-draw-see-and-talk/), Whisper silently plays a crucial role. It's the bridge between my spoken words and ChatGPT's understanding. This seamless interaction is fascinating – Whisper accurately picks up my speech, converting it into text that feeds directly into ChatGPT. This not only showcases Whisper's precision but also its potential to enhance and simplify how we interact with advanced AI systems. It's a practical demonstration of how these technologies can work in tandem to create a more intuitive and natural user experience. ![After activation in the settings you will find an icon to start voice conversations](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/11/ChatGPT-talks.jpg) After activation in the settings you will find an icon to start voice conversations ## Real change in the market Reflecting on the impact of Whisper, it strikes me that its significance goes beyond its technical prowess. It's the combination of its advanced features and its affordability, thanks to being open-source, that really stands out. This blend is what I believe will drive real change in the market. It's not just about offering a sophisticated tool; it's about making such technology accessible to a wider audience. This, in my view, is where Whisper could truly make a difference, transforming how we approach communication and efficiency in business. Stay tuned, as I delve deeper into the practical applications of Whisper, especially in video call transcriptions. The future of digital communication is here, and it's whispering a tale of endless possibilities. 💬 ****A Sidenote on Whisper's Capabilities** While exploring the depths of Whisper's functionality, it's important to note a key distinction: Whisper excels in speech-to-text, but it does not venture into the realm of generating speech. That's a different arena, handled by another unnamed module within OpenAI's suite of tools. For this speech generation aspect, OpenAI has taken a unique approach by employing five professional voice actors. The resulting voices, each with their own distinct character, are named Juniper, Sky, Ember, Breeze, and Cove. This diversification in voice technology complements Whisper's capabilities, together painting a comprehensive picture of OpenAI's advancements in auditory AI. --- ### Further reading - [Turning Radio Mentions into Digital Text with Whisper: My Friday AI Experiment](https://rob-hoeijmakers.ghost.io/whisper-on-my-mac/) ### A year of AI transformation (2023) URL: https://rob-hoeijmakers.ghost.io/a-year-of-ai-transformation/ Last updated: 2026-05-10T08:54:15.000Z It's been an extraordinary year since the introduction of various AI technologies that have profoundly altered the way we interact with the digital world. I've experienced firsthand the transformative power of AI. In this blog post, I'll share my experiences and impressions of how AI, across various applications, has impacted my work and daily life over the past year. ## A year with ChatGPT In the past year, the introduction of [ChatGPT](https://rob-hoeijmakers.ghost.io/try-out-chatgpt-yourself/) into my toolkit has ushered in a revolution in how I approach creativity and receive support across various aspects of my work as a web strategist and content writer. ChatGPT has emerged as a dynamic and indispensable partner, facilitating tasks that range from brainstorming ideas to enhancing my technical skills. Here's a closer look at its role in my professional journey: ### Idea generation and brainstorming One of the standout features of ChatGPT is its ability to assist in idea generation. Whether I'm seeking inspiration for blog topics, marketing campaigns, or creative projects, ChatGPT serves as a virtual brainstorming companion. By providing prompts and engaging in dialogue, it helps me explore new concepts, refine existing ideas, and uncover fresh perspectives. ### Information sorting and organization The volume of information available in today's digital landscape can be overwhelming. ChatGPT plays a pivotal role in [organizing and sorting through this vast sea of data](https://rob-hoeijmakers.ghost.io/chatgpts-code-interpreter-coding-without-code/). It helps me structure my research, categorize information, and distill key insights from diverse sources, enabling more efficient decision-making and content creation. ### Technical support and skill expansion AI-powered technical support is another area where ChatGPT shines. It assists me in various technical aspects of my work, from writing Python code for automation tasks to optimizing SEO strategies for improved online visibility. This technical guidance not only saves time but also empowers me to take on more complex projects and expand my skill set. ### Content generation and revision As a content writer, the quality of my written work is crucial. [ChatGPT](https://rob-hoeijmakers.ghost.io/try-out-chatgpt-yourself/) assists in generating high-quality content by suggesting sentence structures, providing synonyms, and offering alternative phrasing. It serves as a writing companion, aiding in content refinement and ensuring that my written pieces resonate with the intended audience. ### Enhancing SEO strategies Effective [SEO strategies ](https://rob-hoeijmakers.ghost.io/my-basis-for-seo/)are essential for online visibility and reach. ChatGPT's insights into SEO trends, keyword optimization, and content structuring contribute to the enhancement of my SEO strategies. It assists in identifying opportunities for improvement and ensuring that my online content remains competitive in search engine rankings. ### Versatile problem solving Beyond its specific applications, ChatGPT's versatility shines in problem-solving across a spectrum of subjects. Whether I'm troubleshooting technical issues, exploring new marketing tactics, or unraveling complex concepts, ChatGPT's ability to provide clear and concise explanations aids in informed decision-making and problem resolution. - [This week ChatGPT learned to draw, see and talk](https://rob-hoeijmakers.ghost.io/this-week-chatgpt-learned-to-draw-see-and-talk/) - [Increased productivity with AI in chat services](https://rob-hoeijmakers.ghost.io/increased-productivity-and-better-performance-with-ai-in-chat-services/) - [ChatGPT and the art of asking questions](https://rob-hoeijmakers.ghost.io/the-art-of-asking-questions/) - [Try out ChatGPT yourself](https://rob-hoeijmakers.ghost.io/try-out-chatgpt-yourself/) 💡 ****AI: A Diverse Category of Capabilities** Artificial Intelligence (AI) is a diverse category of technologies. From Natural Language Processing (NLP) to computer vision and more, AI capabilities can be combined and stacked to create innovative solutions. For example, ChatGPT uses NLP for text understanding, but it can also listen, speak, and generate images. Explore the versatile world of AI. \- ****Natural Language Processing (NLP)**: Understanding and generating human language. \- ****Computer Vision**: Enabling machines to interpret and analyze visual information. \- ****Generative Models**: Creating content, images, and text using AI-generated data. \- ****Machine Learning**: Algorithms that learn patterns and make predictions. \- ****Robotics**: Combining hardware and AI for physical tasks. \- ****Speech Recognition**: Transforming spoken language into text or commands. \- ****Reinforcement Learning**: Teaching AI through rewards and punishments. \- ****Expert Systems**: Mimicking human expertise for decision-making. \- ****Neural Networks**: Model inspired by the human brain for complex tasks ## **Photo management and creativity** One of the earliest and most impactful changes has come in the realm of photo management and content creation. AI's ability [to search through my photos](https://rob-hoeijmakers.ghost.io/iphone-live-text/), enhance their quality, propose captions, and generate summaries has revolutionized my content creation process. Whether I'm crafting blog posts, social media updates, or marketing materials, AI's assistance has become indispensable, boosting my creativity and productivity. ## Efficient meetings and communication AI-based transcription and summarization tools have revolutionized the way I handle meetings. The ability to automatically transcribe and summarize discussions has not only saved time but has also improved the clarity and accessibility of meeting records. This has made collaboration more efficient and effective. Now I use [Tactiq](https://rob-hoeijmakers.ghost.io/streamline-your-virtual-meetings-with-ai-note-taking/) but I am also evaluating [Fireflies](https://app.fireflies.ai/login?referralCode=fX9OjJzFbU) and [Otter.ai](https://otter.ai). [Review of transcription software: Tactiq vs. Fireflies.aiJoin me on a journey as I test and contrast Tactiq and Fireflies.ai, seeking the best transcription tool for video meetings in a web strategist’s life.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/size/w256h256/2024/05/Logo-Rob-Hoeijmakers-Square-Dark-Glow-3-1.png)Web StrategiesRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/size/w1200/2023/11/IMG_8046-1.jpeg)](https://rob-hoeijmakers.ghost.io/review-of-transcription-software-tactiq-vs-firefliesai/) ## **Multilingual and editorial texts** In a globalized world, effective communication in multiple languages is crucial. AI-powered translation tools have been a game-changer, enabling seamless cross-border collaborations and broadening my audience reach. The ability to **effortlessly translate** between English and Dutch has opened up new opportunities for engagement and understanding. For translations I rely on [DeepL](https://www.deepl.com/translator) and for editorial support on [LanguageTool](https://languagetool.org). In the article '[Blog writing workflow](https://rob-hoeijmakers.ghost.io/blog-writing-workflow/)' I explain my workflow. ## **Explore nature** AI has extended its influence beyond work, enriching my personal life. Apps that recognize trees, flowers, mushrooms, animals, and even clouds have added depth to my outdoor adventures. AI's ability to provide information and insights about the natural world has deepened my connection with the environment. The two main apps I use are [Obsidentify](https://observation.org/apps/obsidentify/) and [PlantNet](https://identify.plantnet.org). The first is more local and also trained on animals, insects and mushrooms. The second is very thoroughly trained on trees, plants and flowers. ![Exporing nature in action. Left Obsidentify and right PlantNet.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/11/AI-Images.jpg) Exporing nature in action. Left Obsidentify and right PlantNet. ## **Access to Up-to-Date Information** In our age of rapidly evolving information, staying informed is vital. I use a tool called Perplexity to access the latest information with reliable sources for day to day research. It ensures that I can provide clients and readers with the most current and credible information available. Here I go on depth what this AI-powered tool can do: [Perplexity, the AI-powered answer engine](https://rob-hoeijmakers.ghost.io/perplexity-the-ai-powered-answer-engine/). ## **The ongoing AI Journey** Reflecting on the past year, I'm astounded by the profound impact AI has had on my work and life. AI technologies, spanning various applications, have empowered me to work more efficiently, think more creatively, and explore new horizons. However, it's crucial to remember that AI is a versatile tool whose impact is shaped by our creative application. The future holds boundless possibilities as AI continues to advance. I eagerly anticipate discovering new ways to leverage these technologies to enhance my work as a web strategist and content writer. AI's potential is limitless, and I'm excited to be part of this ever-evolving journey. In conclusion, a year of AI transformation has taught me that these technologies are not merely tools; they are partners that can amplify our abilities and enrich our lives in unexpected and remarkable ways. Want to know more and see what AI could do for you or your company, [feel free to reach out](https://rob-hoeijmakers.ghost.io/about-rob-hoeijmakers/) or [subscribe to my newsletter for updates](#/portal/signup). I also was a podcast guest and talked about [the topic of using AI in blogging](https://contentandai.com/rob-hoeijmakers/). --- ### Further reading - [Rob Hoeijmakers: using AI to transform blogging workflows](https://contentandai.com/rob-hoeijmakers/) ### Paid subscription on X URL: https://rob-hoeijmakers.ghost.io/paid-subscription-on-x/ Last updated: 2026-04-25T13:16:02.000Z Social media platforms have evolved significantly over the years, and X (formerly known as Twitter) is no exception. As someone deeply immersed in the world of web strategy, online marketing, and chat technology, [my journey on X](https://twitter.com/robhoeij) has been a unique one. In this article, I want to reflect on the benefits of making the leap from a free user to a paid subscriber on X. ## Defining the Purpose of X X, for me, is more than just a platform for socialising. It's a versatile tool that plays a role in my professional life. I use it to socialise, share my written content, and showcase my iPhone 12 mini photography. More importantly, X is an integral part of my online marketing strategy. Let me explain how. ## Benefits of a Paid Subscription Lets's dive into the reasons behind upgrading from a free user to a paid subscriber on X. ### Enhanced Reach One of the primary reasons I decided to become a paid subscriber on X is to break free from the limitations that the platform imposes on the reach of messages. It may feel like "tricking the system" or creating inequalities, but it's a strategic move to ensure that my content doesn't get lost in the noise. X is a platform where attention is currency, and a paid subscription is my way of ensuring my voice is heard. It's an investment in [my personal brand and services as a web strategist](https://rob-hoeijmakers.ghost.io/about-rob-hoeijmakers/). 📉 I also checked the statistics, and they don't directly support my point. But I still argue for that a premium account leads to enhanced reach, but the increase has to be offset by the sharp decline in traffic on X recently. Reports mention a 15% decline. ### Longer Messages and Blog-Like Posts Another useful feature with a paid subscription on X is the ability to write longer messages. No longer do I need to resort to cumbersome thread-based posts; I can now craft messages that resemble blog articles. This feature has made content creation more efficient and has expanded the scope of what I can share with my audience. It's a tool I wholeheartedly embrace. ### The Blue Tick debate The coveted blue tick, once a symbol of expertise and professionalism, has evolved into something of a mixed bag. It's no longer a guarantee of status or expertise; it's simply a sign of being a paid subscriber. While it may be a somewhat doubtful trophy, I chose to display it as a testament to my commitment to X as a medium. It's a way of saying, "I'm here, and I take this platform seriously." ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/11/X-with-Tick.jpg) X account with Blue Tick 🤑 ****Costs of subscriptions** As a side note: Looking beyond X, we can observe a broader trend of users transitioning from free users to paying subscribers on various social media platforms, including Facebook and Instagram. The reasons vary, from privacy concerns to the need to stay connected with an audience. Of course this leads to new considerations. ## Platform at risk For a content creator X is a precarious medium, traffic on it is spiralling down and there have been signals of pulling out of Europe. If you are not in the circumstances to make moral judgments on the path and contents of X it makes sense to use it to the full and consider to get out and land your content and valuable audience somewhere else. 👀 On the moral side of X and social media in general, I refer to [an article by David Rushkoff](https://open.substack.com/pub/rushkoff/p/why-im-finally-leaving-x-and-probably?r=49ph0&utm%5Fmedium=ios&utm%5Fcampaign=post). It is a critical piece, and I found it useful to form an opinion. ## Not worth the money for most My subscription to X is not just a personal choice; it's a strategic move in my professional journey as a web strategist and content creator and partially a necessity. It's a way to ensure my content reaches the right audience and continues to add value. As social media platforms evolve, it's essential to adapt and leverage the available tools to stay ahead. The benefits of a paid subscription on X are clear for someone like me, who relies on it for professional growth and personal brand building. Whether it's the enhanced reach, longer messages, or the blue tick, these features serve as valuable tools in the ever-changing landscape of social media. So, the question remains: Is it time for you to make the transition from a user to a subscriber on X? The answer lies in your own social media usage patterns and professional objectives. But to be honest, **for most users and use cases, for people who want to socialise and keep up to date, I don't believe Premium is worth the money**. ### The paper RFID URL: https://rob-hoeijmakers.ghost.io/the-paper-rfid/ Last updated: 2026-04-25T13:16:42.000Z In the fast-paced world of technology, innovation often takes the form of complexity—cutting-edge gadgets, intricate circuitry, and high-tech solutions. However, every now and then, there comes an innovation that brings simplicity to the forefront, changing the game while staying grounded in the basics. This is precisely what the team at PulpaTronics has achieved with their groundbreaking paper RFID tags. But before we delve into this clever innovation, let's start with the basics. What is RFID, and where is it used? ## The basics of RFID Radio-Frequency Identification (RFID) might sound like a mouthful, but it's a technology that has quietly become a part of our everyday lives. In simple terms, RFID allows for the wireless tracking and identification of objects. Unlike traditional barcode scanning, RFID doesn't require a visual line of sight. Instead, it operates by sending and receiving radio waves to and from RFID tags. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/11/pulpatronics-rfid-tags.webp) RFID tags ## Where you find RFID You'll find RFID technology in various applications, but it's most commonly used in retail, logistics, and inventory management. Ever wondered how those self-checkout machines at the grocery store seem to magically identify your items without scanning? That's RFID at work. It's also instrumental in tracking packages during shipping, managing inventory in warehouses, and even preventing theft in retail stores. Now, here's where things get interesting—the downside of current RFID tags. 📻 ****Difference between RFID and QR codes** Both technologies used to store and retrieve data, but they have distinct characteristics. Similarities lie in their data storage capabilities and applications for contactless interactions. However, RFID uses radio waves for data transmission and can read tags from a distance without direct sight, making it ideal for inventory management and secure access. In contrast, [QR codes are 2D barcodes ](https://rob-hoeijmakers.ghost.io/qr-code-explained/)that rely on visual scanning through smartphone cameras, making them versatile for digital interactions but less secure. In essence, both technologies serve as efficient tools for data retrieval, but RFID excels in security and range, while QR codes offer simplicity and accessibility ## The drawbacks of traditional RFID tags While RFID technology has brought convenience to many industries, traditional RFID tags come with some environmental concerns. These tags typically contain metal and silicon components, and most importantly, a microchip. This combination of materials makes them non-recyclable and prone to ending up in landfills. In essence, they contribute to electronic waste, which we're all striving to reduce. Moreover, traditional RFID tags are energy-intensive. They require power to operate, which means they rely on batteries or active electronic components. This not only adds to their environmental impact but also increases costs for businesses. ## Simplifying RFID with Carbon Now, let's shift our focus to the innovative solution brought to us by PulpaTronics. Their approach to RFID is refreshingly simple yet incredibly clever. PulpaTronics has crafted RFID tags entirely from paper, eliminating the need for metal, silicon, or power-hungry microchips. You might wonder how this is possible. The magic lies in the use of carbon—a substance as humble as a pencil's graphite. Instead of complex components, PulpaTronics uses a laser to mark a carbon-based circuit onto the paper's surface. This laser not only simplifies the manufacturing process but also changes the paper's chemical composition to make it conductive. The result? RFID tags that are as environmentally friendly as a piece of paper with pencil markings. They can be recycled with household waste, reducing the environmental footprint of RFID tag production. 💡 The paper RFID tag, as described in the context of PulpaTronics' innovation, may not be suitable for all the sensitive and critical use cases mentioned earlier. While the paper RFID tag offers a more eco-friendly and simplified approach to RFID technology, it may have limitations when it comes to security, durability, and environmental conditions. ## The impact of PulpaTronics' innovation PulpaTronics estimates that their paper RFID tags can reduce carbon dioxide emissions by a whopping 70% compared to traditional RFID tags. Not only does this innovation benefit the planet, but it also makes business sense. It's expected to cut costs for businesses by half. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/11/pulpatronics-rfid-tags-info.webp) Conventional versus paper RFID ## In conclusion In a world where innovation often seems synonymous with complexity, PulpaTronics reminds us that clever innovation can be refreshingly simple. Their paper RFID tags, crafted with carbon and a touch of laser magic, offer a sustainable alternative to traditional RFID tags without the need for batteries or complex components. To learn more about PulpaTronics and their game-changing technology, visit [their website](https://www.pulpatronics.com). In the quest for eco-friendly solutions, PulpaTronics stands out as a beacon of simplicity and innovation. Keep an eye on this space, because sometimes, it's the simplest ideas that bring about the most profound changes. --- P.S. there is also an RF-tag. Those are used for security and don’t contain information. They need to be deactivated through a subtle form of destruction. You put them in a container and there the RF-tag are exposed to strong electromagnetic waves. ![RF-tag on a Decathlon padel racket ](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/11/IMG_7859.jpeg) RF-tag on a Decathlon padel racket ### Perplexity, the AI-powered answer engine URL: https://rob-hoeijmakers.ghost.io/perplexity-the-ai-powered-answer-engine/ Last updated: 2026-04-25T13:16:35.000Z In the ever-evolving world of AI and technology, a new player has emerged onto the scene, and it goes by the name of [Perplexity.ai](https://www.perplexity.ai). If you're a web strategist, content creator, or simply someone who values accurate and comprehensive information, you'll want to pay attention to this innovative tool. ![Rule of thirds in photography, answer on the iOS app.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/10/Untitled-design-2.jpg) Rule of thirds in photography, answer on the iOS app. ## What is Perplexity.ai? Perplexity.ai is not your ordinary search engine; it's an AI-powered answer engine designed to provide users with precise and informative responses to their questions. Imagine having a search tool that doesn't just offer links but instead serves you with concise and accurate answers directly. That's the promise of Perplexity.ai. Built on the foundation of OpenAI's GPT-3.5 model, Perplexity.ai brings a host of impressive capabilities to the table. It puts users, not advertisers, at the center of its mission, ensuring that you receive valuable information without distractions. 🎭 ****How is Perplexity.ai different from ChatGPT?** While both Perplexity.ai and [ChatGPT](https://rob-hoeijmakers.ghost.io/try-out-chatgpt-yourself/) are powered by AI language models, they have different focuses and applications. Perplexity.ai is geared towards providing direct answers and retrieving information efficiently, while ChatGPT specializes in natural language conversations and interactive interactions. The choice between them depends on the specific use case and the nature of the interaction you want to create. ## How does Perplexity AI work? Here's a step-by-step breakdown of how Perplexity.ai operates: 1. **Inputting a Question**: Instead of traditional keyword searches, users can ask their questions directly to Perplexity.ai via its user-friendly interface. 2. **Searching the Internet**: Upon receiving a question, Perplexity.ai scours the internet for relevant information. It harnesses the power of OpenAI's GPT-3.5 and Microsoft Bing to find the best answers. 3. **Generating Answers**: The magic happens when Perplexity.ai uses large language models (LLMs) to provide precise answers. It's not limited to simple queries; it can handle complex questions that require synthesizing content from multiple sources. 4. **Displaying Sources**: What sets Perplexity.ai apart is its transparency. After delivering answers, it also presents the sources of the information it used, making it easy for users to verify and explore further. ![Are civets cats? The perplexity web app with on top the sources for the answer.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/10/Perplexity.jpg) Are civets cats? The perplexity web app with on top the sources for the answer. ## Pricing model for Perplexity AI Perplexity.ai offers both a free plan and a premium plan tailored to users' needs. Here's a breakdown of the pricing: - **Free Plan**: Basic features are accessible to all users at no cost. However, some advanced features may be limited in the free plan, such as access to GPT-4 and Copilot. - **Premium Plan**: For $20 per month or $200 per year, users can unlock all of Perplexity.ai's features. This includes unlimited Copilot usage, GPT-4 access, dedicated support, and updated information. The yearly plan also offers a free seven-day trial. Whether you're a casual user seeking quick answers or a professional in need of advanced features, [Perplexity.ai](https://www.perplexity.ai)'s pricing model has you covered. 💡 ****What is Copilot in This Context?** Copilot is a valuable feature of Perplexity.ai. It leverages OpenAI's GPT-4 model to provide personalised and interactive responses. Think of it as your intelligent search assistant. It guides you through the search process, asks clarifying questions, summarises information, and curates answers with links to relevant websites and data sources. It's available to both free and premium users, with the premium plan offering unlimited usage. ### Improve your YouTube advertising with Junbi URL: https://rob-hoeijmakers.ghost.io/youtube-advertising-junbi/ Last updated: 2026-04-25T13:16:39.000Z In the fast-paced world of online marketing, staying ahead of the curve is essential. That's why I sat down with [co-founder Daan van der Wiele](https://www.linkedin.com/in/daanvanderwiele/), an expert in helping brands build better YouTube ads with the online tool called Junbi. Junbi can change the way you **analyse and benchmark YouTube commercial videos**. This tool is powered by a predictive AI that is based on thousands of earlier scientific measurements of YouTube ads and so can save you tons of work and help you focus on optimising for results. In this blog post, I will first delve into how Junbi is rooted in neuroscience and then I will zoom in into the three main metrics that are being used in the tool. 💡 ****Who can benefit from Junbi?** Online professionals of all kinds can reap the rewards of Junbi's tool: ****Digital Marketers:** Enhance your YouTube ad campaigns, optimize spending, and achieve better Return on Ad Spend (ROAS) with the insights Junbi provides. ****Content Creators:** Whether you're an influencer or a content producer, Junbi can help you create ads that capture your audience's attention effectively. ****Brands:** Ensure your brand gets the attention it deserves and stay competitive in the ever-evolving digital landscape. ****Advertising Agencies:** Streamline your ad testing processes and deliver exceptional results for your clients. ## The science behind Junbi Junbi uses neuroscience in their product by leveraging AI to provide advertisers with insights into the effectiveness of their YouTube creative. Here's how they incorporate neuroscience into their platform: - **Neuroscience-based effectiveness drivers**: Junbi helps advertisers create better YouTube ads by providing second-by-second insights on their neuroscience-based effectiveness drivers. These drivers are rooted in the learnings from testing over 80,000 ads with neuroscience. - **Eye-tracking results**: The platform offers predictive eye-tracking results, which can be used to optimise the YouTube ads. By analyzing how viewers' eyes move across the screen, advertisers can make informed decisions about visual elements and their placement. - **Benchmarking data**: Junbi provides benchmarking data to help advertisers understand how their YouTube ads perform compared to industry standards. This data allows them to identify areas for improvement and make data-driven optimizations. In short, Junbi uses neuroscience in their product by combining **the power of predictive AI, historic eye-tracking results, and** [**benchmarking**](https://rob-hoeijmakers.ghost.io/where-does-the-word-benchmark-come-from/) **data** to provide advertisers with science-based insights into the effectiveness of YouTube creatives. ## Understanding Junbi's Core Measures Junbi's uses three key measures that provide invaluable insights for online professionals. They show you if you ad gives the proper attention to your brand, if it has the potential to break through the clutter and if the visual story is to be quickly understood. Let's explore these core measures: ### 1\. Brand Attention In the realm of advertising, the ability to capture and maintain your audience's attention is paramount. Junbi's Brand Attention measure is a game-changer in this regard. Here's what it entails: - **Measure the attention on your brand:** In an age where information overload is the norm, ensuring your brand receives the attention it deserves is vital. Junbi's Brand Attention score offers a quantifiable metric to gauge how much attention your brand is getting. - **Optimize brand recall:** Your audience will remember what they've seen, making it crucial to ensure your brand stands out. Junbi helps you easily measure and quantify brand attention, enabling you to optimise brand recall effectively. ![Brand attention](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/10/brand-attention-product-pic-v3-1536x985.png) Brand attention ### 2\. Ad Breakthrough In a sea of visual content, your ad's ability to stand out is a make-or-break factor. Junbi's Ad Breakthrough measure assesses your ad's performance in the context of its surroundings. Here's what it offers: - **Stand out from the noise:** Your ad needs to grab attention, even when surrounded by competing visuals. The Ad Breakthrough score evaluates your ad's contextual placement and measures its effectiveness in standing out. - **Contextual assessment:** Junbi places your ad within its surrounding visuals, allowing you to understand how well it catches the eye in its specific environment. For when the ad is presented on desktop or mobile. And when on mobile, it the orientation portrait or landscape. ![Ad breakthrough visualised ](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/12/ad-breakthrough-visual-1536x1055.jpeg) Ad breakthrough visualised ### 3\. Cognitive Ease The human brain prefers simplicity when processing information. Junbi's Cognitive Ease measure is designed to ensure your ad is easy to digest and remember: - **Ensure your ad is easy to process:** Visually complex ads can be challenging for viewers to process and remember. The Cognitive Ease score evaluates how easily your ad can be understood and recalled by your audience. - **Optimize message retention:** By measuring cognitive ease, Junbi assists you in creating ads that are not only visually appealing but also easy for viewers to comprehend and remember. ![Cognitive ease](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/12/cognitive-ease-visual-1-1536x1008.jpeg) Cognitive ease score In conclusion, Junbi's offers a powerful set of core measures that can improve and benchmark your YouTube advertising campaigns. With Brand Attention, Ad Breakthrough, and Cognitive Ease at your fingertips, you can optimize your ads for maximum impact and audience engagement. > As [a web strategist](https://rob-hoeijmakers.ghost.io/about-rob-hoeijmakers/), I believe Junbi.ai is a vital tool for ensuring cost-efficiency during the early stages of YouTube ad production. It empowers advertisers with invaluable insights to optimize their creative, ultimately leading to more effective and budget-friendly campaigns. Don't miss out on the opportunity to stay ahead of the competition with Junbi, I was truly impressed by what I saw. To learn more about how Junbi can transform your YouTube ad campaigns using these core measures, checkout [the Junbi website](https://junbi.ai/?utm%5Fsource=hoeijmakers.net&utm%5Fmedium=referral) and see if your company qualifies for a demo. ### Orchestrate personal experiences with Conscia URL: https://rob-hoeijmakers.ghost.io/conscia-dx-engine-and-dx-graph/ Last updated: 2026-04-25T13:16:40.000Z In the fast-paced world of digital interactions, businesses are constantly challenged to meet evolving customer demands while managing legacy systems. A possible solution to this challenge lies in adopting modern architectural principles. Enter Conscia's suite of solutions, including the DX Engine and DX Graph, which are not just tools but orchestrators of digital transformation. In a recent demo led by [Sana Remekie](https://www.linkedin.com/in/sana-remekie/), Conscia's landscape and visionary ideas were discussed, shedding light on the orchestration of digital transformation. The orchestration continued with a demo by [Jason Yen](https://www.linkedin.com/in/jasonyen/), showcasing how Conscia's solutions bring these ideas to life. Let me talk you through the tools. ## DX Engine **Empowering Business Teams and Legacy Systems** Legacy systems often require developers to write intricate code to orchestrate content and data, which can be costly and time-consuming. The DX Engine steps into the role of orchestrator, redefining the process. It places operational control in the hands of business teams, freeing developers from crafting complex frontend "glue code." The heart of the DX Engine is its rules engine, empowering businesses to define and manage sophisticated rules for content delivery. This orchestration allows for highly targeted and context-aware experiences. Real-time decisioning ensures customers receive up-to-date, personalized content, further enhancing the orchestration of their digital journey. Additionally, the DX Engine provides zero-code API orchestration, dramatically cutting development costs and simplifying integration. Moreover, data and analytics capabilities, coupled with A/B testing, offer insights for continuous optimization. The DX Engine doesn't just streamline operations; it orchestrates them for peak efficiency. ## DX Graph **Unifying Data and Content for Legacy Systems** In the digital realm, data and content often reside in isolated silos. Here, the DX Graph takes center stage as the chief orchestrator. It centralizes data and content from legacy, home-grown, and proprietary systems into a single hub accessible through real-time APIs. This orchestration is a lifeline for legacy systems. It modernizes them with an API-first layer, extending their lifespan and vitality. The DX Graph simplifies master data management, ensuring data flows smoothly and efficiently. It also orchestrates the creation of a knowledge graph, providing comprehensive insights that further enhance decision-making. By orchestrating the unification of data and content, the DX Graph supports legacy systems, giving them a boost that is often more cost-effective than costly replatforming. It's a testament to the power of orchestration in extending the life of older systems. ![Screenshot from the presentation and the demo by Conscia.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/10/2023-10-11_17-13-04.jpg) Screenshot from the presentation and the demo by Conscia. ## My personal take Conscia's DX Engine and DX Graph are more than just tools; they are orchestrators of digital flows. By placing operational control in the hands of business teams, unifying data and content, and orchestrating the modernisation of legacy systems, Conscia can empower businesses to stay agile and competitive in today's digital landscape. If you like the idea, please [check out the orchestrator modules](https://conscia.ai/contact/) and try to unlock the full potential of your digital ecosystem and extend the life of your legacy systems cost-effectively. ### This week ChatGPT learned to draw, see and talk URL: https://rob-hoeijmakers.ghost.io/this-week-chatgpt-learned-to-draw-see-and-talk/ Last updated: 2026-04-25T13:17:08.000Z It was indeed a glorious week for [ChatGPT](https://rob-hoeijmakers.ghost.io/try-out-chatgpt-yourself/) enthusiasts as three exciting new features were launched, making this AI-powered tool even more versatile and capable. In this blog post, we'll delve into these exciting developments and provide examples to showcase their functionality. I will use a demo approach which will help you discover and use the features described. ### **1\. Image Generation with DALLE-3 Integration** One of the most anticipated additions is the integration of [DALLE-3](https://rob-hoeijmakers.ghost.io/playing-with-ai-system-dall-e/), a cutting-edge image generation model. Now, ChatGPT can generate images based on text descriptions, opening up a world of creative possibilities. Whether you need illustrations for your content or want to visualise your ideas, ChatGPT's DALLE-3 integration has you covered. **Example:** Imagine describing a fantastical scene in your blog post (Show the Amsterdam Rijksmuseum alone in a polder like landscape), and with a simple command, ChatGPT can bring it to life with a stunning image like this: ![Example of image generation with DALLE-E in ChatGPT](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/10/F8LcZ7EXYAAu2rC.jpeg) Example of image generation with DALLE-E in ChatGPT 💡 Please note that the first feature, image generation with DALLE-3 integration, is currently available exclusively on the browser version of ChatGPT and is not yet accessible through the app. To access this feature, simply visit ChatGPT on your web browser, switch to GPT-4 mode, and then select DALLE-3. ### **2\. Image Analysis and Description** Another remarkable feature is the ability to upload images for analysis and description. This functionality can be a game-changer for content creators, marketers, and strategists. ChatGPT can provide detailed insights about the content of an image, making it easier to understand and utilise visual information. **Example:** Suppose you've captured a photo of a some scene, and you want to know more about its characteristics. Upload the image, and ChatGPT can analyse and describe it like this: - **Image Analysis:** “What do we see here?” - **Description:** "The image displays a collection of wooden pencils neatly arranged in wooden compartments or shelves. Each compartment contains pencils with the same coloured tip, indicating that the pencils have been organised by colour. The tips of the pencils are prominently visible, showcasing a vibrant spectrum of colours. The pencils seem to be of the classic hexagonal shape and appear unused or sharpened to a uniform length. The wooden compartments lend a rustic aesthetic to the image. Overall, the picture captures an artistic and orderly presentation of coloured pencils." ![Example of image analysis from within ChatGPT](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/10/1697201014615.jpeg) Example of image analysis from within ChatGPT 💡 I tested this both on browser and app. You select ChatGPT-4, upload an image from the prompt box and start asking questions. ### **3\. Voice Interaction with ChatGPT** The third exciting feature is the ability to have voice interactions with ChatGPT. This takes conversational AI to a whole new level. You can now communicate with ChatGPT using your voice, enabling more natural and dynamic conversations. **Example:** Let's say you're brainstorming ideas for your next blog post. Instead of typing out your thoughts, you can simply speak to ChatGPT, and it will respond with voice, enhancing the flow of your creative process. 💡 Activate this by going to the app, go to settings and choose new features. There you will see a button to activate 'Voice conversations'. ![After activation in the settings you will find an icon to start voice conversations](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/10/ChatGPT-talks.jpg) After activation in the settings you will find an icon to start voice conversations ## Draw, see, and talk These three new features represent significant advancements in the capabilities of ChatGPT. From generating images to analysing them and engaging in voice conversations, ChatGPT is evolving to be a comprehensive tool for content realisation, online marketing, and more. As a [web strategist](https://rob-hoeijmakers.ghost.io/about-rob-hoeijmakers/) I am impressed by the capability to analyse photo’s through prompting. And for the content creator, these new features can undoubtedly amplify the capabilities and bring a new level of creativity to your work. So, check out these exciting developments and explore the endless possibilities they offer in your field. And I hope these simple demo’s help you to keep track of all the innovations and use them. ### Navigating BlueSky URL: https://rob-hoeijmakers.ghost.io/bluesky/ Last updated: 2026-04-25T13:16:41.000Z In the digital age, we've become accustomed to residing in virtual spaces, much like tenants in the real world. Every social media platform we join is a digital home, and like any responsible tenant, it's crucial to understand both the landlord and the rules governing our online abode. In this article, we'll explore the world of BlueSky, a social media app that's changing the game, and we'll emphasize the importance of knowing the rental boss and the rules before moving in. ## Meet Your Digital Landlord: BlueSky BlueSky, often dubbed as the next frontier in social media, is not your typical digital landlord. Founded with a unique vision, BlueSky is more than just a social media platform; it's a pioneering force in reshaping how we engage and interact online. ![The web version of BlueSky](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/10/bluesky-web.jpg) The web version of BlueSky ### The vision and mission of BlueSky At the heart of BlueSky lies a profound mission: **to transform the way social media works, making it more open, decentralized, and user-centric**. Unlike traditional social networks, where a small group of people control everything, BlueSky aims to return power to the users and creators. The driving force behind this transformation is the AT Protocol, a groundbreaking open-source framework for building social apps. It standardizes user identity, follows, and data, allowing for interoperability among apps and easy migration between platforms. Think of it as a passport for the digital world, giving you the freedom to move between digital cities while taking all your belongings—your data and connections—with you. ### The AT Protocol gives portability The AT Protocol empowers users, developers, and communities. It breaks the chains of centralized social networks, where leaving often means starting from scratch. With the AT Protocol, **account portability is a reality**. Imagine moving to a new city without leaving your friends and possessions behind. This is the essence of account portability, one of the most significant differentiators of apps built on the AT Protocol. In addition to account portability, BlueSky is working on algorithmic choice and composable moderation, giving users even more control and transparency over their online experience. ### Who's in charge? BlueSky is not just another faceless corporation. It's a Public Benefit Corporation, owned by Jay Graber and the Bluesky team, with Jack Dorsey and Jeremie Miller serving on the board. This unique structure underscores BlueSky's commitment to its mission of open and decentralized public conversation. 💰 ****BlueSky Business Model** Bluesky PBLLC, the company behind Bluesky Social, is [exploring multiple revenue models for its platform](https://theintercept.com/2023/06/01/bluesky-owner-twitter-elon-musk/). While specific details have not been disclosed, potential monetization avenues include advertising and a paid subscription component. ### The Relationship with Twitter BlueSky's journey began as a project initiated by Jack Dorsey during his tenure as Twitter's CEO in 2019\. Jay Graber was chosen to lead BlueSky, and Twitter supported its development as an open social protocol. However, in late 2022, Twitter and BlueSky chose to go their separate ways. BlueSky continued its mission independently, focusing on "developing and driving large-scale adoption of technologies for open and decentralized public conversation." 🎭 Main differences I, as a publisher and user, now see between X and BlueSky is that the audience is still limited and I miss companies and organisations. But it is still in the invite phase. Not yet DM's, polls and video's. ### How to join BlueSky As of now, BlueSky has welcomed thousands of users onto its platform, and it plans to grow steadily while preserving a focus on protocol development and fostering healthy conversations. The invite system is a key part of this strategy, with invites distributed based on existing users' trustworthiness and contributions to the community. To join BlueSky, sign up for the waitlist at [bsky.app](https://bsky.app/). While some may offer invite codes for sale, it's essential to exercise caution, as legitimacy cannot always be guaranteed. ![BlueSky Mobile app](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/10/BlueSky.jpg) BlueSky Mobile app ### Moderation: Prioritizing User Safety and Control BlueSky's approach to moderation is innovative and multi-faceted. It incorporates automated filtering, manual admin actions, and community labeling. Users and services can define and apply labels to content, giving them control over their experience. This approach aims to strike a balance between user safety and individual control over content. 🤔 This is the part I experienced the least. In our polarised society moderation is required but it also is part of politics itself. ## Your Digital Space, your responsibility In a world where our online presence is as real as our physical existence, understanding the digital spaces we inhabit is important. Just as you would research a landlord before renting a physical apartment, it's essential to know your digital landlord and the rules of engagement on any online platform. BlueSky stands at the forefront of this digital change, offering a new way to experience social media—one that empowers users and promotes openness. By understanding BlueSky's vision, structure, and unique features, you can make an informed decision about your digital tenancy, ensuring a fulfilling and safe online experience. *I am on* [*X (@robhoeij)*](https://twitter.com/robhoeij)*,* [*Threads*](https://www.threads.net/@robhoeij/) *and* [*BlueSky (@hoeijmakers.net)*](https://bsky.app/profile/hoeijmakers.net)*. I share stories, trivia and my iPhone photography. About nature, language and history. And who knows do I meet you there.* ✌️ ****BlueSky versus Mastodon** BlueSky and Mastodon differ in their core approaches; BlueSky primarily concentrates on developing the AT Protocol for decentralized social apps, with the Bluesky app as an illustration, while Mastodon is a well-established decentralized social media platform with a community-driven, federated network of instances. The choice between them hinges on whether you prefer a protocol-first approach with the potential for greater decentralisation or a fully functional decentralised social platform with a diverse user base and community-driven governance. ### Digital Asset Management: Where Art and Science Converge URL: https://rob-hoeijmakers.ghost.io/digital-asset-management-where-art-and-science-converge/ Last updated: 2026-04-14T14:21:53.000Z In today's digital age, where information flows at the speed of light, managing your organization's digital assets efficiently can mean the difference between success and obscurity. This is where [Digital Asset Management (DAM) systems](https://rob-hoeijmakers.ghost.io/digital-asset-manager-dam/) come into play. In this blog post, I explore how a DAM brings together the realms of art and science, as witnessed through the experiences of companies like [Bystronic](https://www.bystronic.com/), cultural institutions like the [Rijksmuseum](https://www.rijksmuseum.nl/nl), and dynamic solutions like [UEFA's](https://www.uefa.com) real-time content delivery. ## The Versatile World of DAM Digital Asset Management is a versatile tool, and the canvas on which art and science seamlessly merge. It's like a digital gallery housing all your media assets, from images and videos to documents and audio files. The central idea is simple: manage your content in one place, but make it come alive on multiple platforms. ## Bystronic's Multifaceted Approach Imagine a company like Bystronic, which not only utilizes a robust DAM system but also collaborates with Smint.io and their local agency for content activations. To gain deeper insights into their approach, I had the chance to interview [Fabienne Flückiger](https://www.linkedin.com/in/fabienneflueckiger/), who shared Bystronic's DAM journey. They educate their staff with up-to-date training materials, engage clients with compelling visuals, inform the press with the latest news, and most vividly, interact with potential employees. All the imagery and videos used in these interactions come straight out of the DAM, in synergy with [Smint.io](https://www.smint.io) and their local agency. It's like a living, breathing representation of the company's ethos. This multifaceted approach not only demonstrates Bystronic's commitment to leveraging technology for engagement but also highlights the efficiency benefits of reusing systems and knowledge. Bystronic can achieve scale while maintaining a lean team, thanks to the seamless integration of their DAM, Smint.io, and their local agency. There digital setup is a testament to the power of technology and collaboration in today's business landscape. ![Career site from Bystronic on a screen as example.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/09/Bystronic-career.jpg) Great example of the DAM coming to life is the [career site from Bystronic](https://www.bystronic.com/int/en/career). ## Bridging Differences with DAM: Bystronic, Rijksmuseum, and UEFA [Bystronic's multifaceted use of DAM](https://www.fotoware.com/blog/how-dam-enables-bystronic-to-keep-brand-consistency-across-channels), coupled with their collaboration with Smint.io and a local agency, showcases the power of digital asset management in corporate and industrial settings. But DAM's influence extends well beyond the corporate realm, and it's exemplified by two entirely different yet equally compelling use cases: the Rijksmuseum in Amsterdam and UEFA. ### Bringing Art to Life: The Rijksmuseum Connection DAM's influence extends well beyond corporate applications. Consider my personal experience with the Rijksmuseum in Amsterdam, where DAM technology enables art to transcend physical boundaries. Visitors can browse and download digital reproductions, opening the doors of artistic heritage to a global audience. It's a prime example of how DAM breathes life into culture and art. The same DAM technology that Bystronic employs for corporate training, client engagement, and employee interaction becomes a conduit for culture at the Rijksmuseum. It's a testament to the versatility of DAM systems, bridging the gap between corporate and cultural domains. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/09/Rijksmuseum.jpg) The rich collection of the Rijksmuseum can be accessed and [browsed in a studio](https://www.rijksmuseum.nl/nl/rijksstudio). ### UEFA's Real-Time Content Delivery But let's switch gears and delve into the world of sports. During a match, photographers attach a special key to their camera's metadata. They upload photos in real-time, and here's where the science comes in: facial recognition technology names the players, making the images instantly available for third parties. It's not just about storing; it's about dynamic content delivery. This real-time content delivery system employed by UEFA is underpinned by the same DAM technology. It showcases how DAM is not only a bridge between art and science but also between entirely different sectors. From corporate to culture to sports, DAM is the common thread that streamlines content management, making it accessible and dynamic. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/09/uefa-action.jpg) Always Knowing the Score. Read more [about this dynamic workflow](https://picturepark.com/digital-asset-management-case-study/uefa). In essence, Bystronic, the Rijksmuseum, and UEFA may seem worlds apart, but they all rely on the same DAM software. It's a powerful example of how a well-implemented DAM system can bridge differences and serve as a unifying force in a diverse digital landscape. Whether you're educating staff, preserving art, or enhancing the sports experience, DAM brings efficiency, accessibility, and life to content in all its forms. ## Cathrine Wessel's Insight from FotoWare To understand the driving force behind the creation of such powerful DAM systems, I had the privilege of interviewing [Cathrine Wessel from FotoWare](https://www.fotoware.com). She shared invaluable insights into how DAM bridges the gap between art and science. What's particularly noteworthy is that the FotoWare Group [recently acquired](https://www.fotoware.com/blog/fotoware-acquires-swiss-digital-asset-management-provider-picturepark) the Swiss DAM company Picturepark and their core product Content Platform, which is the actual solution behind the three examples we've explored. FotoWare's commitment to creating a platform that empowers organisations to [manage, activate, and share their digital assets](https://www.fotoware.com/blog/what-is-content-operations) is evident in the stories we've covered - stretching across industries and countries. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/10/About-FotoWare-Group.jpeg) About FotoWare Group. ## The Essence of "Coming to Life" Throughout these examples, a common theme emerges. DAM isn't just a digital vault; it's a stage where companies, museums, and organizations come to life. It's a place where branding, storytelling, and engagement blend seamlessly. > Art meets science to create a symphony of efficiency and creativity. In conclusion, Digital Asset Management systems are the conduits where art and science converge in the digital landscape. From Bystronic's versatile corporate applications to the Rijksmuseum's cultural enrichment and UEFA's real-time magic, DAM proves its worth in diverse domains. It's not just about managing content; it's about making it breathe, creating a vibrant, living entity out of data. So, if you're still wondering whether DAM is worth the investment, the answer is clear: It's where the magic of art and the precision of science intertwine to bring your content to life. ## About the Author I am an [Amsterdam-based web strategist](https://rob-hoeijmakers.ghost.io/about-rob-hoeijmakers/) with a passion for exploring the intersections of technology, marketing, and content realisation. My expertise spans web builders, SEO, conversion optimisation, marketing technology, and chat technology. If you wish to connect with me or explore collaboration opportunities, please [visit the contact page](https://rob-hoeijmakers.ghost.io/contact/) for various options. ### What are Image Optimization Services for? URL: https://rob-hoeijmakers.ghost.io/what-are-image-optimization-services-for/ Last updated: 2026-04-14T13:25:37.000Z In the ever-evolving world of web development and digital marketing, optimising images is a crucial step towards enhancing user experience and improving website performance. Image optimisation services have emerged as powerful tools to achieve these goals, but how do they compare to [Digital Asset Management (DAM)](https://rob-hoeijmakers.ghost.io/digital-asset-manager-dam/) systems? Let's dive into this topic to shed some light on these essential web technologies. ## Enhancing web performance Image optimisation services are the hidden champions of web performance. They focus on making images smaller, faster to load, and perfectly suited for the online world. These services are essential for improving page speed, creating a seamless user experience, and enhancing your website’s search engine rankings. As a [web strategist and blogger](https://rob-hoeijmakers.ghost.io/about-rob-hoeijmakers/) covering web technologies and online marketing, getting to grips with these tools is crucial. Let me list some of the most used tools. ![Alt TextI uploaded a HEIC image to imagekit.io and asked ChatGPT to provide me with a URL for a properly sized WebP image. ](https://ik.imagekit.io/ivdzvx7f6/IMG_6289_aNM7NBWVr.HEIC?tr=w-1600,f-webp) I uploaded a [HEIC image](https://rob-hoeijmakers.ghost.io/heic-a-blood-type-for-images/) to imagekit.io and asked [ChatGPT](https://rob-hoeijmakers.ghost.io/try-out-chatgpt-yourself/) to provide me with a URL for a properly sized [WebP image](https://rob-hoeijmakers.ghost.io/webp-versus-jpg/). ## Popular image optimisation services Several image optimisation services have made a name for themselves in the world of web development. Here's a brief overview of some of the notable players in this field: 1. [**ImageKit.io**](https://imagekit.io): ImageKit.io offers a comprehensive cloud-based solution for image optimisation and delivery. It automatically optimises images, resizes them, and provides responsive image delivery. 2. [**Cloudimage by Scaleflex**](https://www.scaleflex.com/): Cloudimage by Scaleflex is a cloud-based image processing and delivery service. It optimises, resizes, and delivers images through a content delivery network (CDN). 3. [**Cloudinary**](https://cloudinary.com): Cloudinary is a versatile image and video management platform. It offers features for image transformation, resizing, and content delivery. 4. [**Kraken.io**](https://kraken.io): Kraken.io specializes in image optimisation, reducing image file sizes without compromising quality. 5. [**Imgix**](https://www.imgix.com): Imgix provides real-time image transformation and optimisation capabilities. Each of these services comes with its unique set of features and benefits, catering to various web development needs. 💡 ****Difference Image Optimisation Service and a CDN** Imagine your website as a high-speed delivery service for web content. Image optimization is like packing your goods efficiently to make them lighter and quicker to load, ensuring your customers receive their orders swiftly. On the other hand, a Content Delivery Network (CDN) acts as the global network of distribution centers, strategically placed around the world, ensuring that those optimized goods are delivered from the closest center to your customers, reducing delivery time even further. Together, image optimization and CDNs supercharge your website, providing lightning-fast experiences for your visitors. ## Comparing image optimisation services with DAM Systems While image optimisation services are instrumental in delivering optimised content to users, it's essential to differentiate them from DAM systems. Here's how they compare: ### **Focus and Scope** - *Image Optimisation Services*: These services primarily focus on optimising and delivering images for websites and applications. Their primary goal is to enhance web performance. - *DAM Systems*: Digital Asset Management systems have a broader scope, managing various digital assets, including images, videos, documents, and more. They often serve internal asset management and collaboration needs. ### **Use Cases** - *Image Optimisation Services*: Ideal for external content delivery, these services cater to web and app developers, marketers, and SEO specialists. - *DAM Systems*: DAM systems are commonly used for internal asset management within organisations, enabling teams to organise, collaborate on, and distribute digital assets effectively. ### **Features** - *Image Optimisation Services*: Core features include image optimisation, resizing, and content delivery. They are designed to streamline external content delivery. - *DAM Systems*: DAM systems offer metadata management, version control, asset organisation, and collaboration features, making them suitable for broader digital asset management requirements. ## Choosing the right solution When deciding between image optimisation services and DAM systems, it's crucial to consider your project's specific needs. Factors like the type of content, the scale of operations, and collaboration requirements should guide your decision. Image optimisation services excel at enhancing web performance and user experience, while DAM systems provide comprehensive asset management solutions. ![Cloudinary interface showing some of the transformations available.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/09/cloudinary.jpg) Cloudinary interface showing some of the transformations available. ### Vignettes in iPhone photography URL: https://rob-hoeijmakers.ghost.io/vignettes-in-iphone-photography/ Last updated: 2026-04-25T08:31:49.000Z In the world of photography, we often seek ways to make our images stand out. While many of us are familiar with the various filters and editing tools available on our smartphones, there's one feature that remains relatively unknown but can have a powerful impact on your photos – vignettes. ## What Are Vignettes? Vignettes are a subtle yet effective photography technique that involves darkening or fading the edges of a photo, drawing the viewer's attention towards the center of the frame. It's like creating a natural spotlight on your subject, making it perfect for online professionals who want to enhance their visual content. ![The word vignette comes from vine. In books text or graphics were adorned with a motif. ](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/09/2012FK6128.jpg) The word vignette comes from vine. In books text or graphics were adorned with a motif. ## When to Use Vignettes: As online professionals, you might wonder when to incorporate vignettes into your photography toolkit. Here are some scenarios where they can prove beneficial: 1. **Highlighting a Key Element**: Use vignettes to emphasize a specific element in your web content, such as a product, a person, or a call-to-action button. 2. **Creating a Mood**: Vignettes can set the mood for your online marketing materials. A subtle vignette can add a touch of elegance to a product photo, while a heavier vignette can create a dramatic atmosphere for a blog post header. 3. **Framing Text**: If you overlay text on images for your website or social media posts, vignettes can help frame the text and make it more readable against busy backgrounds. ![Miffy with light vignette, no vignette and a dark vignette.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/09/nijntje.jpeg) Miffy with light vignette, no vignette and a dark vignette. ## Finding Vignettes on the iPhone: Now that we understand the potential of vignettes, let's talk about how to access and use this feature on your iPhone. It's simpler than you might think: 1. **Open the Photo App**: Launch your iPhone's photo app. 2. **Edit photo**: Select editing photo. 3. **Enable the Vignette Effect**: Scroll through the settings until you find "Vignette." Tap on it to apply the effect. 4. **Adjust the Intensity**: After applying the vignette, you can further adjust its intensity by tapping the "Vignette" icon again and using the slider to find the right balance. How to find vignette on the iPhone ## Tips for Effective Vignette Usage: While adding vignettes is straightforward, here are some tips to make the most of this feature: 1. **Keep It Subtle**: Vignettes work best when they are subtle. Avoid overdoing it, as an excessive vignette can make your photo look unnatural. 2. **Experiment**: Don't hesitate to experiment with different levels of vignetting to see what suits your content best. 3. **Think About Composition**: Consider your photo's composition and where you want to draw the viewer's attention before applying the vignette. ## Conclusion: Vignettes are a hidden gem in [iPhone photography](https://rob-hoeijmakers.ghost.io/taking-photos-with-the-iphone-12-mini-a-powerful-camera-in-a-compact-size/) that can significantly enhance your online content. By using them strategically, you can create captivating visuals that engage your audience and convey your message effectively. So, next time you're [capturing images for your website, blog, or social media](https://rob-hoeijmakers.ghost.io/three-ways-to-enrich-your-website-with-top-photos/), give vignettes a try and watch your visuals come to life in a whole new way. ### Google's TextFX gives text effects URL: https://rob-hoeijmakers.ghost.io/googles-textfx-gives-text-effects/ Last updated: 2026-05-26T19:29:51.000Z As [a web strategist](https://rob-hoeijmakers.ghost.io/the-web-strategist-and-the-web-strategy/) and content realization enthusiast, I recently delved into the world of TextFX, and the possibilities it offers are nothing short of fascinating. Whether you're a poet, copywriter, songwriter, or simply someone who values creativity by association, TextFX has something to offer you. ## A Playground for Creative Minds At first glance, TextFX may seem abstract, but with a little experimentation, you'll quickly grasp its potential. The tool consists of **ten creative modules**, each designed to expand your writing process by generating innovative possibilities with text and language. Let's take a closer look at some of TextFX's key features: 1. **Simile:** Create vivid comparisons to enhance your descriptions. 2. **Explode:** Break words into similar-sounding phrases for a playful twist. 3. **Unexpect:** Add unexpected and imaginative elements to your scenes. 4. **Chain:** Connect semantically related items to build powerful narratives. 5. **POV:** Explore topics from multiple points of view for well-rounded content. 6. **Alliteration:** Curate topic-specific words that start with a chosen letter for a unique touch. 7. **Acronym:** Craft creative acronyms from words, adding depth to your text. 8. **Fuse:** Find intriguing intersections between seemingly unrelated ideas. 9. **Scene:** Generate rich sensory details to immerse your readers. 10. **Unfold:** Seamlessly slot words into other phrases, creating a tapestry of language. ![Alliteration with Google TextFX](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/09/textfx.jpg) Alliteration with Google TextFX ## User-Friendly Experience One of the standout features of TextFX is its [user-friendly interface](https://rob-hoeijmakers.ghost.io/user-experience-ux-and-ux-design/). The layout is thoughtfully designed, making navigation a breeze. As you experiment with the various tools, you can generate ideas, pin the results you like, and even download them for future reference or further refinement. But that's not all. TextFX also includes video clips featuring Lupe Fiasco, the rapper who played a role in building the tool. These videos provide valuable insights and guide you through the creative process, making your journey with TextFX both educational and enjoyable. Lupe Fiasco and TextFX ## Beyond Rap: Who Can Benefit? While TextFX may have its origins in the world of rap, its applications extend far beyond. Wordsmiths of all kinds can find value in this innovative tool. Here are some roles and professions that can benefit from TextFX's creative prowess: - **Poets:** Enhance your poetic expressions with imaginative comparisons and sensory details. - **Copywriters:** Craft compelling copy that stands out with unique word choices and acronyms. - **Songwriters:** Elevate your lyrics with unexpected twists and creative wordplay. - **Business Thinkers:** Foster group thinking and brainstorming sessions with fresh ideas and intersections. - **Content Creators:** Fuel your creativity to produce engaging and memorable content. 💡 Exploring TextFX brings to mind the work of Edward de Bono, a pioneer **in the field of lateral thinking**. Just as de Bono encouraged thinking outside the box to solve problems, TextFX encourages thinking outside the ordinary to create captivating text. ## Dive in In the ever-evolving landscape of content creation, creativity is the key to standing out. [Google's TextFX](https://textfx.withgoogle.com) offers a unique and powerful way to infuse your work with fresh ideas and unconventional thinking. It's a playground for wordsmiths, a tool to stimulate creativity, and a resource that can benefit professionals in various fields. So, what are you waiting for? Dive into TextFX, experiment, and let your imagination run wild. You might just discover new dimensions of your creativity that you never knew existed. It's definitely worth playing around with and checking out. Get ready to take your writing to exciting new heights with [TextFX](https://textfx.withgoogle.com)! ### The tangible side of AI in the cloud URL: https://rob-hoeijmakers.ghost.io/the-tangible-side-of-ai-in-the-cloud/ Last updated: 2024-12-17T13:29:50.000Z The [world of AI](https://rob-hoeijmakers.ghost.io/tag/ai/) and the cloud is a dynamic one, constantly evolving and presenting both challenges and opportunities. Recently, I had the privilege of attending a panel discussion on this very topic at a [Dutch Startup Association (DSA) event](https://www.dutchstartupassociation.nl) in Amsterdam, and the insights gained were valuable. ![The panel with DSA chair Lucien Brum and people from Autoscriber, Vultr and NVIDIA](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/09/IMG_6343.jpeg) The panel with DSA chair Lucien Brum and people from Autoscriber, Vultr and NVIDIA. ## AI's Hunger for Resources AI, a vast landscape encompassing **voice recognition, speech synthesis, image generation, and language models**, is a digital powerhouse. However, its insatiable appetite for energy, bandwidth, and processing power can't be ignored. Ideally, the task of satisfying this hunger falls to GPUs, but CPUs can also play their part, albeit less efficiently. > When it comes to AI applications in the cloud, the choice between GPU and CPU matters. CPUs are versatile and handle a wide range of tasks, while GPUs excel at parallel processing, making them ideal for AI workloads like deep learning. ## The Quest for Instant Gratification In a world where instant results are the norm, having access to robust computing resources is paramount. Unfortunately, this power isn't distributed evenly across the globe. This means that startups seeking to harness AI for real-time applications face a significant challenge in ensuring they have the necessary computational firepower at their disposal. ## The Cost Conundrum Startups, in particular, need to be mindful of their budgets. While the cloud offers incredible advantages, it's not without its indirect costs. As a business scales, these expenses can multiply exponentially, potentially posing a financial hurdle. ## The Vendor Lock-In Dilemma Perhaps the most intriguing lesson from the discussion was the challenge of vendor lock-in. Moving from one cloud provider to another can be a complex and costly endeavor, leading many startups to stick with their existing setup, even if it's not the perfect fit. This decision often includes continuing to rely on specific AI applications. ![Concentration of control for the cloud graph ](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/10/IMG_6786.jpeg) Concentration of control for the cloud ## The Power of Choice Choosing the right cloud provider or partner is therefore a critical decision for startups. It's not just about the services they offer but also about the flexibility to adapt as your business evolves. This choice can significantly impact a startup's trajectory. 💡 Learning on the side: The ****Dutch Startup Association (DSA)** and ****TechLeap** serve distinct roles in the Dutch startup ecosystem. DSA is a grassroots initiative driven by startups themselves, operating on a membership basis. Its primary aim is to represent and support the interests of startups in the Netherlands. On the other hand, TechLeap is a government-funded program dedicated to stimulating economic activities within the startup sector. It provides support, resources, and initiatives to help startups thrive and contribute to the Dutch economy. ## A Program Tailored for Amsterdam Startups For Amsterdam-based startups looking to thrive in the AI space, there's exciting news. Vultr, a cloud provider, is rolling out a Digital Startup Program designed especially for those seeking [AI-composable architectures](https://rob-hoeijmakers.ghost.io/mach-alliance/). If your startup is based in the Netherlands, AI is at the heart of your offering, and you're gearing up for rapid growth, this program is worth exploring. In conclusion, the world of AI and the cloud is both exhilarating and challenging. The right choices in terms of technology and partnerships can make all the difference in a startup's journey. Keep innovating, stay informed, and consider what solutions like [Vultr's Digital Startup Program](https://experience.vultr.com/Amsterdam-Startup-100K.html) can offer to propel your Amsterdam startup to new heights. ### Poe: Swiss Army Knife for Mastering AI Magic URL: https://rob-hoeijmakers.ghost.io/poe-app/ Last updated: 2026-05-26T19:34:34.000Z In the ever-evolving landscape of web technologies and digital marketing, staying ahead of the curve is essential. That's where Poe steps in as your versatile AI-infused chatbot companion, ready to assist and inspire. But with a multitude of features and possibilities, it's natural to feel overwhelmed initially. Fear not; I’ve got you covered. In this comprehensive guide, we'll dive deep into Poe, providing insights, hands-on tips, and real-world examples to help you harness its full potential. ## What is Poe? Poe is **a versatile AI-infused chatbot web application that leverages advanced language models, such as GPT-4 and Claude+**, to provide users with quick and tailored responses to their queries and prompts. Its user-friendly interface allows individuals to interact with multiple Language Models (LLMs) within the same platform. 💡 ****Poe is made by Quora.** [Quora](https://www.quora.com) is a popular platform known for its question-and-answer format, where users can ask questions on various topics and receive answers from the community. It's a valuable resource for gaining knowledge and insights on a wide range of subjects. ## Getting Started with Poe Before we embark on our journey into the world of Poe, let's start with the basics: getting started. ### Account Setup To unlock the capabilities of Poe, begin by signing up for an account. You can choose to use your email, Google, or Apple credentials. This step is your gateway to the AI playground. ### Exploring the Interface Once logged in, take a moment to explore the user-friendly interface. Familiarize yourself with the options available and the chatbot lineup. ![Exploring AI bots, creating one yourself, pick the right LLM.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/09/Poe.jpg) Exploring AI bots, creating one yourself, pick the right LLM. ## Using Multiple Language Models Poe boasts an impressive lineup of Language Models (LLMs) like Claude, ChatGPT, Google PaLM, and Llama. The ability to interact with multiple LLMs within a single interface is a game-changer for web professionals. Here's how you can make the most of this feature: - **Diverse Applications**: Each LLM has its unique strengths. Claude is excellent for structured data analysis, ChatGPT excels in natural language understanding, Google PaLM the 540-billion parameters powerhouse, and Llama can assist with your content. Think of them as tools in your digital toolkit, ready to serve different purposes. - **Experimentation**: Don't hesitate to experiment. Try various LLMs to understand their capabilities better. Start with questions or prompts related to your field, and you'll quickly grasp their strengths and weaknesses. ## Preconfigured AI Bots and Customization One of Poe’s standout features is the ability to use preconfigured AI bots or create your custom bots. Let's explore both options: ### Preconfigured AI Bots Dive into the preconfigured bots available within Poe. These bots are designed for specific tasks, from answering questions to generating content. For instance, you can use them for quick insights into SEO strategies or content creation tips. ### Creating Custom AI Bots Here's where the fun begins. Creating your own custom bot opens up endless possibilities. Think of it as crafting a digital assistant tailored to your needs. - Select the LLM that suits your task. For instance, ChatGPT for content-related queries or Google PaLM for more up to date information. - Craft your prompt: This is the key to getting the desired output. Prepare prompts carefully, keeping your objectives in mind. I played around with it and had great fun with it. And next to that it gave me a quick insight into the power of the different LLM's and how to prepared a good prompt for others to converse with my AI bot. ## Examples of Self-Created Bots To give you a glimpse of the potential, let's look at a few examples: 1. **Conversing with Stephen King**: Imagine creating a bot that emulates the writing style of Stephen King. You can have virtual conversations with this literary legend, gaining insights into storytelling and creativity. 2. **Generating Art with Dall-e**: If you're an art enthusiast, set up a bot that assists in generating prompts for [Dall-e](https://rob-hoeijmakers.ghost.io/playing-with-ai-system-dall-e/). This opens the door to creating stunning visual artworks or graphics for your website. 3. **Exploring Stable Diffusion**: Don't limit yourself to text-based models. With Poe, you can work with models like [Stable Diffusion](https://rob-hoeijmakers.ghost.io/synthetic-images-playgroundai/) to generate high-quality images. Create bots to craft prompts for visual content that stands out. ![Left the Stephen King AI bot, and on the right a bot to prompt Dall-e](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/09/Poe-AI-bots.jpg) Left the Stephen King AI bot, and on the right a bot to prompt Dall-e ## Experiment and Discover As a [web professional](https://rob-hoeijmakers.ghost.io/the-web-strategist-and-the-web-strategy/), your primary goal with Poe might be to gain better insights and experiment with the technology. Poe provides the perfect playground for this purpose. You can explore different LLMs, experiment with prompts, and discover the capabilities of AI in real-time. It's not just about the benefits or immediate applications; it's about diving in, playing around, and understanding the technology's possibilities. ## Review Poe is your gateway to the world of advanced AI language models. Whether you're here to gain insights, experiment with AI, or simply explore the fascinating world of machine-generated content, Poe has something to offer. The power to define the AI's path lies in your hands. So, dive in, experiment, and [let Poe be your guide](https://poe.com) on this exciting journey of discovery. ### Measures versus Metrics URL: https://rob-hoeijmakers.ghost.io/measures-vs-metrics/ Last updated: 2026-04-25T13:16:36.000Z In the fast-paced world of online marketing and [web strategy](https://rob-hoeijmakers.ghost.io/the-web-strategist-and-the-web-strategy/), data-driven decisions are the norm. We regularly track website performance, analyze user behavior, and strive for [continuous improvement](https://rob-hoeijmakers.ghost.io/user-experience-ux-and-ux-design/). Amid this data deluge, two terms often come into play: measures and metrics. What's the difference, and why should you care? Let's demystify these concepts and explore their significance for online professionals like you. ## Measures: The raw building blocks Imagine you're taking attendance in a classroom. You make a list of students present and absent—simple, right? In the realm of data analysis, these lists are our "measures." Measures are the raw, unprocessed data points that provide specific information about a particular aspect of your online operation. For instance, if you're monitoring website traffic, a measure could be the number of visitors your website received in a given month. It's the foundation upon which metrics are built, like bricks in a wall. ## Metrics: Constructing insights Now, let's move from the raw data to meaningful insights. Metrics are the result of processing and analyzing those measures. They provide a structured and quantifiable way to gauge performance, often expressed as numbers or percentages. Metrics transform measures into actionable information. They are the key performance indicators (KPIs) and benchmarks you use to evaluate how well your online efforts are working. For instance, your website's conversion rate, which calculates the percentage of visitors who take a desired action, is a metric. ## Etymology: The Origins of Measures and Metrics Understanding a bit of word history can shed light on their meanings. Both "measure" and "metric" have roots in ancient languages: - "Measure" originates from the Latin "mensura," which means "a measuring, a measurement." It entered Middle English through Old French in the 14th century. - "Metric" has a more recent history, derived from the Greek "metrikos," meaning "relating to measurement." It became prevalent in English during the 19th century. ## Why It Matters for Online Professionals Now that we've defined the terms and explored their etymology, you might wonder why this matters for online professionals. Here's why: 1. **Clarity and Communication:** Understanding the distinction between measures and metrics enables you to communicate more precisely. When discussing your online performance with colleagues or clients, using the right term ensures everyone is on the same page. 2. **Data-Driven Decision-Making:** Metrics are the compass guiding [your digital strategies](https://rob-hoeijmakers.ghost.io/the-web-strategist-and-the-web-strategy/). By knowing which metrics matter most to your objectives, you can make data-driven decisions that lead to better outcomes. 3. **Continuous Improvement:** Metrics are your allies in the pursuit of improvement. They help you identify areas that need attention, optimise campaigns, and achieve your online goals more effectively. In conclusion, while the terms "measure" and "metric" may seem interchangeable at first glance, they serve distinct purposes in the world of online professionals. Measures are the raw data points, while metrics are the calculated values that provide insights. By understanding their roles and origins, you'll not only enhance your communication but also empower your data-driven decision-making. Next time you dive into your [website analytics or online marketing reports](https://rob-hoeijmakers.ghost.io/understanding-website-metrics-statistics-analytics-and-insights/), remember the difference between these two essential concepts. It's a small step toward greater clarity and effectiveness in the online world. ### Meaning of Metrics, Statistics, Analytics, and Insights URL: https://rob-hoeijmakers.ghost.io/understanding-website-metrics-statistics-analytics-and-insights/ Last updated: 2026-05-10T08:54:16.000Z I've been in the digital sphere for quite some time, and just when I think I've got it all figured out, I find myself tangled in a web of overlapping terminologies. [Metrics](https://rob-hoeijmakers.ghost.io/measures-vs-metrics/), statistics, analytics, insights—what do these words really mean, and how are they different? If you've found yourself scratching your head over these terms, you're not alone. After some thorough research and head-scratching of my own, I've finally untangled the web. Here's what I've discovered. ## Website Metrics: The Raw Numbers Let's start with the basics. Website metrics are the individual, quantifiable pieces of data that provide a snapshot of your website's performance. Think of metrics as the building blocks that form the foundation of your website's analytics. Pageviews, bounce rate, and click-through rates are all examples of website metrics. They are the "what" in your data story—the raw numbers that indicate performance but don't offer much in the way of interpretation. ### Key Takeaway - Metrics are raw, individual data points. ## Website Statistics: Metrics in Context Website statistics take these raw metrics and place them into a more understandable context. For instance, saying your website had 10,000 visits last month is a metric. But stating that your website traffic increased by 10% compared to the previous month is a statistic. Statistics often involve an element of comparison, trend identification, or other forms of interpretation that lend meaning to the raw metrics. ### Key Takeaway - Statistics are metrics that have been contextualized or analyzed. ## Website Analytics: The Bigger Picture This is where things get interesting. [Website analytics](https://rob-hoeijmakers.ghost.io/cookie-free-analytics/) is the umbrella term that encompasses the collection, measurement, analysis, and interpretation of website metrics and statistics. It's the process of taking those numbers and statistics and turning them into actionable insights. Analytics delve into the "why" and "how" behind the numbers, providing a roadmap for decision-making and strategy. ### Key Takeaway - Analytics involves a comprehensive understanding and interpretation of metrics and statistics. 💡 ****Understanding KPIs in** [****Web Strategy**](https://rob-hoeijmakers.ghost.io/the-web-strategist-and-the-web-strategy/) Key Performance Indicators (KPIs) belong within the realm of ****Analytics**. They are specific, quantifiable metrics that align with strategic objectives, helping to measure and evaluate the effectiveness of digital strategies. KPIs transform raw data into meaningful insights, enabling data-driven decisions. For example, a KPI such as conversion rate directly indicates how well your website converts visitors into customers, guiding improvements and optimizations. ## Website Insights: The Actionable Recommendations Lastly, we have [website insights](https://rob-hoeijmakers.ghost.io/google-search-console-gsc/). Insights are the golden nuggets of wisdom you obtain after thorough analytics. They are the actionable recommendations that you can implement to improve your website's performance. While metrics tell you 'what,' statistics provide 'how much,' analytics answer 'why,' insights tell you 'what to do next.' ### Key Takeaway - Insights are actionable recommendations derived from analytics. ## My personal conclusion Navigating the maze of website data terminology can be confusing, but understanding the differences can significantly impact your digital strategy. **Metrics** give you raw numbers, **statistics** put those numbers into context, **analytics** help you understand what those numbers mean, and **insights** tell you what actions to take. As someone deeply involved in content realization, online marketing, and marketing technology, understanding these terms has given me the clarity I needed to refine my strategies and help others in the digital space. So, the next time you find yourself lost in the jargon, just remember: Metrics, Statistics, Analytics, and Insights are not just words; they are the compass that guides your digital journey. --- ### Further reading - [Google Search Console, a wealth of insights](https://rob-hoeijmakers.ghost.io/google-search-console-gsc/) - [Web analytics made simple by using Plausible](https://rob-hoeijmakers.ghost.io/cookie-free-analytics/) ### WordPress versus Ghost: A comparison for bloggers URL: https://rob-hoeijmakers.ghost.io/wordpress-vs-ghost-for-bloggers/ Last updated: 2026-05-10T08:54:16.000Z In the ever-evolving world of web strategy and content realization, choosing the right blogging platform is key. As [web strategist](https://rob-hoeijmakers.ghost.io/the-web-strategist-and-the-web-strategy/) and blogger, I've had the privilege of delving deep into two prominent solutions: [WordPress](https://rob-hoeijmakers.ghost.io/how-about-wordpress/) and [Ghost](https://rob-hoeijmakers.ghost.io/ghost-blog/). In this article, I'll share my insights and experiences to help you make an informed decision for your blogging journey. ## Overview of WordPress and Ghost WordPress and Ghost are both robust blogging platforms with sizable user bases. WordPress is a versatile giant, powering millions of websites worldwide, while Ghost, a relative newcomer, has gained popularity for its simplicity and focus on publishing. Each platform caters to a unique audience, and it's essential to understand their strengths and weaknesses. ![Ghost interfaces are clean and focussed on writing with our worries.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/09/Ghost-interface.jpg) Ghost interfaces are clean and focussed on writing with our worries. ## Ease of setup and maintenance Setting up a WordPress site offers extensive possibilities but can be a complex journey. During my WordPress experience, I grappled with countless plugins, security concerns, and even the need to collaborate with my hosting provider to implement essential features like Varnish, a reverse proxy. In contrast, Ghost Pro provided a refreshing experience. It arrived with many essential features out of the box, eliminating the need for a plugin overload. The simplicity of Ghost's setup allowed me to focus on content rather than technical intricacies. Moreover, Ghost Pro's automatic maintenance and top-notch support were a game-changer, saving precious time and ensuring my blog stayed up-to-date and secure. 💡 Ghost was founded in April 2013, after a very successful [Kickstarter campaign](https://www.kickstarter.com/projects/johnonolan/ghost-just-a-blogging-platform/description) to create a new platform focused solely on professional publishing. Ghost's mission is to create the best open source tools for independent journalists and writers across the world, and have a real impact on the future of online media. ## Feature comparison Are you eager to uncover the inner workings of your blog platform? In this section, we delve into the core aspects that define your blogging experience: content management, customization, and additional features. From the powerful content management system of WordPress to the intuitive simplicity of Ghost, we explore the merits of each. ### Content Management WordPress boasts a powerful [content management system](https://rob-hoeijmakers.ghost.io/content-management-system-cms/) with a wide range of plugins and themes. Ghost, while simpler, provides an intuitive editor and smooth media handling. Both have their merits. ### Customization WordPress's extensive theme library and plugin marketplace offer unmatched customization options. Ghost, while more streamlined, still allows for personalized themes and integrations. ### Plugins/Extensions WordPress shines in this department, but the sheer number of available plugins can be overwhelming. Ghost's simplicity is a plus, but it may lack some specialized plugins. ### SEO Capabilities When it comes to [SEO (Search Engine Optimization)](https://rob-hoeijmakers.ghost.io/my-basis-for-seo/), Ghost shines right from the start. Out of the box, Ghost offers robust SEO capabilities. It provides clean and optimized HTML markup, structured data, and essential meta tags, all designed to make your content search engine-friendly. Ghost's simplicity extends to its SEO features, ensuring that you're on the right track without the need for extensive plugin installations and configurations. On the other hand, WordPress, while highly customizable, often requires additional plugins and meticulous configuration to achieve similar SEO capabilities. With an extensive library of SEO plugins available, WordPress users have plenty of options, but navigating this sea of plugins can be daunting. While these plugins can be powerful, they also add complexity to your setup and may require ongoing maintenance. ### Localization and multilingual support WordPress, with its vast user base, offers robust localization and multilingual support. Ghost is more English-centric and may require additional effort for localization. ### Newsletter Integration When it comes to [newsletter integration](https://rob-hoeijmakers.ghost.io/web-push-vs-email-newsletters/), Ghost takes the lead with its integrated and user-friendly features. Ghost's built-in newsletter functionality allows you to effortlessly manage and send newsletters to your subscribers directly from the platform. This seamless integration simplifies the process of connecting with your audience through email, making it an attractive choice for bloggers who prioritize newsletter marketing. On the other hand, WordPress, while highly versatile, often relies on third-party extensions or plugins for comprehensive newsletter management. To set up newsletters in WordPress, you may need to search for and install the right plugin, configure it to your needs, and potentially deal with compatibility and maintenance issues. While WordPress offers many excellent newsletter plugins, the need for additional installations can add complexity to your blogging setup. ![Ghost newsletter email](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/09/Ghost-email.jpg) Ghost newsletter email ### Internal Linking Assistance When it comes to internal linking, WordPress excels with built-in features and plugins that make it easy to create and manage internal links within your content. WordPress offers a user-friendly editor that allows you to add links to your posts and pages effortlessly. You can specify attributes such as "nofollow" and "open in a new tab" directly from the editor, giving you precise control over your internal links. ✅ ****UPDATE**: As of this writing, June 2024, Ghost has [linking assistance available](https://ghost.org/changelog/internal-linking/). And the team did a great job! In contrast, Ghost could benefit from similar tools to assist with internal linking. While Ghost provides a clean and straightforward writing experience, it currently lacks some of the advanced internal linking features found in WordPress. Bloggers who prioritize internal linking and want fine-grained control over link attributes may find WordPress more accommodating in this regard. ![Internal linking in wordpress](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/09/internal-linking.jpg) Internal linking in wordpress ### Additional Features WordPress offers an abundance of features, including e-commerce capabilities and community building tools. Ghost focuses more on blogging essentials. What is better in WordPress is the [Site Search](https://rob-hoeijmakers.ghost.io/website-search/) by Jetpack. A better UX and a deeper, more intelligent search with facets etc. Ghost uses a script that searches through headlines and excerpts but not the full text. Another feature of WordPress, also a plugin, that is really good is a plugin that handles related articles. Vital for bringing relevant content to the surface (and improve bounce rates). 💡 ****Streamline Your Writing Workflow with Ulysses** Both Ghost and WordPress offer the flexibility you crave when it comes to choosing your writing tools. If you're a fan of the streamlined and [distraction-free writing experience that Ulysses provides](https://www.hoeijmakers.net/blog-writing-workflow/), you're in luck! Ulysses seamlessly integrates with both platforms, allowing you to craft your content in the familiar, writer-friendly environment you love. So whether you prefer Ghost's intuitive simplicity or WordPress's robust features, your writing process can remain consistent and efficient with Ulysses by your side. ## Time and skills investment My experience with WordPress required a considerable time investment. To optimize and maintain my blog, I had to delve into web technologies, including schema markup, sitemaps, canonical URLs, and more. While this was enjoyable for me as a web strategist, it might not be everyone's cup of tea. ## Support and updates Ghost Pro's hands-off approach to updates and responsive support greatly impressed me. WordPress, on the other hand, demands manual updates and plugin management, which can be a hassle. ## Verdict In my journey of comparing WordPress and Ghost, [Ghost Pro](https://rob-hoeijmakers.ghost.io/self-host-ghost-blog/) emerged as my preferred choice. Its simplicity, out-of-the-box features, and top-notch support aligned perfectly with my blogging goals. However, the decision between these platforms depends on your unique needs and preferences. Remember that both Ghost and WordPress are **open source and offer reasonable import and export tools**, providing you with the flexibility to switch platforms when necessary. Evaluate your needs carefully to determine if the move to Ghost is the right step for your blogging journey. ### *Author bio and contact information* *I'm* [*Rob, an Amsterdam-based web strategist*](https://rob-hoeijmakers.ghost.io/about/) *with a passion for blogging. I also compared* [*Wordpress to Squarespace*](https://rob-hoeijmakers.ghost.io/squarespace-vs-wordpress/)*.* --- ### Further reading - [Squarespace versus WordPress](https://rob-hoeijmakers.ghost.io/squarespace-vs-wordpress/) ### Structured Content URL: https://rob-hoeijmakers.ghost.io/structured-content/ Last updated: 2026-05-26T19:35:19.000Z When it comes to web development, one crucial concept that often gets overlooked is structured content. At first glance, it might seem like a dull and abstract topic, but think of it as the mosaic tiles of a website. Just like assembling colorful mosaic tiles to form intricate patterns, structured content allows you to piece together content components strategically to create a [compelling digital experience](https://rob-hoeijmakers.ghost.io/user-experience-ux-and-ux-design/). Let's dive into this topic using the mosaic analogy, and along the way, we'll explore what structured content means, how it differs from structured data, and why it's the backbone of a successful online presence. ## The Mosaic Analogy Imagine yourself with a set of colorful mosaic tiles, each tile featuring unique geometric shapes and colors. These mosaic tiles represent structured content in web development. Like a skilled mosaic artist, you can puzzle these tiles together to form intricate and beautiful patterns. This mirrors the flexibility and creativity that structured content brings to web design and content creation. Just as mosaic tiles can be arranged to create unique art pieces, structured content enables web professionals to craft various content layouts, templates, and designs for different channels and platforms. These content components serve as reusable building blocks, much like multichannel reuses in structured content. ![Wooden mosaic tiles where the pieces can give endless variations.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/09/houten-mozaiek-puzzel-99169400.jpg) Wooden mosaic tiles where the pieces can give endless variations. ## Understanding Structured Content Structured content is the practice of organizing and presenting content in a consistent, standardized format. It ensures that content remains coherent and adaptable across various digital channels and devices. Structured content embodies the COPE (Create Once, Publish Everywhere) principle, allowing content creators to craft content once and seamlessly adapt it for diverse channels, both now and in the future. ![Structured content at work in Ulysses, the tool I use to write and produce content.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/09/Ulysses-structured-content.jpeg) Structured content at work in Ulysses, the tool I use to write and produce content. ### Structured Content vs. Structured Data It's essential to distinguish structured content from [structured data](https://rob-hoeijmakers.ghost.io/semantic-web/). While structured content focuses on the organization and presentation of content components for human consumption, structured data primarily serves search engines and machine interpretation. Structured data, often in the form of schemas and markup, helps search engines understand the context and meaning of web content, enhancing search results with rich snippets and knowledge graph entries. Structured content, on the other hand, caters to the human experience. It ensures that the content you create is not only informative but also visually appealing and consistent across different touchpoints, from websites to social media. ## The Benefits of Structured Content Now that we've established the mosaic analogy and clarified structured content let's explore why it's so essential in web development: ### Consistency and Efficiency Structured content guarantees that your content remains consistent across all channels and platforms, much like how mosaic tiles maintain their geometric patterns. Your brand message and style stay uniform, whether it's on your website, social media, or email campaigns. This consistency extends to content presentation, ensuring a seamless user experience. ### Improved SEO and Site Performance Structured content improves your website's [SEO performance](https://rob-hoeijmakers.ghost.io/my-basis-for-seo/) by making it more understandable to search engines. Metadata and taxonomy provide clear signals about content relevance and context. This can boost your SEO rankings and increase your website's visibility in search results. Additionally, structured content enhances site performance by increasing click-through rates (CTR) and potentially earning rich snippets and featured snippets on search engines. These visual enhancements make your content more eye-catching and trustworthy. ### Content Reusability and Scalability Structured content allows for content reusability, just as you can repurpose mosaic tiles for different patterns. Content components become building blocks that can be reused across your website or even in different projects. This not only saves time but also ensures content consistency and reduces the risk of errors. As your web presence grows, structured content makes scalability effortless. You can add new content components without disrupting the existing structure, ensuring a seamless user experience and adaptability to emerging channels. ### Benefits for AI and Machine Learning Structured content plays a pivotal role in enabling artificial intelligence (AI) and machine learning. It provides the semantic meaning needed for computers to interpret content accurately and connect it to related content. With structured content, machines can learn, identify patterns, and make decisions with minimal human intervention, accelerating the growth of AI and machine learning. ## Conclusion In conclusion, structured content is not just an abstract concept; it's the practical foundation that can take your web development projects to new heights. Much like assembling mosaic tiles to create intricate patterns, structured content allows you to piece together content components strategically, ensuring consistency, adaptability, and visual appeal. As the digital landscape evolves, structured content remains the key to staying relevant, efficient, and effective in the world of web development. ### Content Activation URL: https://rob-hoeijmakers.ghost.io/content-activation/ Last updated: 2026-04-25T13:48:47.000Z In the vast digital landscape of today's enterprises, valuable content often lies hidden in dark corners, reminiscent of warehouses where you need forklifts to retrieve something. However, with the advent of content activation, a new era of content management is dawning, offering a SaaS-level solution that seamlessly fuses content from different sources, including images, videos, text, and documents. 💡 ****SaaS** stands for ****Software as a Service**. It's a cloud computing model where software applications are hosted and provided to users over the internet on a subscription basis. Instead of users installing and maintaining software on their local devices or servers, they access the software via a web browser. ## The Challenge of Content Silos Enterprises typically accumulate an immense amount of content over time, stored in various [Digital Asset Management (DAM) systems](https://rob-hoeijmakers.ghost.io/digital-asset-manager-dam/), cloud storage platforms like Google Drive, Sharepoint, or Box, and sometimes scattered across different departments. This dispersed content leads to inefficiencies, wasted resources, and missed opportunities for content utilization. ## Content Activation: Shedding Light on the Dark Corners Content activation is the answer to this problem. It's a strategy that involves creating a unified solution to access, manage, and distribute content from diverse sources seamlessly. One such SaaS-level solution, smint.io, has emerged as a game-changer in this field. Recently, I had the privilege of taking a guided tour of this innovative platform with [Gert Fahrnberger](https://www.linkedin.com/in/gertfahrnberger/). > Use Smint.io to activate content silos and enhance the content experience between brands, teams, employees, agencies, investors, channel partners and journalists. ### A Guided Tour of smint.io Gert Fahrnberger walked me through the functionalities of smint.io and showcased how it can transform [content management](https://rob-hoeijmakers.ghost.io/content-management-system-cms/). Here are some key takeaways from our tour: 1. **Fusion of Content:** Smint.io brings together content from various sources, including [DAMs](https://rob-hoeijmakers.ghost.io/digital-asset-manager-dam/) like Tenovos, Cloudinary, Picture Park, Bynder, and cloud storage platforms, making it easily accessible in one centralized hub. 2. **Design Integration:** The platform allows for seamless integration of designs, ensuring that your content [looks visually appealing](https://rob-hoeijmakers.ghost.io/user-experience-ux-and-ux-design/) and aligns with your brand's aesthetics. 3. **Enhanced Search Capabilities:** A [robust search](https://rob-hoeijmakers.ghost.io/website-search/) feature simplifies the process of locating specific pieces of content, eliminating the need to sift through endless digital repositories. 4. **Diverse Use Cases:** Smint.io caters to a range of use cases, including [press portals, media centers, e-learning platforms, brand portals](https://rob-hoeijmakers.ghost.io/unlocking-digital-asset-managements-versatility/), and [content hubs](https://rob-hoeijmakers.ghost.io/exploring-specialised-content-hub-platforms-and-vendors/), making it versatile for different industry needs. ### A Real-World Example: Bystronic To illustrate the power of content activation in action, let's look at a real-world success story. [Bystronic](https://www.bystronic.com), a Swiss machine manufacturer, embraced content activation to bring seven different use cases to life for their online presence. This demonstrates the adaptability and effectiveness of content activation in addressing diverse business needs. It needs another article to take a deep dive into their solutions. ![Bystronic website](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/09/DraggedImage.png) Bystronic website ## Conclusion In today's digital age, where content is king, it's essential to have a robust strategy for managing and utilizing your valuable assets efficiently. Content activation, exemplified by solutions like [smint.io](https://www.smint.io), is the beacon of light that dispels the darkness of content silos, providing a unified, user-friendly approach to content management. As enterprises continue to recognize the importance of content activation, we can expect more success stories like Bystronic, where businesses harness the true potential of their content to drive growth, engage audiences, and stay ahead in the competitive landscape. It's time to embrace the content activation revolution and unlock the hidden treasures within your digital warehouses. [You can reach out to me if you want to learn more.](https://rob-hoeijmakers.ghost.io/contact/) 🚀 ****Content Activation** is a strategic process that involves optimizing, organizing, and distributing digital content from various sources to maximize its impact, engagement, and usability. This approach ensures that content is readily accessible, searchable, and aligned with specific business goals, making it a valuable asset for driving audience engagement, brand awareness, and desired outcomes. Content activation often utilizes technology solutions to streamline content management and distribution across digital platforms. ### Typography insights URL: https://rob-hoeijmakers.ghost.io/typography-insights/ Last updated: 2026-05-10T08:54:16.000Z If you're anything like me, you've come to realize that fonts are more than just a backdrop for words—they're the unsung heroes of visual communication. In this post, I'll be sharing some exciting insights I've gained about fonts, from their role in web design to the recent Microsoft Office font switch. ## The Art of Typography in Web Design As a web strategist, I'm constantly exploring ways to enhance [user experiences](https://rob-hoeijmakers.ghost.io/user-experience-ux-and-ux-design/) online. One crucial aspect that often takes the spotlight is typography. Fonts can influence how users perceive a website, affecting readability, mood, and even trust. Here are a few key takeaways of my own learnings that might reshape your approach to web fonts: ### Webfonts: Google Hosting vs. Local The choice between hosting fonts via Google Fonts or using local hosting has more implications than you might think. While Google Fonts offer convenience and a wide array of options, local hosting can provide better performance and control over loading times. The decision ultimately hinges on finding the right balance for your website's needs. 🔤 I've implemented the 'Just Any Font' plugin on my [WordPress](https://www.hoeijmakers.net/how-about-wordpress/) site, mirroring the default behavior in [Ghost](https://www.hoeijmakers.net/ghost-blog/). With this clever tool, I ensure that all my web fonts are loaded from the same domain name. This not only contributes to faster page load times, enhancing the overall user experience, but also prioritizes user data privacy. ### Variable Web Fonts: Embracing Flexibility Variable fonts are a game-changer, allowing a single font file to offer multiple weights, styles, and even widths. This not only reduces page load times but also offers creative freedom for designers. Imagine a font that adapts harmoniously to various screen sizes and contexts! 🔤 A ****variable font** is a type of font format that allows for a single font file to encompass a wide range of styles and variations. Unlike traditional fonts, where each style (such as bold, italic, light, etc.) requires a separate font file, a variable font combines all these styles into one file. This is achieved through the use of adjustable design axes, which are like sliders that control various attributes of the font, such as weight, width, and slant. > Decagram [pic.twitter.com/4nqF3yXtEi](https://t.co/4nqF3yXtEi) > > — Emtype Foundry (@emtype) [December 16, 2022](https://twitter.com/emtype/status/1603698924023422976?ref%5Fsrc=twsrc%5Etfw) Variable font demonstrated ### Fixed vs. Swap: A Delicate Balance Fonts can be set as "fixed" or "swap" depending on how crucial they are for your content's legibility. Fixed fonts load first, maintaining the design's integrity, while swap fonts load later, ensuring content remains readable even during loading. It's all about achieving a balance between aesthetics and functionality. ## Microsoft's Font change: Calibri to Aptos Recently, news broke that Microsoft Office is bidding farewell to the long-serving Calibri font and welcoming the new Aptos as its default typeface. This change has already stirred discussions and anticipation within the design community. While Aptos may be the fresh face, we can't ignore its potential lineage to the timeless Helvetica font. This transition is a reminder that even established brands acknowledge the importance of staying relevant and embracing change. ![Aptos, formerly known as Bierstadt](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/09/Aptos.jpg) Aptos, formerly known as Bierstadt ## The face of communication At the heart of all these font discussions lies a fundamental truth: fonts are more than just visual elements. They're vehicles for emotions, ideas, and messages. A font choice can evoke nostalgia, establish credibility, or create a sense of modernity. It can bridge cultures, convey urgency, and even spark curiosity. Fonts are, in essence, the voice of your content—a powerful communicator that resonates with your audience. I'm reminded that fonts are more than just design choices; they're integral to the stories we tell and the connections we make. So next time you're choosing [a font for your website, blog, or project](https://rob-hoeijmakers.ghost.io/typeface-to-serif-or-not-to-serif/), remember the vast impact it can have on how your message is received. 🖥️ A nice place to [play around with variable fonts](https://design.ubuntu.com/font) is Ubuntu. They have there own variable, [sans-serif](https://www.hoeijmakers.net/typeface-to-serif-or-not-to-serif/) and [mono-spaced font](https://www.hoeijmakers.net/single-spaced-fonts/) which is used in the Operating System and in their marketing communications like the website. The font is aptly called Ubuntu. ### Further reading - [Typography, in motion](https://rob-hoeijmakers.ghost.io/typography-in-motion/) ### Webpages require strong headings URL: https://rob-hoeijmakers.ghost.io/html-headings/ Last updated: 2026-04-25T13:16:05.000Z In the ever-evolving landscape of online content, the synergy between [search engine optimisation (SEO)](https://rob-hoeijmakers.ghost.io/my-basis-for-seo/) and [user experience](https://rob-hoeijmakers.ghost.io/user-experience-ux-and-ux-design/) is paramount. A fundamental yet often underestimated element in this balance is the strategic use of headings (or headers) within your web documents. Content can literally fall flat without headings. By ensuring that your headings are not just visually appealing, but also well-structured and optimised, you can significantly enhance both SEO rankings and user engagement. ## Introducing HeadingsMap Despite the fact that headings define the supporting structure of a web page, they are actually surprisingly hard to see. Design and structure are only loosely coupled. To streamline your heading optimisation efforts, I am happy to share the "HeadingsMap" Chrome browser extension. This simple yet powerful tool provides a clear outline for any web page. With colour-coded indicators and an interactive interface, HeadingsMap enables you to quickly identify any issues with your heading structure. ![HeadingsMap showing the outline and the headings.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/08/HeadingsMap.jpg) HeadingsMap showing the outline and the headings. ### Using HeadingsMap: A Step-by-Step Guide 1. **Install the Extension**: Search for "[HeadingsMap](https://chromewebstore.google.com/detail/headingsmap/flbjommegcjonpdmenkdiocclhjacmbi?hl=en&pli=1)" in the Chrome Web Store and install the extension. 2. **Navigate to Your Web Page**: Open a web page that you want to assess. 3. **Activate HeadingsMap**: Click on the HeadingsMap icon in your browser's toolbar to activate the tool. 4. **Visualise the Hierarchy**: As you browse the page, HeadingsMap displays a visual hierarchy of headings on the right side of your screen. Different levels of headings are colour-coded for easy identification. 5. **Spotting Issues**: Easily identify missing or incorrectly ordered headings, allowing you to make immediate adjustments. With HeadingsMap, you can now embark on a journey of enhanced heading optimisation, ensuring that your content not only captivates your readers but also resonates with search engines. With this tool I can quickly verify structure and it becomes easier to work with web builders . Those can be a little opaque when it comes down to headings. You might pull in visual elements, like a read more block or newsletter promo, and things mess up quickly. The tool gives you a better view on what is happening behind the curtains and straighten the structure out again. [HeadingsMap - Chrome Web StoreTo show, browse and audit (for accessibility and SEO) the headings structure![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/icon_144px.png)Chrome Web Store![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/09q9d8b_KTCkAiuj7VFpyPK25oQPnOffPhY6iclGMog5f-g9kB3IcjDNdi7TjGPNHGzwxgp5o1TOH35uW8Pk7RGE-s128-rj-sc0x00ffffff)](https://chromewebstore.google.com/detail/headingsmap/flbjommegcjonpdmenkdiocclhjacmbi?hl=en&pli=1) ## Crafting effective headings Now that you have the powerful tool of HeadingsMap at your disposal, let’s go through three top tips for crafting headings that maximise both SEO and usability: ### Tip 1: Clear hierarchy for improved navigation Effective headings act as signposts, guiding readers through your content seamlessly. Establish a clear hierarchy by utilising HTML heading tags (H1, H2, H3, etc.) in the proper order. This helps both users and search engines understand the importance and relevance of each section. ### Tip 2: Concise and informative text Craft headings that succinctly convey the essence of each section. Use keywords that resonate with your content's theme, as these keywords can boost SEO rankings. However, avoid keyword stuffing—keep your headings concise while maintaining their ability to capture the essence of the content. ### Tip 3: Reflecting content structure Headings should accurately reflect the structure of your content. Each major idea or subtopic should have a corresponding heading, allowing readers to anticipate what's ahead. This not only aids comprehension but also reinforces the theme for search engines. ## Order and overview If you are like me you want some order and grip on content. And even more if this leads to better SEO and Usability. I never fully grasped titles, headers and headings, but with the extension I get visual feedback on what the structure is and how it is worded. I was able to spot a number of errors (after being warned by an Ahrefs report) in the structure of pages and corrected it very quickly. 😳 ****Head (** **` `** **):** The `` element is not usually visible on the web page itself. It's a container for metadata and other head elements, such as the title of the page, links to stylesheets or scripts, and other information that helps browsers and search engines understand and render the page correctly. It's like the behind-the-scenes [metadata](https://webstrategiesblog.com/metadata/) for the webpage.****Header (** **`
`** **):** The `
` element is a visible part of the webpage, usually found at the top of a section or the top of the entire page. It often contains branding elements like logos, navigation menus, and introductory text. The `
` is a way to present important introductory content to your visitors.****Heading (** **`

`** **,** **`

`** **,** **`

`** **, etc.):** Headings are used to structure and organize the content within the body of a webpage. They range from `

` (the highest level, often used for main titles) to `

` (the lowest level, typically used for subheadings). These elements help both users and search engines understand the hierarchy of your content. The topic of this article. *Also check out* [*my (well structured) article on topic clusters and pillar pages*](https://rob-hoeijmakers.ghost.io/the-tower-of-seo-success-pillar-pages/) *to see how all content matches together.* 😵 Readers scan texts and one of the patterns that matches headers is called Layer cake pattern. Interesting article by Rita Kind-Envy called [People Don’t Read Online—They Scan. This Is How to Write for Them](https://uxplanet.org/people-dont-read-online-they-scan-this-is-how-to-write-for-them-80a75069c14e). ### Identity unmasked URL: https://rob-hoeijmakers.ghost.io/identity-unmasked/ Last updated: 2026-05-10T08:54:16.000Z In the digital age, the word "identity" has taken on new dimensions, but its roots reach back to ancient times. From Latin "idem," meaning "the same," identity has always been about sameness and continuity. Today, we find ourselves navigating two distinct realms of identity: administrative and online. And with the advent of face recognition technology, these two worlds are converging in unexpected ways. ## The Etymology of Identity – A Journey Through Time The word "identity" comes from the Latin "identitas," derived from "idem." It's a concept that has evolved through the ages, reflecting our need to define and understand ourselves in relation to others. In the administrative world, identity is a formal, legal construct. It's your name on a birth certificate, your face on a driver's license, your signature on a contract. It's tangible, concrete, and governed by laws and regulations. In the online world, identity is more fluid and self-curated. It's your profile picture on Facebook, your witty bio on X (Twitter), your curated feed on Instagram. It's a digital expression of who you are—or who you want to be. ![X Profile](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/08/2023-08-24_21-24-22.jpg) X Profile ## Face recognition – Bridging two worlds With the rise of face recognition technology, these two realms of identity are beginning to intersect. Your administrative identity, captured in official documents, can now be linked to your online identity, scattered across social media platforms. This convergence has opened new doors for convenience and security but also raised serious questions about privacy, consent, and accuracy. > Rather, one way to think about this might be that AI makes practical at a massive scale things that were previously only possible on a small scale. This might be the difference between the police carrying wanted pictures in their pockets and the police putting face recognition cameras on every street corner - a difference in scale can be a difference in principle. [Benedict Evans](https://www.ben-evans.com/benedictevans/2023/8/27/generative-ai-ad-intellectual-property). ## A cautionary tale A recent incident brings these concerns to life. During a demonstration, photographs were taken, capturing the faces of attendees. Police then used face recognition technology to search social media profiles, matching faces with online identities. Warnings were sent to the matches, leading to confusion, fear, and controversy. Was this an innovative use of technology to maintain order, or an overreach that infringed on personal freedoms? The incident sparked a heated debate, highlighting the complex interplay between administrative identity, online identity, and emerging technology. [Identity Wallet: A Tool for Empowerment or a Mechanism of Control?Freedom or surveillance? Digital IDs offer convenience but come with risks. Are they tools of empowerment or state control over your personal data?![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-24.png)Rob Hoeijmakers (robhoeij)Rob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_8641-3.jpeg)](https://rob-hoeijmakers.ghost.io/digital-identity-id-empowerment-or-state-control/) ## Conclusion: Navigating the Identity Landscape The tale of "identity" is a rich and evolving narrative, stretching from ancient Latin roots to cutting-edge technology. As we continue to explore and define ourselves in both administrative and online realms, we must be mindful of the connections and collisions between these worlds. The story of the demonstration serves as a reminder that our digital footprints are more than mere pixels on a screen. They are part of our broader identity, intertwined with our legal and social selves. As we embrace the conveniences of modern technology, let's also embrace the responsibility to use it wisely, respecting the delicate balance between innovation and privacy, between being the same in two worlds yet unique in our own right. ### Further reading - [Understanding Digital Identity in the EU: eID, eIDAS, and the EUDI Wallet](https://rob-hoeijmakers.ghost.io/eu-eid-eidas-eudi-wallet/) ### Account, profile and identity URL: https://rob-hoeijmakers.ghost.io/accounts-profiles-identity/ Last updated: 2026-05-13T15:19:02.000Z In the bustling marketplace of the internet, two terms often float around: "account" and "profile." They may seem interchangeable, but they serve different roles in our online interactions. Add to that the process of creating an account and then proving you're the rightful owner, and things can get a bit complex. Let's break it down. ## Accounts vs. Profiles Ever wondered about the difference between accounts and profiles? ### Your Account: The Digital Passport An account is like your digital passport. It's the key that unlocks the door to a specific online service, be it email, social media, or online banking. As free user or a paid subscriber. When you create an account, you're establishing your credentials—usually a [username and password](https://rob-hoeijmakers.ghost.io/passwords-passkeys/)—that grant you access. Think of your account as a membership card. It says, "I belong here, and I have the right to access these services." ### Your Profile: A Digital expression A profile, on the other hand, is like a digital expression. It's the way you present yourself within the service you've accessed through your account. Your profile might include your profile picture, bio, interests, and other personal details. Think of your profile as the outfit you wear once inside a club. It says, "This is who I am, and this is how I want to be seen." A profile also resembles [the business card](https://rob-hoeijmakers.ghost.io/business-cards-in-a-digital-world/). You want to make a favourable expression of who you are and where to be found. ![Account versus profile](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/08/Account-and-Profile.jpg) Account versus profile ## Establishing an account vs. proving yourself I will be discussing the essential duo: setting up an account and proving that you're the rightful user. Let's break it down step by step. ### Establishing an account: The Sign-Up Creating an account is like signing up for a membership. You provide your credentials (username and password) and often some basic personal information. This process establishes your right to access the service. It's like getting your membership card. You've shown you want in, and you've been given the key. ### Proving yourself: The Log-In Logging in is the act of proving that you're the rightful owner of the account. You enter your username and password (or [Passkey](https://rob-hoeijmakers.ghost.io/passwords-passkeys/)), and sometimes additional verification like a code sent to your phone ([two-factor authentication](https://rob-hoeijmakers.ghost.io/two-factor-authentication-2fa/)). It's like showing your membership card at the door. You're proving that you're the same person who signed up, and you have the right to enter. ![Login versus Registration, @CodingLab](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/08/image-4.png) Login versus Registration, @[CodingLab](https://www.youtube.com/@CodingLabYT) ## Navigating the online world with confidence Understanding the distinction between an account and a profile, and the difference between establishing an account and proving yourself, helps you navigate the online world with confidence. Your account is your passport, your key to access. Your profile is your persona, your way of expressing yourself. And the process of creating and then verifying your account is a digital handshake, a way of introducing yourself and then proving you're the real deal. In this bustling marketplace of the internet, knowing these distinctions empowers you to engage with online services more effectively and securely. It's all part of the rich tapestry of our digital lives, where accounts, profiles, and log-ins are the threads that weave us into the online world. ### Further reading - [How to get verified on LinkedIn](https://rob-hoeijmakers.ghost.io/how-to-get-verified-on-linkedin/) - [LinkedIn Verification: When Identity Becomes a Profile](https://rob-hoeijmakers.ghost.io/linkedin-verification-when-identity-becomes-a-profile/) ### CMS, Content Platform, or DXP? URL: https://rob-hoeijmakers.ghost.io/cms-content-platform-or-dxp/ Last updated: 2026-05-10T08:54:17.000Z In the rapidly evolving landscape of digital experiences, finding the right tools to manage, optimize, and deliver content is key. Whether you're a blogger, a marketer, or an enterprise, understanding the differences between a [Content Management System (CMS)](https://rob-hoeijmakers.ghost.io/content-management-system-cms/), a [Content Platform](https://rob-hoeijmakers.ghost.io/content-platforms/), and a [Digital Experience Platform (DXP)](https://rob-hoeijmakers.ghost.io/dxp/) is crucial for crafting a successful online presence. In this article, I will explore these concepts, highlight their unique features, and provide real-world examples to help you make an informed decision for your digital strategy. ## CMS: Content Management System A CMS is the foundation of content creation and publishing. It empowers individuals and small businesses to manage their digital content seamlessly. Popular examples like [WordPress](https://rob-hoeijmakers.ghost.io/how-about-wordpress/), Joomla, and Drupal offer user-friendly interfaces and pre-designed templates to simplify the process. A CMS is perfect for bloggers, content creators, and small websites. Read more: [The right CMS for your online content](https://rob-hoeijmakers.ghost.io/content-management-system-cms/) ## Content Platform A content platform expands beyond content creation, focusing on optimizing content experiences. It includes tools for content distribution, personalization, and integration with marketing technologies. Contentful is a prime example of a content platform that lets businesses tailor content for various channels and devices, enhancing the overall user experience. In essence, the composability of a content platform is a key feature that sets it apart from traditional CMS systems, providing a more agile and adaptable solution for managing and delivering content across various channels and platforms. **Read more**: [Understanding Content Platforms](https://rob-hoeijmakers.ghost.io/content-platforms/) ## DXP: Digital Experience Platform A DXP takes content management to the next level by offering a comprehensive suite of tools for creating seamless digital experiences. It combines content management with features like personalization, e-commerce, analytics, and customer data management. Enterprises looking to deliver sophisticated and tailored experiences across various touchpoints can benefit from a DXP. Notable examples include Sitecore Experience Platform. **Read more**: [Intro DXP](https://rob-hoeijmakers.ghost.io/dxp/) ## Comparing the Concepts Let's compare the three options on key concepts. ### Focus and Purpose: - CMS: Primarily manages content creation, editing, and publishing. - Content Platform: Emphasizes content optimization and distribution across channels. - DXP: Offers a holistic approach to creating cohesive digital experiences. ### Target Audience: - CMS: Ideal for individual bloggers and small businesses. - Content Platform: Suited for marketing teams and businesses aiming to optimize content experiences. - DXP: Geared towards enterprises seeking comprehensive digital experience solutions. ### Key Features: - CMS: Content creation, templates, basic content organization. - Content Platform: Content distribution, personalization, integration with marketing tools. - DXP: Content management, personalization, e-commerce, analytics, customer data management. ## Real-World Examples The differences can be fuzzy because there is overlap. Let me give some real-world examples: 1. **WordPress (CMS):** Rob, a tech-savvy blogger, uses WordPress to effortlessly manage and publish articles about web technologies and online marketing. Its simple interface and customizable themes suit his content creation needs perfectly. 2. **Contentful (Content Platform):** A marketing team at an e-commerce company relies on Contentful to create, optimize, and distribute content tailored for various devices and channels. This ensures a consistent and engaging experience for customers. 3. **Sitecore Experience Platform (DXP):** A multinational corporation chooses Sitecore Experience Platform to provide personalized digital experiences for its diverse audiences, integrating content management, analytics, and e-commerce capabilities. ## Conclusion When selecting the right platform for your [digital strategy](https://rob-hoeijmakers.ghost.io/the-web-strategist-and-the-web-strategy/), consider your goals, audience, and the complexity of your online presence. A CMS is perfect for bloggers and smaller websites, while a content platform like Contentful suits marketing teams aiming to optimize content experiences. For enterprises seeking comprehensive solutions, a DXP like Sitecore Experience Platform offers a wide array of tools to create seamless and engaging digital journeys. By understanding these concepts and their unique features, you can make an informed decision that aligns with your digital ambitions. --- ### Further reading - [The Difference Between Headless CMS & Content Platforms](https://agilitycms.com/resources/posts/headless-cms-vs-content-platforms) ### Understanding Content Platforms URL: https://rob-hoeijmakers.ghost.io/content-platforms/ Last updated: 2026-04-24T18:36:57.000Z In the ever-evolving digital landscape, content management has transcended traditional boundaries. Enter the content platform, a revolutionary concept that's reshaping how businesses create, manage, and deliver content. But what exactly is a content platform, and how does it differ from conventional [content management systems (CMS)](https://rob-hoeijmakers.ghost.io/content-management-system-cms/)? Let's dive in. ## What is a Content Platform? A content platform is an advanced system designed to manage and deliver digital content across various channels. Unlike traditional CMS, a content platform emphasizes cutting-edge technology, flexibility, and composability. ### 1\. Headless Architecture A content platform often incorporates a [**headless architecture**](https://rob-hoeijmakers.ghost.io/headless-cms/), where the content management backend is decoupled from the presentation layer. This separation allows content to be accessed and displayed through APIs on any channel using any front-end technology. **Example:** A company using a content platform can manage content centrally and display it on websites, mobile apps, digital kiosks, and even IoT devices without having to modify the backend. ### 2\. API Mesh An integral part of a content platform is the **API mesh**, ensuring that different services and third-party systems work together harmoniously. It orchestrates how various components interact, creating a personalized and dynamic user experience. **Example:** An e-commerce site using a content platform can integrate inventory management, payment gateways, customer relationship management, and marketing automation tools through an API mesh, ensuring seamless operation. ### 3\. Composability Content platforms are [**composable**](https://rob-hoeijmakers.ghost.io/what-is-a-composable-dxp/), meaning they can combine various components or services flexibly. This composability ensures that different tools and systems can be integrated and orchestrated to create a unified digital experience. **Example:** A media company can compose a content platform with video streaming services, social media feeds, analytics tools, and advertising networks to create a rich and engaging multimedia platform. *Read more: [CMS, Content Platform, or DXP?](https://rob-hoeijmakers.ghost.io/cms-content-platform-or-dxp/)* ## Why choose a Content Platform? Let me give three reasons you might want to opt for a Content Platform. ### 1\. Enhanced Flexibility Content platforms offer unparalleled flexibility, allowing developers to use various front-end technologies and platforms. It's a future-proof solution that adapts to emerging trends and user needs. ### 2\. Improved User Experience With the ability to create tailored content for specific user segments and integrate various services, content platforms enhance the overall user experience, making content more relevant and engaging. ### 3\. Greater Automation and Efficiency Content platforms offer automation capabilities, streamlining processes across departments, reducing manual labor, and improving communication between teams. ## Content Platforms in Action: Real-World Examples If you are interested in Content Platforms and you want to dive in further you can check out these examples. Often the provide a demo of their products. - **Agility CMS:** A popular content platform that offers robust user-management features, advanced content editing tools, and ensures quick deployment while keeping data secure. - **Contentful:** Known for its API-first approach, Contentful provides a scalable solution that facilitates composability and seamless integration with various tools and services. - **Adobe Experience Manager:** Adobe's content platform offers a comprehensive system for managing and delivering content, utilizing cutting-edge technology to provide high-speed and personalized experiences. ## Conclusion Content platforms are redefining the way businesses handle digital content. With their headless architecture, API mesh, and composability, they offer a flexible, efficient, and advanced solution for modern digital needs. Whether you're a developer looking for a scalable way to manage content or a business aiming for a full-fledged platform that interacts with various stakeholders, a content platform could be the right call. In a world where content is king, content platforms are the crown jewels, providing the tools and technologies to create a kingdom of engaging and effective digital experiences. ### Etymology and meaning of "Data" URL: https://rob-hoeijmakers.ghost.io/etymology-of-data/ Last updated: 2023-09-13T10:31:04.000Z In the contemporary digital landscape, the term "data" has become a fundamental pillar. Yet, behind its modern usage lies a captivating etymology that harks back to ancient times, offering insights into its original meaning and usage. ## Origins in Latin: "Dare" and "Datum" The word "data" finds its linguistic roots in the Latin word "datum," the past participle of the verb "dare," which translates to "to give." This linguistic connection provides a glimpse into the historical context that gave rise to the term. In the Latin context, "data" also held significance as a date formula, such as "dabam Romae prid. Kal. Apr.," meaning "I gave this at Rome on the last day of March." This formula evolved into "data Romae," which meant "given at Rome." It was from this concept that the usage of "data" transitioned to denote a specific time (date) when something occurred. Over time, this transition evolved further, and "data" came to represent not just the timing of an event but also the information given at a specific time. This transformation underscores how the term has journeyed from denoting a date to signifying "given" information. ## From scrolls to screens This historical perspective aligns with the contemporary usage of data, albeit in a different form. Today, data is not only about information but also about the act of transmitting it. From electronic messages to digital files, the underlying concept of giving and sharing information remains intertwined with the term. > Data is not the new oil. It’s the new sand. Tim O'Reilly In essence, the etymology of "data" echoes the ancient practice of handing over knowledge, preserving the core idea of providing information. As we navigate the intricacies of the modern digital age, understanding the origins of "data" allows us to appreciate its enduring significance—a concept that continues to bridge the gap between the past and the present. *I also wrote an article on [metadata](https://rob-hoeijmakers.ghost.io/metadata/), which is basically data on data.* 💡 Now we know where the word data comes from, should I use it as singular or plural? I wish there was a simple answer: The usage of "data" as a singular or plural noun depends on the specific context, style guide preferences, and the formality of the writing. The language is still evolving, and [both singular and plural forms are considered acceptable](https://grammarist.com/usage/data/) in various contexts. ### The crucial role of technical writing in composable platforms URL: https://rob-hoeijmakers.ghost.io/the-crucial-role-of-technical-writing-in-composable-platforms/ Last updated: 2026-04-14T13:25:13.000Z In today's fast-paced digital landscape, [composable platforms](https://rob-hoeijmakers.ghost.io/what-is-a-composable-dxp/) with modules and APIs have completely changed the way online professionals build and integrate software solutions. These platforms provide flexibility and agility, allowing developers to create tailored solutions by combining pre-built modules and leveraging APIs. However, the success of such platforms heavily relies on one often overlooked aspect: [**technical writing**](https://rob-hoeijmakers.ghost.io/professional-writer-roles-for-the-web/). In this article, I will explore the significance of good technical writing in the context of composable platforms, focusing on a real-life example where technical documentation played a pivotal role in the success of a digital commerce platform. ## Emporix and their Digital Commerce Platform (DCP) During a conversation with Nicolas Berney, a pre-sales representative at Emporix, I had the opportunity to delve into their composable Digital Commerce Platform (DCP). Nicolas showcased how different modules fit together seamlessly, creating a robust ecosystem for their clients, particularly in the B2B space. As impressive as the product was, it was the technical documentation that truly left a lasting impression. ## The power of well-crafted documentation Emporix's technical documentation was extensive, well-crafted, and immaculately organised in a [documentation portal](https://developer.emporix.io). As someone with over three decades of experience, I had never encountered such exceptional documentation. This level of attention to detail and clarity showcased the company's commitment to providing developers with the tools they need to succeed. It was evident that Emporix recognised the essential role of technical writing in supporting their composable platform. ![Example of a page from of the Emporix documentation portal.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/08/Emporix-docs.jpg) Example of a page from of the Emporix documentation portal. ## Embedding technical writers Nicolas informed me that Emporix maintains two full-time technical writers who are fully integrated into the software design and development process. This approach ensures that documentation is prioritised from the beginning to the end of the project lifecycle. Having technical writers as part of the development team allows for continuous communication, enabling them to gain a deeper understanding of the product's intricacies and write documentation that resonates with developers. ## The significance of MACH-like architecture Emporix's composable platform follows the [MACH (Microservices, API-first, Cloud-native, Headless)](https://rob-hoeijmakers.ghost.io/mach-alliance/) architecture principles. In such a modular and dynamic environment, having comprehensive technical documentation becomes even more critical. It serves as a blueprint for developers, providing a clear understanding of how each module interacts with others and how APIs can be leveraged to build custom solutions. Without such documentation, developers may struggle to navigate the platform effectively, leading to delays and inefficiencies. ## Transparency and developer flow In the fast-paced world of software development, speed and efficiency are paramount. Developers need to be able to turn things around quickly and seamlessly flow through their work. Emporix's dedication to exceptional technical documentation fosters transparency and empowers developers to work efficiently. With easy access to comprehensive and well-structured documentation, developers can remain in their flow, optimising productivity and staying ahead in their competitive landscape. ## Recap on the craft of technical writing As composable platforms with modules and APIs continue to gain prominence, technical writing emerges as a crucial element for success. Emporix's composable Digital Commerce Product serves as an exemplary case study, highlighting the profound impact of comprehensive and well-crafted technical documentation. By embedding technical writers and prioritising documentation from the outset, Emporix has set a standard for other companies in the industry to follow. In the MACH-like architecture, this documentation becomes the indispensable plan and manual for developers, helping them build, integrate, and succeed efficiently in today's rapidly evolving digital landscape. ### Unlocking Digital Asset Management's versatility URL: https://rob-hoeijmakers.ghost.io/unlocking-digital-asset-managements-versatility/ Last updated: 2026-04-25T08:29:40.000Z In the fast-paced digital world, online companies constantly strive for growth and innovation. As their business expands, so do their needs for efficient management and organisation of digital assets. Enter [Digital Asset Management (DAM)](https://rob-hoeijmakers.ghost.io/digital-asset-manager-dam/), a powerful solution that proves to be more than just a static tool. As online professionals, it is essential to understand how DAM scales to meet new requirements and how it can serve multiple purposes, acting as an engine to different vehicles in your digital ecosystem. ## The Limitations of Integrated Media Libraries When online companies are in their early stages, [integrated media libraries](https://rob-hoeijmakers.ghost.io/wix-squarespace-ghost-digital-assets/) within [Content Management Systems (CMS)](https://rob-hoeijmakers.ghost.io/content-management-system-cms/) can often suffice. However, as these companies grow, their needs evolve, and they face new challenges that a standard media library may struggle to address. Issues such as data sprawl, difficulty in locating assets, lack of version control, and inefficient collaboration arise. These constraints can hinder productivity and creativity, putting a strain on the marketing and creative teams and ultimately impacting the business's bottom line. This is where a robust DAM comes into play, providing a centralised solution that scales seamlessly with your growing requirements. ## Scalability of DAM During [a meeting with industry leaders Tenovos and Smint.io](https://tenovos.com/events/sharing-is-caring-maximizing-content-distribution-consumption-and-reuse-with-internal-and-external-teams/), the concept of DAM's scalability was discussed by Michael Waldron and Gert Fahrnberger, revealing the system's potential to adapt to ever-changing business needs. Contrary to popular belief, DAM is not just a rigid tool but, in fact, a highly flexible and adaptable solution. Think of it as a "headless" engine that can power various aspects of your online business effectively. Or [**activate your content**](https://rob-hoeijmakers.ghost.io/content-activation/) as it is called. Just like an engine can be installed in different types of vehicles, a DAM system can serve multiple purposes within your digital ecosystem. Let's explore some of these applications: ### **Brand Portal** A brand portal is a centralised hub that holds all brand-related assets and guidelines. As your business grows, maintaining brand consistency across various marketing channels becomes crucial. A DAM serves as the ideal brand portal by offering a single source of truth for logos, images, videos, and brand guidelines. Marketing teams can access and distribute approved assets, ensuring a coherent brand experience across all touch points. ### **Media Center** In the fast-paced world of media and press releases, instant access to high-quality assets is paramount. A DAM system as a media center provides journalists, bloggers, and influencers with a well-organised repository of images, press releases, videos, and other media assets. This streamlined access not only saves time but also enhances media coverage and improves the brand's reputation. ### **Press Portal** Similar to the media center, a press portal is a dedicated platform that allows PR teams to share press releases, company news, and multimedia content. By utilising a DAM for this purpose, PR professionals can efficiently manage and distribute assets, allowing journalists to access the most up-to-date and accurate information about your company. ### **Content Buying Portal** In some cases, businesses may need to grant access to specific assets for external partners or stakeholders, such as distributors, retailers, or licensing agencies. A DAM can function as a content buying portal, ensuring that the right users can access the relevant assets while maintaining control over usage rights and permissions. Some other uses could be Sales Enablement Portal, Event Portal, Product Launch Portal, E-learning Portal or Campaign Portal. ![Options for content activation.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/08/Content-Activation-options.jpg) Options for content activation by smint.io ## Wrapping it up As online companies grow and evolve, so do their requirements for managing digital assets efficiently. Integrated media libraries within CMS may suffice in the early stages, but as business complexity increases, a Digital Asset Management system emerges as the go-to solution. Not only does DAM scale effortlessly to meet new demands, but it also serves as[ a versatile engine powering various outlets](https://rob-hoeijmakers.ghost.io/digital-asset-management-where-art-and-science-converge/) in your digital ecosystem. By leveraging DAM as a brand portal, media center, press portal, or content buying portal, you can streamline asset management, improve collaboration, and ensure brand consistency. Embrace the power of DAM, and watch your online business flourish with greater efficiency and creativity. *Also read:*[ ***Digital Asset Management: Where Art and Science Converge***](https://rob-hoeijmakers.ghost.io/digital-asset-management-where-art-and-science-converge/)*.* *Explore DAM's role in uniting diverse content purposes. From corporate training to art preservation to sports, DAM delivers versatile solutions.* ### Regular and online calendar invitations URL: https://rob-hoeijmakers.ghost.io/understanding-digital-calendar-invitations/ Last updated: 2026-05-10T08:54:17.000Z In the not-so-distant past, digital calendars were mere offline tools for keeping track of our schedules. However, as technology has advanced, these [calendars have evolved into powerful organizers](https://rob-hoeijmakers.ghost.io/digital-diary-and-calendars/), often connected to servers that enable seamless synchronisation across devices. This has changed the way we manage appointments and invitations, and also posed some challenges to keep things in check. In this article, I will explore the differences between regular invitations (offline) and those sent through Google Calendar (0nline), highlighting the benefits and pitfalls of navigating the synchronised era of digital calendars. ![Typical flow for an invitation from mail to calendar. On iPhone.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/08/Calendar-invite-on-mobile-phone-3.jpg) Typical flow for an invitation from mail to calendar. On iPhone. ## Regular Invitations (ICS files) Regular invitations in digital calendars often come in the form of an ICS (iCalendar) file that is opened and saved locally in the calendar app. When you receive a regular invitation, you get all the details of the event, such as the date, time, location, and purpose of the meeting. You have the option to accept or decline the invitation. The downside of regular invitations is that any changes made to the event details require manual communication with all participants. For example, if the meeting time is rescheduled, the sender must inform each participant individually about the update. This can be time-consuming and may lead to confusion if not managed effectively. 💡 ****Google Calendar or ICS file** If someone sends you an invitation via e-mail via Google Calendar, you can respond in two ways. You can press "yes" and then the invitation will be saved in Google Calendar. Or you can press the ICS file, and then you can choose in which calendar you want to save something. In the first case, adjustments can be made by you and the others. In the second case, not. It is a choice that gives you more control. ## Google Calendar invitations: An online option Google Calendar, as one of the leading cloud-based calendar systems, offers an online solution for handling invitations. When you receive an invitation through Google Calendar, the event details are not only added to your personal calendar but are also stored on Google's servers. This synchronisation feature enables real-time updates and automatic notifications for all participants. If the event's organizer makes any changes to the meeting details, such as rescheduling the time or updating the location, Google Calendar takes care of the rest. All participants receive automatic updates, ensuring everyone is informed without the need for manual notifications. Also Outlook 365, Outlook.com and Yahoo offer online calendar options. ![Example by addevent clearly shows the different regular (Apple/Outlook) and online invitations (Office 365, Outlook.com and Yahoo)](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/08/2023-08-02_14-57-23.png) This option by addevent clearly shows the different regular (Apple/Outlook) and online invitations (Office 365, Outlook.com and Yahoo) 💡 ****Google Calendar or ICS file** If someone sends you an invitation via Google Calendar via e-mail, you can respond in two ways. You can press "yes" and then the invitation will be saved in Google Calendar. Or you can press the ICS file and then you can choose in which calendar you want to save something. In the first case, adjustments can be made by you and others. In the second case, they can't. It's a choice that gives you more control. Which system you prefer may vary and it also has to do with a network effect. What do most others I work with use? In my case, I prefer the Google ecosystem to Apple's. ## Benefits of online invitations 1. **Effortless Updates:** With Google Calendar's synchronisation, event organizers can make changes confidently, knowing that all participants will be promptly notified. This eliminates the hassle of individually informing each participant and reduces the likelihood of misunderstandings or missed meetings. 2. **Collaborative Efficiency:** Online invitations are particularly advantageous for group events and appointments involving multiple participants. It fosters collaborative efficiency by ensuring everyone is on the same page, facilitating smooth coordination among team members, colleagues, or friends. 3. **Seamless Sharing:** Google Calendar's synchronisation simplifies the process of sharing calendars with others. Whether it's colleagues using Outlook or fellow Google Calendar users, the shared event information is automatically updated for everyone, promoting seamless collaboration. ## Cons of online invitations 1. **Fluid Calendars:** One of the drawbacks of the synchronised era is the fluidity of calendars. With automatic updates, events may be rescheduled or altered without much notice, potentially disrupting personal plans or causing conflicts in busy schedules. This constant change may lead to a sense of unpredictability and the need to frequently adapt to shifting appointments. 2. **Over-reliance on Technology:** While digital calendars offer convenience, relying heavily on synchronised invitations can lead to over-dependence on technology. Technical issues, such as server outages or connectivity problems, may result in missed updates, leaving participants uninformed about changes to the event details. 3. **Privacy Concerns:** Cloud-based, online calendar systems like Google Calendar store event information on external servers. While these platforms prioritise security, some individuals may have reservations about sharing sensitive or private event details on third-party servers, leading them to opt for local calendars to maintain greater control over their data. ## Happy scheduling! Whatever your preference or encounter with digital calendars, I hope this information makes navigating the recurring options easier. Whether you embrace the seamless synchronisation of Google Calendar or choose a more controlled approach with regular invitations, the goal remains the same: to make the most of your valuable time and stay organised in a rapidly changing world. ### Further reading - [Digital scheduling calendars compared and explained](https://rob-hoeijmakers.ghost.io/digital-diary-and-calendars/) - [When AI Makes the Calendar Click](https://rob-hoeijmakers.ghost.io/gemini-google-calendar/) - [From Calendly to scheduling with Google Calendar](https://rob-hoeijmakers.ghost.io/from-calendly-to-scheduling-with-google-calendar/) ### Google Search Console, a wealth of insights URL: https://rob-hoeijmakers.ghost.io/google-search-console-gsc/ Last updated: 2026-04-25T13:16:33.000Z As an online professional, you understand better than anyone how important it is to measure, optimise and improve the performance of your website. Google Search Console (GSC) is a powerful tool that helps you do just that. Every day, I use GSC to gain insight into my website's performance and discover valuable opportunities. In this article, I explain how I use the versatility of Google Search Console and how it can provide invaluable value for any professional pursuing online success. ![Google Search Console example](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/08/image.png) Google Search Console example ## Finding your audience and fulfilling intents One of the most important aspects of online success is understanding your audience. With GSC, you can gain insight into which search terms your visitors use to find your website and which pages generate the most traffic. By analysing this data, you can better understand what intents you need to fulfil to effectively reach and satisfy your audience. ## Optimisation opportunities Thanks to GSC, you can gain insight into the searches for which your website is displayed and how often they are clicked. If you notice that certain searches are not driving the desired traffic to your website, you can adjust your content and [SEO strategy](https://rob-hoeijmakers.ghost.io/my-basis-for-seo/) to better match your target audience's intentions. ## Fix technical problems with indexing Indexing issues can have a big impact on your website's visibility in search engines. Fortunately, GSC alerts you to any technical issues that may be hampering your site's indexing. This way, you can take quick action to fix any issues and ensure your website is indexed optimally. ![GSC on mobile](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/08/image-1.png) GSC on mobile ## Fill content gaps A good content strategy is all about providing valuable information to your target audience. With GSC, you can discover which searches and topics are popular but for which your website does not currently provide relevant content. By filling these content gaps (omissions in your content), you not only increase your online visibility, but also demonstrate your authority in your field. ## Insight into inbound and internal referrals GSC allows you to see the websites that link to your site, providing valuable insights into your backlink profile. In addition, you can also analyse internal referrals, which helps to optimise your website's internal link structure and improve the user experience. ## Integration with Ahrefs and WordPress (Site Kit Plugin) As icing on the cake, GSC can be integrated with other valuable tools such as Ahrefs and WordPress (Site Kit Plugin). These integrations provide even more insights and opportunities to analyse and improve your website. ## Who is Google Search Console interesting for? Google Search Console is interesting and valuable for different audiences and professionals involved in managing and optimising websites: - **Website owners**: They want to know how their site performs in search results and which search terms their target audience uses to find the website. - **SEO specialists**: For them, GSC is an essential tool to analyse search data, adjust SEO strategies and increase online visibility. - **Content Marketers**: They can use GSC to discover popular topics and optimise their content strategy. - **Web Developers and Technical Teams**: GSC provides insight into technical issues and helps optimise the website for a better user experience. - **Online Marketers and Digital Strategists**: They can use GSC for data-driven decision-making and to measure the impact of marketing campaigns. - **Webmasters and Web Administrators**: GSC is a central hub to monitor website health and resolve any issues. - **Business owners**: For them, GSC helps understand and increase the online visibility of their business. In short, [Google Search Console](https://search.google.com/search-console/about) provides valuable data and insights that can be used by a wide range of professionals to improve their website's performance and achieve their online goals. 💡 Google Search Console also provides insight into [Google Discover](https://rob-hoeijmakers.ghost.io/get-found-in-google-discover/). Suddenly I had hundreds more visitors and through GSC I discovered why. ### Introduction to GDPR URL: https://rob-hoeijmakers.ghost.io/gdpr/ Last updated: 2026-05-26T19:43:37.000Z The General Data Protection Regulation (GDPR) is a landmark regulation enacted by the European Union to govern data protection and privacy. Implemented on May 25, 2018, this comprehensive law aims to give individuals greater control over their personal data while standardizing data protection laws across EU member states. For the online profession, this means a new paradigm of responsibilities and obligations concerning data collection, storage, and processing. Understanding GDPR is crucial for anyone involved in the online business, from web strategists and marketers to e-commerce owners. Non-compliance isn't an option; it can result in severe financial penalties and reputational damage. This guide will delve into the consequences of ignoring GDPR, its geographical scope, and what it means for businesses globally. ## What is GDPR and What Does It Mean for the Online Profession? GDPR stands for General Data Protection Regulation, a comprehensive legal framework that aims to protect the privacy and personal data of individuals within the European Union (EU) and the European Economic Area (EEA). The regulation touches on various aspects of data management, including data collection, storage, processing, and sharing. For professionals in the online sector, GDPR has several implications: 1. **Data Collection:** Clear consent is required for collecting personal data. This affects online forms, [lead magnets](https://rob-hoeijmakers.ghost.io/gdpr-and-lead-magnets/), and any tool or method used for data gathering. 2. **Data Storage:** Businesses are expected to store data securely and are accountable for any breaches. This impacts how data is encrypted and stored. 3. **Data Processing:** Professionals must have a clear understanding of why and how data is processed, ensuring it aligns with the consent provided by the individual. 4. **Data Sharing:** Sharing data with third parties is subject to stringent guidelines, affecting partnerships and vendor relationships. 5. **Transparency:** Organizations are required to be open about their data usage policies, often necessitating clear privacy policies and consent forms. 6. **Accountability:** Regular data audits may be necessary to ensure ongoing compliance, affecting operational processes and potential resource allocation. Understanding and complying with these aspects are crucial for any online business, not only to avoid penalties but also to build trust and credibility with your audience. ## What Happens if You Ignore GDPR? Ignoring GDPR is not an option any responsible business should consider. Non-compliance could subject companies to a range of enforcement actions, including substantial financial penalties. **Under GDPR, businesses can face fines of up to 4% of their annual global turnover or €20 million, whichever is higher.** These fines apply to serious violations such as processing data without sufficient customer consent or other core GDPR breaches. Moreover, it's not just the financial penalties that you should worry about. Non-compliance can also result in significant reputational damage. Trust is hard to gain but easy to lose. Failing to comply with GDPR can erode customer trust, putting your business at a competitive disadvantage. To avoid these pitfalls, businesses must take proactive steps to protect personal data. This involves implementing technical and organizational measures, such as conducting regular data protection impact assessments and training employees on data protection practices. ## Where Does GDPR Apply? The reach of GDPR extends far beyond the borders of the EU and the EEA. Any company that processes the personal data of EU citizens must comply with GDPR, irrespective of its geographical location. This means that even if your company is based outside the EU, you are still required to comply with GDPR when dealing with EU citizens. Additionally, the GDPR applies to companies offering goods or services to EU citizens, even if those companies are not based within the EU. For instance, a US-based e-commerce site selling to EU customers must also be GDPR compliant. #### EU vs. EEA: What's the Difference? Understanding the difference between the EU and the EEA can help clarify the scope of GDPR. The EU is a political and economic union comprising 27 member states primarily located in Europe. The EEA, on the other hand, includes EU member states as well as three European Free Trade Association (EFTA) countries: Iceland, Liechtenstein, and Norway. Both regions are part of a single market allowing free movement of goods, services, people, and capital. ![Overview of different European organisations](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/09/GDPR-EU.jpeg) Overview of different European organisations In summary, the GDPR is applicable to both the EU and the EEA, making it essential for companies operating in or targeting these regions to be compliant. *Also read: [Lead Magnets and GDPR](https://rob-hoeijmakers.ghost.io/gdpr-and-lead-magnets/)* ### Topic clusters and pillar pages URL: https://rob-hoeijmakers.ghost.io/the-tower-of-seo-success-pillar-pages/ Last updated: 2026-04-25T13:16:34.000Z In the vast ocean of the internet, standing out as an online professional can be quite challenging. However, with the right [SEO strategy](https://rob-hoeijmakers.ghost.io/my-basis-for-seo/) and a well-organised website, you can elevate your online presence to new heights. One effective technique that can make a significant impact on your website's visibility and authority is the use of pillar pages. In this article, we will explore the concept of pillar pages and how they resemble a tower of glasses, creating a smooth flow of "link juice" to enhance your SEO efforts. ## What are pillar pages and their importance? Pillar pages are comprehensive, in-depth pieces of content that cover a specific topic in its entirety. These pages act as the central hub for a particular subject, providing valuable information and linking to related subtopics. The pillar pages form a **topic cluster**. By creating pillar pages, you can establish your website as an authoritative resource in your niche, boosting your chances of ranking higher in search engine results. Imagine each pillar page as a glass in a tower. The top glass is the primary topic, and as you pour the "link juice" of authority into it through internal and external links, it trickles down to the glasses below, reinforcing the strength of your entire website. ![This fountain shows the concept of "link juice" flowing through corner stond and pillar pages..](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/07/27DACEA3-A2C7-4BA7-8D27-24CB4F67C0F6_1_201_a.jpeg) This fountain shows the concept of "link juice" flowing through corner stond and pillar pages. ## Creating cornerstone content for a solid foundation Just like the foundation of a tower holds everything together, your cornerstone content serves as the base of your website's structure. These pages cover broad topics related to your field, such as "Online Content," "Online Marketing," and "Generative AI," showcasing your expertise and providing a comprehensive overview for your visitors. In doing this you build topic clusters. Linking your cornerstone content in your website's menu ensures easy access from any page, making it user-friendly and search engine-friendly. As the "link juice" flows through these cornerstone pages, it spills over into other related topics, boosting their credibility as well. ![Structure of the blog with pillar pages in the center.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/07/2023-07-30_20-29-46.png) Structure of the blog with pillar pages in the center. You can see topic clusters. ## Pillar pages cluster and make cross-links To further fortify your tower of SEO success, create essential pillar pages focusing on specific topics within your niche. For instance, pages like "[Squarespace](https://rob-hoeijmakers.ghost.io/squarespace-the-experiences-after-a-year/)," "[Wix](https://rob-hoeijmakers.ghost.io/why-choose-wix-to-manage-your-valuable-content/)," "[WordPress](https://rob-hoeijmakers.ghost.io/how-about-wordpress/)," "[SEO](https://rob-hoeijmakers.ghost.io/my-basis-for-seo/)," "[CRM](https://rob-hoeijmakers.ghost.io/customer-relationship-management-crm/)," and "[CMS](https://rob-hoeijmakers.ghost.io/content-management-system-cms/)" delve deeper into these subjects, providing detailed insights and practical advice. By interlinking these pillar pages with your cornerstone content and vice versa, you enable visitors and search engines to navigate your website seamlessly. This interconnection allows "link juice" to flow throughout your site, elevating the ranking potential of all related pages. ## The power of internal and footer links The beauty of the tower lies in its interconnectedness, and the same principle applies to your website. Link your content strategically, ensuring that visitors can effortlessly navigate between your pillar pages and other relevant content. Internal links help search engines crawl and index your site efficiently, while also encouraging users to explore more of your valuable resources. Don't forget to utilize the footer section of your website effectively. Include links to all your essential content, blog posts, and a compelling call-to-action, encouraging visitors to get in touch or subscribe to your updates. ![Content about the blog and me is placed, on every page, in the footer.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/07/2023-07-30_20-19-41.png) Content about the blog and me is placed, on every page, in the footer. ## Watching it grow In the world of SEO, the tower of glasses metaphor beautifully illustrates the power of pillar pages and strategic linking. By crafting informative cornerstone content and interlinking it with smaller pillar pages, you can strengthen your website's authority, improve user experience, and enhance your search engine rankings. As you pour the "link juice" of authority into the top glass of your tower, watch as it cascades down, reinforcing each page in its path. Embrace the concept of pillar pages, and you'll find your online presence soaring to new heights, attracting more visitors and establishing yourself as an influential online professional in your field. *Also check out my* [*article on headings*](https://rob-hoeijmakers.ghost.io/html-headings/) *and find out how information flows the page level.* ### WordPress pros and cons URL: https://rob-hoeijmakers.ghost.io/how-about-wordpress/ Last updated: 2026-05-10T08:54:17.000Z WordPress, a name that resonates with millions of online professionals, has been an integral part of the digital landscape for decades. As someone who uses it daily, I have experienced the highs and lows of this powerful content management system. Join me on a journey through its history, delve into the world of plugins, and explore the profound impact it has had on the lives of online professionals. ## The genesis of WordPress: In the early 2000s, two developers, Matt Mullenweg and Mike Little, joined forces to create a blogging platform known as WordPress. Its humble beginnings as a simple tool for bloggers soon evolved into a full-fledged [content management system (CMS)](https://rob-hoeijmakers.ghost.io/content-management-system-cms/). WordPress's open-source nature allowed it to flourish, attracting a vibrant community of developers, designers, and enthusiasts worldwide. ## The power of plugins One of WordPress's greatest strengths lies in its extensive plugin ecosystem. With over tens of thousands of plugins available, online professionals have a seemingly infinite array of tools at their disposal. Whether it's enhancing [SEO](https://rob-hoeijmakers.ghost.io/my-basis-for-seo/), improving [website performance](https://rob-hoeijmakers.ghost.io/measure-page-speeds-with-core-web-vitals/), or adding eye-catching visual elements, there's a plugin for nearly everything. This adaptability makes it easy for users to customize their websites according to their specific needs and preferences. ## The Pros of WordPress Let me give a list with pros of using WordPress: ### 1\. User-Friendly Interface: Even for novices, WordPress offers an intuitive dashboard and user-friendly interface that simplifies content creation and website management. ### 2\. Flexibility and Scalability: From a simple blog to a fully-fledged e-commerce website, WordPress can scale to accommodate various needs. ### 3\. Thriving Community: The vast and active WordPress community provides a wealth of knowledge, support, and inspiration, making troubleshooting and learning easier. ### 4\. Regular Updates: WordPress frequently releases updates to improve security, performance, and functionality, ensuring users stay up-to-date with the latest advancements. ## The Cons of WordPress There are always two sides to a medal. The cons: ### 1\. Security Vulnerabilities: As an immensely popular platform, WordPress becomes a prime target for hackers, necessitating extra precautions to maintain website security. ### 2\. Plugin Overload: The abundance of plugins can be a double-edged sword. While they enhance functionality, using too many plugins can slow down a website and lead to compatibility issues. ### 3\. Learning Curve for Advanced Customization: While basic usage is straightforward, mastering the more intricate customizations may require time and effort to learn. ## My love-hate relationship with WordPress As someone who relies on WordPress daily, I have grown to appreciate its versatility and convenience. The ease of content creation and the vibrant plugin marketplace have undoubtedly made my professional life more manageable. However, my love for WordPress is tempered by moments of frustration, especially during security incidents and occasional plugin conflicts. Despite the challenges, its immense potential keeps me coming back. ![My first blog in WordPress was in 2005.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/07/Kielzog-in-WordPress.jpg) My first blog in WordPress was in 2005\. It was a personal blog and a struggle so now and again. ## My conclusion on WordPress WordPress has undeniably transformed the way online professionals build and manage websites. Its journey from a simple blogging platform to a versatile [CMS](https://rob-hoeijmakers.ghost.io/content-management-system-cms/) has been nothing short of remarkable. Embracing its history, leveraging the power of plugins, and understanding its pros and cons can empower online professionals to make the most of this beloved platform. In closing, I hope this journey through the world of WordPress has been as enlightening for you as it has been for me. As an [Amsterdam-based web strategist ](https://rob-hoeijmakers.ghost.io/about-rob-hoeijmakers/)deeply involved in content realization and online marketing, I've not only explored the ins and outs of WordPress but have also compared it to other platforms like [Squarespace](https://rob-hoeijmakers.ghost.io/squarespace-vs-wordpress/) and [Ghost](https://rob-hoeijmakers.ghost.io/wordpress-vs-ghost-for-bloggers/). If you're looking to dive deeper into these topics or simply want to discuss anything related to web strategy, online marketing, or content platforms, I invite you to connect with me. Let's continue this conversation and together, we can navigate the ever-evolving digital landscape. [Feel free to reach out](https://rob-hoeijmakers.ghost.io/about-rob-hoeijmakers/), and I look forward to engaging with you further. --- ### Further reading - [WordPress vs. Ghost: A comparison for bloggers](https://rob-hoeijmakers.ghost.io/wordpress-vs-ghost-for-bloggers/) - [Squarespace vs. WordPress](https://rob-hoeijmakers.ghost.io/squarespace-vs-wordpress/) ### My basis for SEO URL: https://rob-hoeijmakers.ghost.io/my-basis-for-seo/ Last updated: 2026-04-25T13:17:10.000Z As an online professional, it is crucial to make yourself and your brand stand out in the endless sea of websites and content. Fortunately, there is a magic ingredient that can help you do just that: SEO! Or search engine optimisation. Search Engine Optimisation (SEO) is a great way to make your website more findable by search engines and your target audience. In this foundational article, I'm going to introduce you to the world of SEO in a light-hearted way and give you the outline to get started. ## Keywords and Search Intention Step one in SEO is understanding your target audience and what they are looking for. This is where keywords and search intent come in. Do research on what keywords your target audience uses to search for your services or products. Use tools like Google's Keyword Planner to find relevant keywords and also pay attention to search intent, that is, understanding what exactly the user is looking for. Are they looking for information, do they want to buy something, or are they looking for instructions? ## Content is King When you think of SEO, think of the three Cs: Content, Content, and Content! Write quality, relevant and valuable content that matches your target audience's search intent. Make sure to work with [an outline and clear headings](https://rob-hoeijmakers.ghost.io/html-headings/). Make use of found keywords in your texts, but be subtle and avoid overuse. On your blog, offer information, tips, and solutions that your target audience is looking for. *Related article:* [*The tower of SEO success: Pillar Pages*](https://rob-hoeijmakers.ghost.io/the-tower-of-seo-success-pillar-pages/) ## Technical magic SEO also has a technical side. Make sure your [website loads quickly](https://rob-hoeijmakers.ghost.io/measure-page-speeds-with-core-web-vitals/) and is mobile-friendly. Secure your website with HTTPS and use a clear and logical URL structure. Also think about [optimising images](https://rob-hoeijmakers.ghost.io/crisp-and-fast-loading-images-with-squoosh/) by compressing them and using alt text. ## Link building Link building is like creating a web of trustworthiness. Work on getting quality backlinks from other websites to yours. These links act as votes of trust and can improve your ranking in search engines. Focus on creating valuable content that others will naturally want to share. ## Dive into the data What you don't measure, you can't improve! Dive into the world of [Analytics](https://rob-hoeijmakers.ghost.io/cookie-free-analytics/) and [Google Search Console](https://rob-hoeijmakers.ghost.io/google-search-console-gsc/) to gather important data about your website traffic and performance. Analyse the data to gain insights into what works and what doesn't, and adjust your strategy based on this information. ![Google Analytics 4 dashboard](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/01/GA4.jpg) Google Analytics 4 dashboard Dive deep into the data and discover the undiscovered of your audience. Data gives you the key to their search behaviour, intentions and needs. It allows you to understand what they are looking for and what they crave. With the power of data analytics, you can gain valuable insights and discover what may be missing from your current approach. Let data be your compass on the way to a better understanding of your audience and a more [effective online strategy](https://rob-hoeijmakers.ghost.io/the-web-strategist-and-the-web-strategy/). *Interested in SEO for your own website,*[ *get in touch with me*](https://rob-hoeijmakers.ghost.io/about/)*!* ### Exploring Content Hub Platforms URL: https://rob-hoeijmakers.ghost.io/exploring-specialised-content-hub-platforms-and-vendors/ Last updated: 2026-04-24T18:37:12.000Z In today's digital landscape, content hubs have revolutionised the way organisations manage and distribute their content. These centralised platforms offer a streamlined approach to managing, organising and delivering content. In this article, we delve into the world of specialised content hub platforms and vendors that offer dedicated solutions for creating exceptional content experiences. ## Understanding Content Hubs Specialised Content Hubs are purpose-built platforms designed specifically for content management, consolidation, and distribution. While conventional [Content Management Systems (CMS)](https://rob-hoeijmakers.ghost.io/content-management-system-cms/) can also serve as content hubs, specialised platforms offer unique features and capabilities tailored to content hub functionalities. These solutions provide advanced tools for curating, organising, and delivering content to engage audiences effectively. ## Benefits of Content Hubs 1. Centralised Content Management 2. Enhanced User Experience 3. Multi-Channel Distribution 4. Advanced Content Curation 5. Robust Analytics and Insights ## Prominent Content Hub vendors 1. **Contentful**: Contentful offers a dedicated Content Hub platform that enables organisations to create, manage, and distribute content seamlessly. With content modeling, flexible APIs, and multi-channel publishing capabilities, Contentful empowers businesses to deliver compelling content experiences. 2. **Acquia Content Hub**: Acquia Content Hub provides a specialised solution for building content repositories and distributing content across digital properties. It offers content syndication, content governance, and content distribution capabilities, enabling businesses to streamline their content management processes. 3. **Bynder**: Bynder is a [digital asset management platform](https://rob-hoeijmakers.ghost.io/digital-asset-manager-dam/) that also serves as a specialised Content Hub. It offers features for organising and distributing content, collaborating on content creation, and managing digital assets in a centralised hub. 4. **Third Light IMS**: Third Light IMS is a specialised [digital asset management platform](https://rob-hoeijmakers.ghost.io/digital-asset-manager-dam/) that can be leveraged as a Content Hub. With its content organisation, metadata management, version control, and content distribution capabilities, Third Light IMS facilitates seamless content management and distribution. 5. **Picturepark**: Access a complete, current and centralized record of any type of information at any time: Images, videos, documents, product data and more. ## Choosing the right Content Hub vendor When selecting a specialised Content Hub vendor, consider the following factors: - Specific requirements and goals of your organisation - Scalability and performance capabilities of the platform - Integration options with other systems and platforms - User-friendliness and content authoring features - Support, training, and future scalability offered by the vendor ## My take on Content Hubs Specialised Content Hubs provide organisations with dedicated platforms for efficient content management, distribution, and engagement. Vendors like Contentful, Acquia Content Hub, Bynder, and Third Light IMS offer specialised solutions designed to enhance content experiences. By leveraging these platforms, businesses can streamline their content workflows, deliver personalised content across multiple channels, and gain valuable insights into content performance. ### The Essence of User Experience (UX) and UX Design URL: https://rob-hoeijmakers.ghost.io/user-experience-ux-and-ux-design/ Last updated: 2026-04-25T13:16:34.000Z User Experience (UX) is the art and science of designing digital products and services that provide an enjoyable and effective experience for users. It's not just about the aesthetics of a website or app, but also about understanding users' needs, goals, and behaviour, and creating a seamless, intuitive, and engaging interaction. A crucial element within UX that shapes the experience is UX design. ## What is UX Design and why is it important for an online project? UX Design is the discipline focused on creating an optimal user experience for digital products and services. It encompasses the entire process of research, design, and evaluation to ensure that a product not only functions but also adds value and has a positive impact on the user. Why is UX Design so important for an online project? The answer lies in the core of the user experience: 1. **User-Centricity**: UX Design puts the user at the center of the design process. In-depth research into the target audience and their needs enables designers to gain insights that form the foundation for a user-friendly design. 2. **Optimal Interaction**: The goal of UX Design is to create intuitive and seamless interactions between users and digital products. Thoughtful design makes it easy for users to achieve their goals without confusion or frustration. 3. **Visual Appeal**: UX Design also encompasses the visual aspect of a product. An attractive design appeals to users and contributes to a positive brand perception. 4. **Focus on Efficiency**: UX Design aims for efficiency and minimalism. It eliminates unnecessary steps and elements, allowing users to complete their tasks more quickly. 5. **Responsiveness**: Good UX design considers [different devices and screen](https://rob-hoeijmakers.ghost.io/resolutions-for-screens-and-images/) sizes. It ensures users have a consistent experience, regardless of the device they use. 6. **Emotional Impact**: UX Design can also have an emotional impact on users. A well-designed product can evoke joy, trust, and even loyalty. 7. **Innovation and Competitive Advantage**: A strong UX design can provide a competitive advantage. It allows room for innovation and differentiation in an increasingly competitive digital world. ## The Impact of AI on UX Design In the era of Artificial Intelligence (AI), we witness an exciting evolution of UX Design. AI-driven technologies expand the possibilities and further enhance the user experience. [Chatbots and virtual assistants](https://rob-hoeijmakers.ghost.io/how-chatbots-understand-human-language/) offer fast and contextual support, while personalisation algorithms adapt content, recommendations, and experiences based on user behaviour and preferences. AI also empowers designers with advanced tools to analyse user behaviour and identify patterns that might otherwise be overlooked. By leveraging AI technologies, designers can create UX designs that are even more tailored to the individual user. ## The key for an online project User Experience Design is the key to creating valuable interactions and the success of every online project. It involves understanding users' needs, designing user-friendly and attractive interfaces, and adapting to the continuously changing user requirements. With the rise of AI technologies, UX designers can create even more personalised, efficient, and engaging experiences, making users feel truly understood and valued. ### ChatGPT's Code Interpreter: Coding Without Code URL: https://rob-hoeijmakers.ghost.io/chatgpts-code-interpreter-coding-without-code/ Last updated: 2026-04-25T13:16:38.000Z ## Unraveling Python with ChatGPT's Code Interpreter The landscape of digital innovation is always expanding, and the latest gem, or maybe better bomb shell, that caught my attention is ChatGPT. Specifically its new beta feature: **the Code Interpreter**. 📣 ****Name Change Alert!** ****Code Interpreter is now Advanced Data Analytics!** 🚀 In the wake of OpenAI's big announcement about ChatGPT Enterprise, they also made an important feature name change that almost slipped under the radar. If you've been using ChatGPT's Code Interpreter, don't worry—the feature you love has just been rebranded. This function is a beautiful fusion of machine learning and programming, opening up Python coding to people who, like me, aren't programmers. 💫 ChatGPT Code Interpreter is a new beta functionality integrated within ChatGPT that ****allows users to interact with Python code using natural language**. This feature makes it possible to perform complex tasks, such as ****data analysis and manipulation**, without having to directly write or understand Python code. Users simply input their requests or queries in everyday language, and the Code Interpreter translates that into Python code and executes the required operations. The results can be viewed directly within the interface or exported for further use. At the moment of writing the feature is ChatGPT 4 beta (requires subscription) and so you need to enable it in the ChatGPT settings. The same way as how to [enable plugins](https://rob-hoeijmakers.ghost.io/exploring-the-power-of-chatgpt-plugins-unleashing-the-beta-features-for-subscribers/). ![It is beta so you need to enable it in the ChatGPT settings. ](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/07/DraggedImage-2.png) It is beta so you need to enable it in the ChatGPT settings. ## A blogger's experiment: Digging deep into digital data To give you an idea of its power and simplicity, let me share a real-life example from my blogging experience. I needed to find out which pages on my blog were live but not showing up in Google Search Console results. They were either not indexed or Google's algorithms didn't deem them worthy of traffic. The process to solve this issue traditionally involves exporting two CSV files: one from Screaming Frog with a list of all live pages on my blog, and another from Google Search Console, containing data about my blog's indexed pages. But this time, I just uploaded these files to ChatGPT without any cleaning or preprocessing. From there, the process was as easy as talking to a colleague. I asked the AI to filter out certain URLs such as tags, categories, and pages, and within moments, I had my results. They weren't a complicated mess of URLs, but neatly formatted titles. ![ChatGPT Code Interpreter working on Google Search Console data](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/07/DraggedImage.png) ChatGPT Code Interpreter working on Google Search Console data Excited by this initial success, I decided to probe deeper. I asked Google Search Console for a list of indexed URLs and again, turned to ChatGPT to request a list of all live pages not included in this index. ![ChatGPT Code Interpreter working on Google Search Console data](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/07/DraggedImage-1.png) ChatGPT Code Interpreter working on Google Search Console data The result was almost perfect. Experiencing this level of sophistication made me a bit dizzy. Such activities do not exceed my knowledge but more my experience. I only use it so now and then, by the time I need it again I need to look it up again and fumble my way through it. This consumes loads of time. With this I really feel empowered to check things quickly without the cognitive load of something I will forget again. Side note: While my experience with the Code Interpreter was largely positive, I did encounter one peculiar instance. The AI had mistakenly changed the author in the title and URL of a review I had written. In a strange twist, Bent Flyvbjerg, the actual author of the book 'How Big Things Get Done', was replaced with 'Bill Fisher' in my review's URL. It's a small reminder that while AI is powerful and incredibly helpful, it's not foolproof and may sometimes produce unexpected results. ## My experience with the Code Interpreter The task, which would normally require advanced Python programming knowledge and substantial time, had become something I could manage in a few simple steps. All thanks to ChatGPT's Code Interpreter. But the Code Interpreter is more than a translator or debugger; it's also a teacher and a partner. It's an excellent tool for Python learners at all levels, translating natural language instructions into Python code and helping users like me to better understand how Python can be applied practically. ChatGPT's Code Interpreter is an innovative blend of AI and programming that simplifies Python coding, making data manipulation an achievable task for everyone. Whether you're a business analyst wrangling data, a student learning Python, or a blogger like me just wanting to dip your toes into the coding world, the Code Interpreter is your guide through the uncharted territories of Python programming. P.S. An interesting aspect of my adventure with the Code Interpreter was that I didn't need to write a single line of code. However, it's worth mentioning that while interacting with this tool, you're not entirely separated from the underlying Python code. By simply clicking on specific buttons, the Python code that powers your requests is revealed. So, for those who are curious or learning Python, this feature serves as an invaluable learning aid, providing real-time Python examples for the tasks at hand. You can even copy the Python code for your own reference or use. I've included a screenshot below to illustrate this unique feature. ChatGPT Code Interpreter sharing the actual (hidden) Python code ### Example of Code Interpreter at work Here you can see ChatGPT Code Interpreter at work on some very messy CSV data on Robert de Niro and his ratings in Rotten Tomatoes. It starts cleaning the data and then I ask to plot it in a graph. [Denirodeniro.csv2 KBdownload-circle](https://rob-hoeijmakers.ghost.io/content/files/2023/07/deniro.csv "Download") ### Review of "How Big Things Get Done" by Bent Flyvbjerg URL: https://rob-hoeijmakers.ghost.io/review-of-how-big-things-get-done-by-bent-flyvbjerg/ Last updated: 2026-04-29T08:35:01.000Z While attending the [MACH Conference in Amsterdam](https://rob-hoeijmakers.ghost.io/mach-alliance/), June 2023, I was introduced to the book "How Big Things Get Done" by Bent Flyvbjerg. Intrigued by its relevance to my professional endeavors, particularly in setting up an organization to promote digital literacy and managing various projects in my personal life, I eagerly immersed myself in its pages. ## Think Slow, Act Fast At the core of Flyvbjerg's book is the principle of "think slow, act fast." This concept resonated deeply with me, considering the complexity of the projects I'm involved in. It emphasized the importance of taking the time to deliberate, plan, and analyze potential challenges before executing actions. By doing so, I've discovered the value of informed decision-making, risk mitigation, and the ability to seize opportunities swiftly and efficiently. 💡 In "How to Get Big Things Done," Flyvbjerg vividly contrasts the experiences of two renowned architects, Jørn Utzon and Frank Gehry, in the creation of iconic architectural landmarks. Utzon's tumultuous journey with the Sydney Opera House serves as a cautionary tale, highlighting the importance of thoughtful, well-planned processes (iterations and models) over impulsive decision-making. The book underscores the value of thinking slow and acting fast, emphasizing the significance of meticulous planning and avoiding the pitfalls of underestimating costs and succumbing to political challenges. Utzon's inexperience and the subsequent troubles faced in the Sydney Opera House project starkly illustrate the critical need for experience, whether in the age of individuals or the depth of expertise within a team or organization. Flyvbjerg's narrative powerfully underscores the essence of choosing wisely when embarking on significant projects, leaving readers with a profound lesson in the pursuit of ambitious endeavors. ## The pitfalls of wrong predictions Flyvbjerg's exploration of the dangers associated with incorrect predictions based solely on optimism or an internal perspective struck a chord with me. The book shed light on the disappointment that stems from such misguided forecasts. This realization prompted me to prioritize grounded predictions backed by historical data, utilizing reference class forecasting to improve accuracy and decision-making. It also was a bit consoling for the horrible times you feel like getting stuck in the machine. ![Frank Gehry's Guggenheim in Bilbao. The whole building was modelled out, physically and digitally.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/10/IMG_7479.jpeg) Frank Gehry's Guggenheim in Bilbao. The whole building was modelled out, physically and digitally. ## The power of reference class forecasting (RFC) Through reference class forecasting, Flyvbjerg demonstrated the power of leveraging historical data from similar projects. This approach enabled me to avoid biases and over-optimism, fostering more realistic expectations. By incorporating this method, I gained confidence in making data-driven decisions and enhancing overall project performance. RFC feels like a very solid formula that can be use in many different projects and life situations. ## Harnessing the power of experience Flyvbjerg's insights on the significance of experience resonated deeply with my professional journey. Acknowledging the collective knowledge and lessons learned from past projects, I found inspiration in fostering a culture of experience sharing within my organization. Leveraging the expertise of teams has proven instrumental in improving decision-making, managing risks, and achieving successful project outcomes. ## Review In my opinion, "How Big Things Get Done" is a well-written book, rich with powerful ideas honed through years of experience. While not light reading, it offers valuable insights that I'm determined to internalize and carry forward. Also, the book elucidated the irreversible nature of projects once they are set in motion, even with flawed predictions. These insights have reshaped my perspective and underscored the importance of robust planning, thoughtful execution, and leveraging reference data to inform decision-making. I wholeheartedly recommend "How Big Things Get Done" to those seeking to navigate large projects. Rather than searching for groundbreaking innovations, this book offers time-tested formulas supported by thorough research. It is a valuable resource for professionals looking to enhance their project management skills and achieve success in the face of big challenges. ### Squarespace login URL: https://rob-hoeijmakers.ghost.io/squarespace-login/ Last updated: 2026-04-21T09:06:52.000Z Squarespace has emerged as a popular website builder, empowering individuals and businesses to create stunning and professional-looking websites with ease. I like squarespace a lot, but if you are like me, I keep forgetting where to log in and Squarespace, probably to not distract new customers, has not a clear login button. In this article I share the steps and link for [Squarespace](https://rob-hoeijmakers.ghost.io/squarespace-the-experiences-after-a-year/) and provide a step-by-step guide on how to log in to your Squarespace account via account.squarespace.com. With these steps you log into the dashboard behind your Squarespace site(s). ### Logging in to Squarespace **Step 1: Access account.squarespace.com** To begin, open your web browser and type "account.squarespace.com" in the address bar or simply click on the link provided here: [account.squarespace.com](https://account.squarespace.com) This link will take you to the official Squarespace login page. **Step 2: Choose Your Login Option** Upon arriving at the login page, you will notice several options to access your Squarespace account: Google, Apple, and Facebook. Select the option you have previously linked to your Squarespace account during the registration process. If you signed up using your email address, choose the "Email" option and enter the email and password associated with your account. - **Google**: If you have linked your Google account to Squarespace, click on the "Continue with Google" button. You will be redirected to the Google login page, where you need to enter your Google account credentials. - **Apple**: For Apple users, click on the "Continue with Apple" button. A pop-up window will prompt you to log in with your Apple ID and complete the authentication process. - **Facebook**: If you linked your Facebook account to Squarespace, select the "Continue with Facebook" option. You will be redirected to the Facebook login page, where you need to enter your Facebook credentials. - **Email**: For users who signed up with their email address, select the "Continue with Email" option. You will be prompted to enter the email and password associated with your Squarespace account. ![Squarespace login screen. From here you login to Squarespace.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/08/Squaresapce-login.jpg) Squarespace login screen. From here you login to Squarespace. **Step 3: Complete Two-Factor Authentication (Optional)** For added security, Squarespace may implement [two-factor authentication (2FA)](https://rob-hoeijmakers.ghost.io/two-factor-authentication-2fa/). If you have 2FA enabled on your account, you will receive a verification code via email or text message. Enter the code in the provided field to proceed to the next step. **Step 4: Access Your Dashboard** After successfully logging in, you will be directed to your Squarespace [dashboard](https://rob-hoeijmakers.ghost.io/why-dashboards/). Here, you will see an overview of your website(s), analytics, design options, and more. From the dashboard, you can access and manage your existing sites or create new ones. ### Four ways to build your website or blog (2023) URL: https://rob-hoeijmakers.ghost.io/four-ways-to-build-and-maintain-website/ Last updated: 2026-04-24T18:50:24.000Z In this article, I will give a general overview of the possibilities for building and maintaining your website. After reading the article below, you will have a good understanding of the different ways and methods to build and maintain a website. The main paths are: 1. [Online web builders](#online-web-builders) 2. [Blogging platform or newsletter service](#blogging-platform) 3. [Content Management System (CMS)](#content-management-system) 4. [Build with WYSIWYG editors](#build-locally-with-wysiwig-editor) ## ## Online Web Builders This option is emerging the most. The construction and management of the website takes place completely online. So via a browser (plus any app) and no longer locally on your own computer or mobile. With this method, you usually do not start from scratch, but there are numerous templates available that you can adjust for use. They also often have a collection of photos and images that you can use. The major players at the time of writing (it is a dynamic market) are [Wix](https://rob-hoeijmakers.ghost.io/why-choose-wix-to-manage-your-valuable-content/), [Squarespace](https://rob-hoeijmakers.ghost.io/squarespace), [Webflow](https://webflow.com) and [Jimdo](https://www.jimdo.com/). In addition to these paid services, many hosting providers offer a similar service. An interesting aspect with these Online Website Builders is that the hosting of the website is an integral part. That is very handy and can speed up the setup enormously. ![Web Builders like Webflow, Squarespace, Wix and Jimdo](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/07/wix-squarespace-jimdo-webflow.jpg) Web Builders like Webflow, Squarespace, Wix and Jimdo ## Blogging platform or newsletter service For a number of people, a Blogging Platform such as [Medium](http://medium.com/), [Substack](https://substack.com), [Tumblr](https://www.tumblr.com/), [Ghost](https://rob-hoeijmakers.ghost.io/ghost-blog/) or [Blogger](https://www.blogger.com/) will be an option. These companies provide the tools for publishing the content and often also provide a platform with audience and eventual reward. This works very well if you mainly want to publish content, and it can be combined with your own site or next to a webshop, for example. You do have less say, and it creates a degree of dependence. ## Content Management System (CMS) If a lot of adjustments and extensions are needed after the site has been created, a hosted [Content Management System](https://rob-hoeijmakers.ghost.io/content-management-system-cms/) is a good option. There are numerous systems and the best known are [WordPress](https://wordpress.org/), [Joomla](https://www.joomla.org/), [Django](https://www.django-cms.org) and [Drupal](https://www.drupal.org/). Teams can easily work on it and the information is stored in such a way that it is easily searchable for the visitor but also for the authors. Setting it up is often quite a job. You can arrange the hosting yourself or purchase it as a service. For example, from [WordPress](https://rob-hoeijmakers.ghost.io/how-about-wordpress/) itself or from your hosting provider. Popular in the last years, especially amongst developers, are **static site generators** like [Jekyll](https://jekyllrb.com), [Hugo](https://gohugo.io) and [Gatsby](https://www.gatsbyjs.com). Rather than dynamically serving out a site, they generate it. This makes these sites run very stable and speedy. The CMS and the website generation are separate, and such an architecture is also called headless. 💡 This article discusses the concept of static site generators (SSGs) and their significance in the Jamstack ecosystem. SSGs are tools that help build static websites by generating HTML pages at build time. The use of SSGs offers several benefits, including improved website speed, enhanced security, and scalability. Compared to traditional dynamic websites, static sites served from a content delivery network (CDN) load faster and perform better. SSGs work well with third-party APIs, allowing the addition of dynamic elements to static sites. In addition, the article emphasizes the importance of [headless CMS ](https://www.hoeijmakers.net/headless-cms/)([Content Management System](https://www.hoeijmakers.net/content-management-system-cms/)) solutions in conjunction with SSGs. A headless CMS enables content managers and editors to easily update and manage the content of a static site, providing a user-friendly interface for non-technical users. By combining an SSG with a headless CMS, users can leverage the simplicity and performance of static HTML while still having the flexibility to manage and update the site's content effectively. [Static Site Generators explained](https://bejamas.io/discovery/static-site-generators/) ## Build with WYSIWYG editor In addition to building and maintaining a website online, you can also do it locally. That is to say, on your own computer with a program that offers WYSIWYG. You immediately see what you are building. For years this was the standard and widely used was [Adobe’s Dreamweaver](https://www.adobe.com/nl/products/dreamweaver.html). A more current variant is [Pinegrow](https://pinegrow.com/). In such a program you create the website on your computer, it offers a number of powerful tools for templates, photo editing and links with other systems. As soon as you have finished building or modifying it, put it live. With this method, you can also use existing templates. ## Conclusion Many ways lead to Rome. I still switch between solutions, and that can be cumbersome and time-consuming. But goals, brands, opportunities often change and flexibility is important. One thing I would always consider when checking for a solution is to see to which degree you will be locked in when you want to change. Also, I combine methods and applications. I use [Ghost](https://rob-hoeijmakers.ghost.io/your-bio-site/) for blogging and both [Squarespace](https://rob-hoeijmakers.ghost.io/squarespace-the-experiences-after-a-year/), [WordPress](https://rob-hoeijmakers.ghost.io/how-about-wordpress/) and [Wix](https://rob-hoeijmakers.ghost.io/why-choose-wix-to-manage-your-valuable-content/) for websites. My[ bio site](https://rob-hoeijmakers.ghost.io/your-bio-site/) I build with just HTML and CSS. ### Squarespace versus WordPress URL: https://rob-hoeijmakers.ghost.io/squarespace-vs-wordpress/ Last updated: 2026-05-10T08:54:17.000Z An important factor to consider is cost. In this article, we will list the costs of [Squarespace](https://rob-hoeijmakers.ghost.io/squarespace-the-experiences-after-a-year/) and [WordPress](https://rob-hoeijmakers.ghost.io/how-about-wordpress/) so you can make an informed decision. And to be honest, I would love to summarise it more succinctly, but that would not do justice to the reality. And both products are not a one-off purchase, but a relatively long-term choice. So let's take a deep breath and walk through it together. ## Platform cost One of the first things people look at when comparing Squarespace and WordPress is price. WordPress itself is free and open source. This means you can download and install the software for free on your own web hosting. However, it is important to note that there may be other costs, such as hosting and domain registration. When using WordPress on a self-hosted website, you need to choose a hosting provider. The cost of this varies, depending on the chosen provider and hosting package. You also need to register a domain name, which is usually an annual cost. However, some hosting providers offer free domain registration when you take out a hosting package. At Squarespace, hosting and a domain name are included in the monthly or annual subscription plan. This makes it easy and convenient as you don't have to find a separate hosting provider or pay domain fees. However, keep in mind that if you want to use a custom domain instead of a standard Squarespace domain, there may be additional costs for registering and managing that domain. Apart from hosting and domain, there are also costs associated with email services. If you want to have an e-mail address linked to your domain name, you can choose to use an external e-mail provider such as [Google Workspace](https://rob-hoeijmakers.ghost.io/choosing-google-one-to-use-google-meet/) (formerly G Suite). This service has a monthly cost per user. With Squarespace, you can integrate email services from Google Workspace directly into your website, but there are additional charges for this. For self-hosted WordPress sites, you can also integrate email services through your hosting provider. Some hosting packages offer free email accounts as part of the hosting package, while other hosting providers may charge extra for email services. It is important to consider these additional costs for domain registration and email when comparing the total cost of Squarespace and WordPress. 💡 ****Rough estimation** As there are multiple pricing plans, it is difficult to give an exact breakdown of costs but you probably want to have a rough estimate. Squarespace with email comes out to plus minus 200 euros per year with usually a discount in the first year. A simple WordPress website with a hosting provider comes to plus minus 100 euros per year, but with more work and maintenance. ## Maintenance and technical aspects Another aspect to consider is the maintenance and technical side of your website. WordPress requires more commitment in this area. You are responsible for maintaining and updating the software, themes and plugins you use yourself. This may require extra time and technical knowledge. However, if you like having full control over your website and like tinkering with the technical aspects, this can actually be an advantage. Squarespace, on the other hand, takes care of the technical aspects of your website. They take care of hosting, software updates and security, so you can focus on creating content and designing your website. This makes Squarespace a good choice for people who have no technical background or prefer not to deal with technical maintenance. ![WordPress versus Squarespace](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/07/squarespace_vs_wordpress.jpg) WordPress versus Squarespace ## Extensions and functionalities Adding extra functionalities to your website is often an important aspect when choosing a platform. WordPress has an extensive ecosystem of themes and plugins that you can use to extend your website. Many of these themes and plugins are affordable or even free. This gives you the flexibility to customise your website to your specific needs. Squarespace also offers some extensions to extend the functionalities of your website, but the range is generally more limited. Moreover, the prices of these extensions can be higher. What sets Squarespace apart, however, is the integration and compatibility of the available functionalities. Because everything is built within the platform itself, Squarespace ensures a consistent user experience without having to worry about incompatible plugins or themes. ## Complexity and purpose of the website The complexity and purpose of your website also play a role in choosing between Squarespace and WordPress. If you want to set up a simple, brochure-style website with limited complexity, Squarespace can be a good choice. It offers intuitive tools and templates that allow you to quickly create a professional-looking website even without much technical knowledge. On the other hand, if you want more complexity and customisation options, and if you have access to time and expertise, WordPress may be more suitable. With WordPress, you can dive deep into the code, create custom themes and plugins, and fully customise your website. This makes it a powerful platform for people with specific requirements and deep technical knowledge. ## My personal conclusion It is clear that both [Squarespace](https://rob-hoeijmakers.ghost.io/squarespace-the-experiences-after-a-year/) and [WordPress](https://rob-hoeijmakers.ghost.io/how-about-wordpress/) come with their own benefits and costs. It is important to carefully consider your needs, ambitions and budget when making your choice. If you are looking for a user-friendly platform with less technical hassle and want to set up a simple website quickly, Squarespace may be the best option for you. It offers integrated functionalities and takes care of technical maintenance. On the other hand, if you want more control and customisation options, and you have the time and expertise to invest in managing and maintaining your website, WordPress is probably the better choice. It offers extensive opportunities to customise your website to your liking. But remember, there is no definitive right answer. It's all about your specific needs and what best suits your goals and capabilities. Do your research, possibly try out both platforms and make an informed decision that suits you and your website. If you are interested in getting help choosing the right platform, you can also [contact me](https://rob-hoeijmakers.ghost.io/about-rob-hoeijmakers/). *Read more: I also* [*compared WordPress to Ghost for blogging*](https://rob-hoeijmakers.ghost.io/wordpress-vs-ghost-for-bloggers/)*.* --- ### Further reading - [WordPress versus Ghost: A comparison for bloggers](https://rob-hoeijmakers.ghost.io/wordpress-vs-ghost-for-bloggers/) ### Use ChatGPT plugins right away URL: https://rob-hoeijmakers.ghost.io/exploring-the-power-of-chatgpt-plugins-unleashing-the-beta-features-for-subscribers/ Last updated: 2026-04-25T13:16:42.000Z As a devoted writer and enthusiast of digital affairs, I recently stumbled upon an exciting beta feature in my OpenAI account that I'm eager to share with my fellow [ChatGPT](https://rob-hoeijmakers.ghost.io/try-out-chatgpt-yourself/) Plus subscribers. It turns out that, as part of the beta offerings, subscribers now have access to an innovative capability called ChatGPT plugins. In this blog post, I will delve into the details of this exclusive benefit, shedding light on the nature of these plugins, how subscribers can activate them, and share my personal surprise in discovering their availability. ## Access granted: ChatGPT Plugins in Beta Deep within the realms of my OpenAI account as a ChatGPT Plus subscriber, I discovered an intriguing option under the GPT-4 button—ChatGPT plugins. It is important to note that this feature is currently in the beta phase, allowing subscribers to explore and experience its potential firsthand. By being part of the ChatGPT Plus program, you have exclusive access to this exciting and evolving feature. ## From waiting list to instant activation To my surprise, I had initially placed myself on the waiting list for access to the ChatGPT plugins beta, hoping to be granted permission to explore this new frontier. However, despite my anticipation, I never received a response from OpenAI. That's why it was even more astonishing to discover that as a ChatGPT Plus subscriber, I had immediate access to activate and start using the plugins without any additional steps or waiting. ## Introducing ChatGPT Plugins ChatGPT plugins are specialized add-ons designed to augment your AI interactions within the ChatGPT environment. These plugins cater to a diverse range of professional needs, enabling you to tap into AI capabilities and seamlessly integrate them into your workflow. From language translation and sentiment analysis to content generation and coding assistance, each plugin brings unique functionalities to enhance your productivity and efficiency. ## Activating ChatGPT Plugins To activate ChatGPT plugins, simply navigate to the settings and activate plugins. Then go to the GPT-4 section within your OpenAI start screen. From there, you will discover a collection of plugins available for you to explore. By selecting and activating the desired plugins, you can unlock their specific features and witness their transformative potential firsthand. This seamless integration ensures that you can leverage the power of AI without leaving the familiar ChatGPT environment. Unlock ChatGPT Plugins beta ## Unlocking the full potential As a ChatGPT Plus subscriber, you now have the opportunity to embrace the future of AI interaction. By utilizing ChatGPT plugins, you can optimize your workflows, streamline tasks, and harness AI's vast capabilities to accomplish more. Whether you're a content creator seeking assistance with translation or a developer looking for coding support, these plugins provide a gateway to new dimensions of productivity and creativity. ## Explore further In conclusion, I encourage you to seize the opportunity to explore the world of ChatGPT plugins. As a ChatGPT Plus subscriber, you have exclusive access to this exciting beta feature, allowing you to shape the future of AI interaction. Activate the plugins that align with your professional needs and witness how they can revolutionize your AI-assisted journey. *Note: The ChatGPT plugin feature is currently in beta and exclusively available to ChatGPT Plus subscribers. The availability of immediate access may vary.* ### Increased productivity with AI in chat services URL: https://rob-hoeijmakers.ghost.io/increased-productivity-and-better-performance-with-ai-in-chat-services/ Last updated: 2026-05-10T08:54:17.000Z Generative AI systems, such as ChatGPT, can quickly lead to increased productivity in chat services. Recent studies show that the deployment of such systems has a significant impact on employee performance and organisational efficiency. In this article, I discuss the key findings and what they mean for businesses. ## AI accelerates learning and supports employee performance [Research](https://www.nber.org/system/files/working%5Fpapers/w31161/w31161.pdf) shows that generative AI systems can **improve productivity in certain office functions by as much as 14%**. They can also boost the performance of lower-performing employees. These include AI-assisted chat functions for support staff. These then receive assistance via AI recommendations. AI also accelerates the learning curve, **allowing employees to gain six months of experience in just two months**. In positions with relatively high turnover, this is naturally extra beneficial. In addition, AI deployment ensures that high-performing employees create valuable knowledge that is reinforced through AI systems, benefiting the entire organisation. ![Graph showing the results described. Source is Brynjolfson, via exponentialview.co](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/06/c112f42e-d1a6-41db-bdb7-5c3a1e7d664f_1600x1273.jpg) Graph showing the results described. Source is Brynjolfson, via exponentialview.co ## Adjustments in organisational structure and management Organisations will need to adjust their performance appraisal, reward structures and human resource management to account for the impact of AI on employee productivity and performance. Middle managers, traditionally the custodians of implicit knowledge, may need to rethink their roles and responsibilities as AI takes on more knowledge transfer tasks. Alternative approaches to workforce structure and training may be needed, rather than just focusing on downsizing management layers. 💡 ChatGPT can be used for both automation and augmentation within the context of helpdesk. ****Automation** means using ChatGPT to run processes without human intervention, such as answering frequently asked questions or providing automated help to customers. This saves time and reduces the workload for human employees. ****Augmentation**, on the other hand, refers to using ChatGPT to support people and enable them to provide better answers to customer queries. This is done by using ChatGPT to find relevant information and present it to the human employee, improving the quality of customer service and making customers more satisfied. ## Uncertainties and opportunities for the future The overall impact of these changes on productivity and dynamics within the workforce remains uncertain. However, the initial findings offer valuable insights and raise important questions for companies to consider. By proactively anticipating these developments and implementing AI appropriately, companies can take advantage of the opportunities AI offers for chat services and strengthen their competitive position. In the coming months, we will soon learn more about the opportunities, possibilities and challenges of AI for (smaller) companies. We will be sure to share those findings here, subscribe to stay informed. --- ### Further reading - [🐙 Generative intelligence can propel productivity](https://www.exponentialview.co/p/generative-intelligence-can-propel) - [Verhoogde productiviteit en betere prestaties met AI](https://www.chatvoorbedrijven.nl/verhoogde-productiviteit-en-betere-prestaties-met-ai-in-chat-dienstverlening/) ### Taking photos with the iPhone 12 Mini URL: https://rob-hoeijmakers.ghost.io/taking-photos-with-the-iphone-12-mini-a-powerful-camera-in-a-compact-size/ Last updated: 2026-05-10T08:54:18.000Z My experience with the iPhone 12 Mini has convinced me that size is not always everything, especially when it comes to photography. In this post, I share why this compact smartphone has become my favourite camera. ## Pocket size is an advantage I really appreciate the compactness and lightness of the iPhone 12 Mini. Its size makes it convenient to carry everywhere and allows me to shoot with one hand, something that can be trickier with larger devices. Having a capable camera handy in my pocket at all times is incredibly convenient. ![With the convenient size of iPhone 12 mini camera, I can go to the edge 😉](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/07/DraggedImage-2.jpg) With the convenient size of iPhone 12 mini camera, I can go to the edge 😉 ![And shoot with one hand.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/06/image.png.webp) And shoot with one hand. ## Technical capabilities Despite its size, the iPhone 12 Mini is not technically lightweight. Its dual camera, zoom and wide-angle capabilities give me considerable flexibility when shooting. No matter whether it's a close-up or a wide landscape, the 12 Mini has it all. One aspect I really appreciate is the iPhone 12 Mini's ability to take excellent photos in low light. It's not unusual to try to capture an important moment but struggle because of insufficient lighting. Fortunately, the 12 Mini has a Night Mode that automatically turns on when needed. This feature allows for longer exposure times, allowing more light to hit the sensor, and combined with Apple's smart software algorithms, noise and motion blur are minimised. This allows me to take beautiful photos even in less than ideal lighting conditions, which is a huge plus. ![Mill at dusk.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/06/IMG_2973-1.jpeg.webp) Mill at dusk. ## Clever software What I also appreciate is the built-in software offered by the iPhone 12 Mini. The easy-to-use editing tools allow me to adjust my photos to my liking, [add a vignette](https://rob-hoeijmakers.ghost.io/vignettes-in-iphone-photography/) and [share them instantly](https://rob-hoeijmakers.ghost.io/heic-a-blood-type-for-images/). This is a great advantage in the digital age, where speed and efficiency are paramount. But it's also nice and quick and convenient, I post photos to Twitter or Instagram as I walk away. Or I quickly add [another colour palette](https://rob-hoeijmakers.ghost.io/coolors-color-palette/). *Also read:* [*High Dynamic Range (HDR) in photo's*](https://rob-hoeijmakers.ghost.io/high-dynamic-range-hdr/) ![Quick editing of colours.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/06/Untitled-design.jpg.webp) Quick editing of colours. ## Possible limitations No device is perfect, and the iPhone 12 Mini is no exception. One of the limitations I have encountered is its megapixel capacity. While this is more than adequate for digital playback and standard print formats, it can be a limitation when printing larger formats such as posters. Also, with the lack of an iPhone 14 Mini in recent releases, there is the possible limitation that Apple may not develop this format in the long run. 💡 ****iPhone 13 mini** The main difference between the iPhone 12 mini and the iPhone 13 mini is that the iPhone 13 mini has a number of improvements and upgrades over its predecessor. Some of these improvements include a faster A15 Bionic chip, a better camera with new features and improved battery performance. ### Final verdict The iPhone 12 Mini may be small, but it is certainly capable. I appreciate the combination of portability, technical capabilities and user-friendly software that this smartphone offers. Despite the limitation in megapixels for large prints and the possible lack of future developments in this format, the iPhone 12 Mini remains an excellent choice for my photography needs, especially in a predominantly digital landscape. *Also read:* [*Catching the magic of magenta with your iPhone*](https://rob-hoeijmakers.ghost.io/catching-the-magic-of-magenta-with-your-iphone/) The lion's share of the photos on this blog were taken with the iPhone, as well as all my Instagram photos: > [ Dit bericht op Instagram bekijken ](https://www.instagram.com/p/CuPwJrXrLus/?utm%5Fsource=ig%5Fembed&utm%5Fcampaign=loading) > > [Een bericht gedeeld door Rob Hoeijmakers (@robhoeij)](https://www.instagram.com/p/CuPwJrXrLus/?utm%5Fsource=ig%5Fembed&utm%5Fcampaign=loading) ### Further reading - [From Mini to Pro: My Experience Switching to the iPhone 16 Pro](https://rob-hoeijmakers.ghost.io/iphone-mini-to-pro/) 🇨🇳 【中文摘要】本文介绍了 iPhone 12 mini 的摄影功能，强调其小巧机身与专业拍摄性能的平衡。作者探讨了在日常生活中使用这一设备的体验与意义。 ### Review: “How to Write Copy That Sells" by Ray Edwards URL: https://rob-hoeijmakers.ghost.io/review-how-to-write-copy-that-sells-by-ray-edwards/ Last updated: 2026-05-14T21:01:13.000Z In the digital landscape where persuasive copy is crucial for sales and conversions, "How to Write Copy That Sells" by renowned copywriter Ray Edwards offers a trusted and practical resource for digital experts and entrepreneurs. This article will provide a comprehensive review of the book, exploring Ray Edwards' expertise, the practical insights shared, the relevance of the book for those working with copywriters, and the P.A.S.T.O.R. Copywriting Framework™. ## What is the book about? Ray Edwards' book stands as a testament to his experience and credibility in the copywriting business. While marketed as a step-by-step system, it primarily serves as **a valuable reference tool and checklist for digital professionals** seeking to enhance their copywriting skills. Edwards' expertise shines through as he shares practical insights that can be directly applied to the creation of compelling sales copy for the web. ## Ray Edwards' Experience and Quality One of the book's standout features is Ray Edwards himself—an accomplished copywriter with a wealth of experience. His extensive knowledge and success in the industry lend credibility to the advice he imparts. As readers delve into the book, they can trust that they are learning from a seasoned professional who has earned his wings in the copywriting world. ## Practical Insights for Digital Experts "How to Write Copy That Sells" is a treasure trove of practical insights tailored specifically for digital experts and entrepreneurs. Edwards tackles various essential aspects of copywriting, including the psychology of persuasion, crafting attention-grabbing headlines, and structuring persuasive sales letters. His advice is clear, actionable, and directly applicable to the digital landscape. 💡 Copy Thesis™ by Ray Edwards: Any \[YOUR AUDIENCE\] can \[SOLVE THEIR PROBLEM\] by using \[YOUR PRODUCT\], because \[HOW IT SOLVES THE PROBLEM\]. Example: Any \[solo entrepreneur\] can \[be successful online\] by using \[my course\], because \[it will equip them with the right tools, support and knowledge\] ## Relevance for Digital Professionals For those working with copywriters and sales copy in the digital realm, this book serves as a valuable resource. Digital experts can benefit from Edwards' tips and tricks, using them as a benchmark for evaluating the quality of sales copy produced by copywriters. By understanding the principles outlined in the book, digital professionals can collaborate more effectively with copywriters, ensuring that their sales copy aligns with their marketing goals. > “An irresistible offer is one that’s so appealing, it sells itself. You don’t ask people to buy. They ask you.” ## The P.A.S.T.O.R. Framework Ray Edwards introduces the P.A.S.T.O.R. Copywriting Framework™, which serves as a central theme throughout the book. The acronym stands for the major sections of effective copy: 1. Person, Problem, Pain 2. Amplify, Aspirations 3. Story, Solution, System 4. Transformation, Testimony 5. Offer 6. Response This framework provides a structured approach to crafting persuasive copy that resonates with the target audience, addresses their pain points, offers solutions, and leads them to take action. While the book is not solely focused on this framework, Edwards incorporates its principles and strategies throughout, enhancing its practicality and applicability for digital experts seeking to create compelling sales copy. I personally really like it and try to work with it as often as possible. ## My review "How to Write Copy That Sells" by Ray Edwards is a must-read for digital experts, entrepreneurs, and anyone seeking to enhance their copywriting skills. While the book may not strictly adhere to a step-by-step system, it really works as a trusted reference guide and checklist. Ray Edwards' expertise, combined with the practical insights shared, makes this book an invaluable resource for digital professionals working with copywriters and sales copy in the online sphere. [121: How to Write Web Copy That Sells \[Podcast\]](https://rayedwards.com/121/) ### Coolors: colour palettes in a snap URL: https://rob-hoeijmakers.ghost.io/coolors-color-palette/ Last updated: 2026-04-25T13:16:37.000Z In the colorful world of creativity and visual aesthetics, finding the perfect colour palette is essential. Luckily, there is an app that helps you generate and share stunning colour schemes effortlessly - Coolors! Whether you're a designer, artist, marketer or just a creative soul, Coolors can inspire you . ## What is Coolors? Coolors is your ultimate sidekick for generating colour palettes. This amazing app is designed to help you explore, create and share beautiful colour combinations. With Coolors, you have a tool to really bring your visual projects to life! 💡 The ways of exporting and sharing the palettes created is very well done. You can share it as a URL but also as a PDF, SVG or image. In addition, you link it with Procreate or with Adobe via an ASE (Adobe Swatch Exchange) file. ## How does Coolors work? Coolors is super easy to use. Here's how it works: 1. **Discover**: Dive into the world of colours with Coolors. Get inspired by randomly generated palettes and discover exciting new colour combinations. When you see a colour you like, save it and Coolors instantly generates a new palette based on that colour! You can also work with your own photos. 2. **Create**: Give your creativity free rein with Coolors' customisation options. Grab an existing photo and select the colours you would like to see in your palette. Create a collage of photo and palette together. 3. **Share**: Share your beautiful creations with the world! Export your colour palettes and share them on social media like Twitter, where others can enjoy your artistic colour choices. Get inspired by the Coolors community and discover what others have created. ![Under "other tools" you will find the option to create a palette using your own photos.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/06/Untitled-design-2.jpg.webp) Under "other tools" you will find the option to create a palette using your own photos. ## What can you do with Coolors? With Coolors, you can bring a wide range of creative projects to life. Here are just a few ideas: 1. **Graphic design**: Give your designs a visual boost with beautiful colour palettes. Whether you're designing a logo, poster, brochure or website, Coolors will help you find the perfect combinations that make your project stand out. 2. **Interior design**: Create a harmonious atmosphere in your home with the help of Coolors' colour expertise. Get inspired by the palettes on offer and give your interior a stylish and cohesive look. 3. **Art and illustration**: Express your creativity with Coolors! Whether you paint, draw or create digital art, experiment with different colour palettes to really bring your artwork to life. 4. **Social media and branding**: Make your online presence on social media eye-catching and recognisable using Coolors. Define your brand colours and ensure consistency in your posts and branding to make a strong impression on your followers and potential customers. 5. **Fashion and textile design**: For fashion designers and textile artists, choosing the right colours is essential. Coolors offers inspiration and helps you create colour palettes that suit your style and designs, whether it's clothing, accessories or textile prints. 6. **Inspiration and sharing**: Coolors is not only a great tool, but also a thriving community of creative minds. Share your colour palettes with others and discover inspiring creations from fellow users. Get inspired by the diversity and originality of colour choices and keep growing as a creative visionary. ![The palettes used are inspiring and easy to export for use elsewhere.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/06/Untitled-design-3-1.jpg.webp) The palettes used are inspiring and easy to export for use elsewhere. ## Start working with Coolors! Now that you're familiar with the versatility and creative possibilities of Coolors, it's time to get started yourself! You can easily download the app on your mobile device or visit the Coolors website to directly access the powerful colour generator. Get inspired by the randomly generated palettes, experiment with colours, and create your unique colour schemes to suit your style and projects. Share your creations with the Coolors community and get inspired by others. So what are you waiting for? Discover the magic of colours with Coolors and let your creativity shine! Note: Coolors is available for both iOS and Android devices and can also be used via the official website of [Coolors](https://coolors.co/). 💡 Coolers lets you generate palettes in several ways but that requires a Pro subscription. The ways are: Monochromatic, Analogue, Complementary, Shared complementary, Triadic, Tetradic, Square ### What is SVG and what is it used for? URL: https://rob-hoeijmakers.ghost.io/what-is-svg-and-what-is-it-used-for/ Last updated: 2026-04-14T13:19:21.000Z To begin with, SVG is an abbreviation for Scalable Vector Graphics. It is an image format for vector graphics that is supported in all modern web browsers. SVGs are defined in XML, a language that is human-readable and search engine-friendly. This makes them easy to create and edit with any text editor. ## Vector graphics and the advantages of SVG Unlike raster graphics, which are made up of a grid of pixels, [vector graphics are based on mathematical formulas](https://rob-hoeijmakers.ghost.io/the-difference-between-raster-and-vector-graphics/). These formulas represent lines and shapes, so unlike pixel-based images, they are not distorted or blurred when enlarged or resized. SVG files are therefore exceptionally suitable for logos, icons, infographics and any image that needs to be able to be displayed at different sizes without sacrificing quality. Moreover, SVG files tend to be smaller than their pixel-based counterparts, making them ideal for use on the web, where load time is crucial. ![Large logos and letters require vector files. ](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/05/B3510A12-A8CA-4A93-AEDE-C9832CF9FEA9_1_201_a-scaled.jpeg.webp) Large logos and letters require vector files. ## SVGs and raster images Although SVGs are primarily intended for vector graphics, they can also contain raster images. However, this can cause problems. Because SVGs are essentially text files, they can grow exponentially in size when they contain raster images. Moreover, some platforms and services that support SVG may have problems displaying the embedded raster images correctly. For these reasons, it is usually better to keep raster images as separate files and not embed them in SVGs. 💡 SVG images have the additional advantage that they can have transparency just like PNG and WebP. Again, PNG images can be included in an SVG. ## The flexibility of SVG: text and fonts One of the many advantages of SVGs is their flexibility with regard to text. SVGs can contain text as part of the vector image. This text is selectable and searchable, just like the text on a web page, which can be an added advantage for search engine optimisation (SEO). When including text in an SVG, creators can choose to reference the font or include the letters themselves as vectors. By referencing a font, the size of the SVG can be reduced. However, if the user does not have the font installed on their system, the text will not display correctly. By including letters as vectors, it can be guaranteed that the text is always displayed correctly, regardless of the fonts installed on the user's system. The downside is that this may increase the size of the SVG. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/02/SVG-vectors.jpg) An SVG can simply be called up in a text editor and you will see the vectors. 🎨 I needed to quickly convert an SVG from colour to grayscale and the online tool that really worked was [De-editor](https://deeditor.com/). ****Deeditor**: Online, SVG color editing has never been easier, faster, and more powerful! ## Embedding SVG in a webpage When embedding SVGs in a web page, there are several approaches you can use, depending on your specific needs and desired use. In this section, I will explore some of the most commonly used methods. ### 1\. Inline SVG Inline SVG is a method where the SVG code is inserted directly into the HTML structure. This provides full control over the SVG elements and allows manipulation with CSS and JavaScript. ```html ``` Two examples: ### 2\. ``\-tag Een andere eenvoudige manier om SVG's in te bedden is door gebruik te maken van de ``\-tag. Hierbij wordt het SVG-bestand opgegeven via het `src`\-attribuut. `Beschrijvende tekst` An example: ![Beschrijvende tekst](https://webstrategieblog.nl/wp-content/uploads/2023/05/hoeijmakersnet-high-resolution-logo-color-on-transparent-background-1.svg) ### 3\. ``\-tag The`-tag can also be used to embed an SVG. Het voordeel hiervan is dat je toegang hebt tot de SVG als een documentobject, waardoor je deze kunt manipuleren met JavaScript.` `` An example: ![This is your image alt](https://webstrategieblog.nl/wp-content/uploads/2023/05/apple1.svg "Your image title tag") ### 4\. SVG as background image You can also set an SVG as the background image for an HTML element using CSS. This can be useful if you want to use the SVG as the background for a specific element, such as a div or a button ```html ``` Choosing the right method for embedding SVGs depends on your specific needs and desired use in your web page. Try out different methods and choose the one that best suits your requirements. ## My personal view on SVG In a world increasingly focused on digital visuals, SVG offers a powerful and flexible alternative to traditional image formats. Its ability to present images at any scale without losing quality, coupled with its compact file size and flexibility in handling text, makes **SVG an essential tool for any digital professional**. However, like any technology, SVG should be used correctly. This means avoiding raster image embedding and carefully considering how text is included is essential for optimising performance and compatibility. With a proper understanding of these elements, SVG can take your digital graphics to the next level, resulting in a sharper, more responsive and user-friendly user experience. ### The difference between raster and vector graphics URL: https://rob-hoeijmakers.ghost.io/the-difference-between-raster-and-vector-graphics/ Last updated: 2026-04-14T14:20:23.000Z Below, I will give you a clear explanation of these two types of images and discover which format is best to choose for different situations. I'm sure this information will come in handy to you sooner or later, it's a constant feature in almost all digital productions. ## Raster images: pixels in action Let's start with raster images. Raster images consist of a grid of small squares, which we call pixels. Each pixel has a specific colour value and together they form the visual representation of an image. Think of pictures you take with your digital camera. If you look very closely, you can see the individual pixels that make up the image. The important feature of raster images is that they are made up of a fixed number of pixels. When enlarging a raster image, the pixels are simply stretched, which can result in a blurry or blocky image. This is because the original pixels are magnified, making the image less sharp. ![Imagining a grid.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/06/IMG_2716.jpg) The grid made visible. ## Vector graphics: perfection without limits Now we move on to vector graphics. Vector graphics are different from raster graphics because they are not based on pixels. Instead, they are formed by mathematical equations that define points, lines and curves. This means that vector graphics are scalable, meaning they can be enlarged or reduced without loss of quality. The uniqueness of vector graphics is that they offer perfection without limits. You can endlessly resize a vector image and still keep its sharpness and clarity. This makes **vector graphics ideal for creating logos, icons and other graphic elements** where accuracy and flexibility are important. ![Example of an illustration with vectors.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/05/vector_graphic_416.png) Example of an illustration with vectors. ## Which format is the best choice Now that you understand the difference between raster and vector graphics, it's time to decide which format is best for different situations. Here are some guidelines: 1. **Raster images (JPG, PNG, GIF and WebP)**: Raster images are ideal for capturing photographs and detailed images. They are suitable for situations where the exact representation of pixels is important, such as photography, website images and digital artwork. JPG is often used for photos with many colour gradations, while PNG is suitable for images with transparency and GIF for animated images. WebP is an emerging image format developed by Google that supports both lossy and lossless compression. It often offers a smaller file size than JPG and PNG, allowing web pages to load faster. 2. **Vector graphics (SVG, AI, EPS)**: - **SVG (Scalable Vector Graphics)**: [SVG is a popular vector format for web use](https://rob-hoeijmakers.ghost.io/what-is-svg-and-what-is-it-used-for/). It offers scalability without losing quality, which means you can endlessly resize SVG images without them becoming blurry. SVG images are ideal for logos, icons and graphics on websites. They can also be animated using CSS or JavaScript. - **AI (Adobe Illustrator)**: AI is the native file format of Adobe Illustrator, a widely used graphic design programme. AI files contain vector-based information and are useful for creating and editing complex illustrations, logos and graphic designs. - **EPS (Encapsulated PostScript)**: EPS is a versatile vector format that can be used in both print and digital media. EPS files can be opened and edited with various design software and are suitable for print materials such as brochures, posters and business cards. When choosing the right format, it is important to consider the intended use of the image. If you want to use an image on a website, I recommend choosing raster images such as JPEG or PNG for photography and detailed images, and SVG for scalable graphics. For print and design, vector graphics such as AI and EPS are more suitable because of their scalability and editability. Keep in mind that the right format depends on the specific needs of your project. Consider the intended use, scalability and desired quality when selecting the format. With the right choice, you can ensure that your images look their best and load quickly, whether they are displayed digitally or printed. 💡 In an electronic book, the text is usually stored as vector-based information. This means that the text is scalable and customisable without loss of quality. This allows the text to adapt to different screen sizes and allows readers to customise the text experience to their needs. An electronic book may contain a combination of vector- and raster-based elements, where vector-based text elements are used for scalability and adaptability, while raster-based graphics are used for visual display. ### A personal experience with phishing URL: https://rob-hoeijmakers.ghost.io/a-personal-experience-with-phishing/ Last updated: 2026-04-25T13:17:14.000Z As digital professionals, we are well aware of the myriad threats that await us online. Sometimes, no matter how vigilant we are, we may fall into a cleverly designed trap. Today, I share a recent experience that once again highlights the importance of two-step verification (2FA). ## An unusual email It all started when I received an email on my iPhone informing me that a domain I manage was about to expire. The sender appeared to be my internet hosting provider's customer service, which I did not question at the time. ![Urgent message from hosting provider.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/05/phishing.jpg.webp) Urgent message from hosting provider. ## A perfectly designed message The content of the e-mail was professionally and flawlessly formatted, without the usual spelling mistakes or strange sentence structures that usually ring alarm bells in phishing attempts. Without thinking, I clicked on the link in the email and landed on a login screen that looked perfectly legitimate. Systems like ChatGPT have greatly improved phishing attacks. Whereas language errors and strange wording were still an indication of deception, that will soon be a thing of the past. ## The trap of phishing I use my iPhone to store my passwords, so I was looking for my login details for the website in question. Since my phone did not recognise the website, it took me a while to find the right credentials. Once I logged in, I suddenly realised that I had possibly fallen into a trap. ## The saving second factor This is where the [second factor of two-step authentication (2FA)](https://rob-hoeijmakers.ghost.io/two-factor-authentication-2fa/) proved its true value. In the context of 2FA, "second factor" refers to an extra layer of security required in addition to the default password. A second something. This can be one of the following: - Something you know (e.g. a personal PIN or answer to a security question) - Something you have (e.g. a physical key, smart card or a device that generates a temporary code as in my case) - Something you are (e.g. a biometric identifier such as a fingerprint or facial recognition) Although I might have disclosed my password to a malicious party, I had secured my account with 2FA. This means that despite possessing my password, a potential intruder would not be able to log in without the second authentication factor. In my case, a [unique numeric code generated by my computer or mobile](https://rob-hoeijmakers.ghost.io/2fa-within-apple-keychain/). ![Example of the code presented](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/05/DraggedImage-8.png) Example of the code presented ## The importance of unique passwords Moreover, another habit that helped me was generating a unique password for each account. Even if the phisher had managed to access my password, they could not have used it to access my other accounts. ## Conclusion: the value of 2FA While this was an unpleasant experience, it also served as a useful reminder of the importance of good security practices, such as 2FA and unique passwords. I share this experience in the hope that it serves as a warning to others to be careful and to emphasise the importance of two-step verification. It may sometimes take some extra time and effort, but this extra layer of security can make the difference between security and risking losing valuable data and privacy. In the face of increasing cyber threats, [2FA](https://rob-hoeijmakers.ghost.io/two-factor-authentication-2fa/) is not only recommended - it is an absolute necessity. Incidentally, of course, I immediately changed the password for security. ### The influence of blue light: how to protect your eyes and sleep? URL: https://rob-hoeijmakers.ghost.io/the-influence-of-blue-light-how-to-protect-your-eyes-and-sleep/ Last updated: 2026-04-25T13:16:38.000Z In a digital age, where we spend more and more hours behind different screens, it is important to think about how these screens affect our health. A much-discussed topic is the blue light coming from LED screens of computers, mobile phones and e-readers like the Kobo. But what is the real impact of this blue light on our health and sleep? And how can we minimise this influence? ## Blue light and the sleep cycle The blue light emitted by LED screens is essentially not harmful. However, it is associated with a possible disruption of our sleep cycle. Blue light suppresses the production of melatonin, the hormone that helps us sleep, more than any other type of light. As a result, exposure to blue light, especially in the hours leading up to bedtime, can lead to difficulty falling asleep or a disrupted sleep pattern. ## Less blue light on Mac and iPhone This is why Apple, for example, introduced the "Night Shift" feature, which automatically adjusts the colour of the screen to the warmer end of the spectrum after sunset. This reduces the amount of blue light emitted by the screen, reducing its impact on the sleep cycle. ![Night Shift settings on the Mac.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/05/DraggedImage-6.png) Night Shift settings on the Mac. ## Less blue light on e-reader For those who, like me, like to read on an e-reader (in the dark before sleeping on a Kobo Clara), there is another, possibly even more effective solution: ComfortLight Pro. Reading on the Kobo Clara uses physical black or white dots (e-ink). The light visible from the screen really only works as illumination of the screen. But this illumination is LED and also consists of a combination of red, green and blue light (RGB) like the screen of a mobile phone. Comfortlight Pro makes the screen light warmer and less bright (more natural) as the set sleep time approaches. This not only reduces the amount of blue light, but also reduces overall eye strain, especially during prolonged use. Setting the ComfortLight Pro properly is crucial to maximise these benefits. [![The settings for more natural light on the Kobo Clara.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/05/IMG_1729-Large.jpeg)](https://help.kobo.com/hc/nl/articles/360017481174) The settings for more natural light on the Kobo Clara. ## Blue light in perspective While it is important to be aware of the potential impact of blue light on our sleep, the amount of [blue light emitted by LED screens](https://rob-hoeijmakers.ghost.io/invention-of-blue-led/) should also be put into perspective. Comparisons with indirect sunlight show that the absolute amount of light emitted by LED screens is very modest. Moreover, although LED screens have a peak blue light, this should not be exaggerated. [![LED light compared with indirect sunlight. ](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/05/fires-bluelight-05.jpg)](https://www.ies.org/research/fires/a-reality-check-on-blue-light-exposure/) LED light compared with indirect sunlight. ### Catching the magic of magenta with your iPhone URL: https://rob-hoeijmakers.ghost.io/catching-the-magic-of-magenta-with-your-iphone/ Last updated: 2026-04-25T08:32:12.000Z Updated: 18 september 2024 Colour is an intriguing phenomenon that enriches and influences our daily lives. From the soft green of nature to the vibrant blue of the sky, colours arouse emotions and create atmospheres around us. One colour that often evokes fascination and wonder is magenta. But did you know that magenta does not actually exist as a single colour in the spectrum of light? ## The colour spectrum and monochromatic light To understand the mystery of magenta, we must first take a look at the colour spectrum. The visible colour spectrum, as represented in a rainbow, consists of different colours ranging from red to violet. When white light is refracted through a prism, it is split into these colours, resulting in the colour spectrum. ![Colour spectrum. Without magenta.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/05/DraggedImage-5.png) Colour spectrum. Without magenta. Monochromatic light, on the other hand, consists of a single wavelength or colour. In the case of magenta, which is between red and blue, it is caused by a specific combination of wavelengths. This means that magenta is actually a colour perceived in our brain as a result of the fusion of different colours in the light spectrum. When the brain detects a mix of red and blue **but no green**, it fills the gap by creating the colour magenta—a psychological response to the absence of green light between the two wavelengths. > Purple is purely a mental construct of the human brain. > > We invented the color because of a lack of a sensor signal. [pic.twitter.com/Dxc2aQeO5R](https://t.co/Dxc2aQeO5R) > > — Brian Roemmele (@BrianRoemmele) [September 18, 2024](https://twitter.com/BrianRoemmele/status/1836268556570497284?ref%5Fsrc=twsrc%5Etfw) How magenta is made completely in the mind. ## The mystery of the purple line When we look at a traditional ring-shaped colour spectrum, we see a seamless transition from red to violet. This gradual change implies that magenta (red arrow pointing at it) should be part of the spectrum. ![Colour circle with seamless transition between red and blue.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/09/Kleurencirkel2.png) Colour circle with seamless transition between red and blue. However, if you study the spectrum closely, you will notice that there seems to be a gap between red and violet. This gap is called the "purple line" and represents the missing part of the spectrum where magenta should normally lie. The view below shows it well. The red line represents the monochromatic colours and their wavelengths. Between violet and red a line called the purple line. ![Colour spectrum showing the "Purple Line".](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/05/DraggedImage-1-1.png) Colour spectrum showing the "Purple Line". ## Photographing magenta with an iPhone Now that we understand why magenta is such a mysterious phenomenon, we can turn to the challenges in [photographing this colour with an iPhone](https://rob-hoeijmakers.ghost.io/taking-photos-with-the-iphone-12-mini-a-powerful-camera-in-a-compact-size/). An iPhone's auto focus is designed to respond to the most common colours in the visible spectrum. Because of magenta's unique properties, formed by the combination of red and blue, the camera can sometimes struggle to find the right focus point. To successfully capture magenta tones, it may help to apply a little trick. By **placing your finger next to the subject**, letting the camera focus on your finger and then withdrawing your finger before taking the picture, you can help the camera find the right focus point. This trick has often helped me with photographing lilac, magenta and purple flowers in particular. ![The same trick with very small subjects. Then I sometimes leave the finger in the picture to indicate scale.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/05/IMG_0793-scaled.jpeg) The same trick with very small subjects. Then I sometimes leave the finger in the picture to indicate scale. Photographing magenta tones with an iPhone can be a challenge, but with some patience and experimentation you can achieve beautiful results. Capturing the magic and beauty of magenta in images can be a worthwhile artistic challenge. So, the next time you want to capture the enchanting colours of magenta, you now know how to make the most of your iPhone camera. ![This one does seem to float away from the image. ](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/05/IMG_4864-scaled.jpeg) This one does seem to float away from the image. ### Reading web pages on a Kobo e-reader: a guide for Mac and iPhone URL: https://rob-hoeijmakers.ghost.io/reading-web-pages-on-a-kobo-e-reader/ Last updated: 2026-04-14T13:21:58.000Z This guide is intended for users with a Mac or iPhone/iPad. I assume connecting your Kobo e-reader to a computer. There is a way to connect an iPhone to an e-reader, but for that I refer you to [a dedicated article](https://rob-hoeijmakers.ghost.io/load-free-ebooks-on-a-kobo-using-the-iphone/). ## Creating a readable PDF on the Mac: 1. Open Safari on your Mac and navigate to the web page you want to read on your Kobo e-reader. 2. Click the Reader icon in the address bar to put the web page in "Reader view". 3. Adjust the font size by pressing "Command" and "+" eight times until the text is comfortable to read. 4. Go to the menu bar, click "Archive" and choose "Export as PDF". Save the PDF file in a folder of your choice. Create a readable PDF on the Mac ## Create a readable PDF on the iPhone/iPad: 1. Open Safari on your iPhone/iPad and navigate to the web page you want to read on your Kobo e-reader. 2. Put the web page in "Reader mode" by tapping the Reader icon in the address bar. 3. Choose to print by tapping the share icon (the square with the upward arrow) and selecting "Print". 4. Adjust the font size by entering, for example, 150% in Enlarge/Reduce. 5. Choose share icon again and select "Save to Files" and save the PDF file in a folder of your choice. Create a readable PDF on the iPhone. ## Connect Kobo e-reader to computer 1. Connect your Kobo e-reader to your computer using a USB cable. Make sure your e-reader is switched on. 2. Drag and drop the PDF file to your Kobo e-reader's drive. You can choose any folder. 3. Disconnect your Kobo e-reader from the computer. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/03/IMG_1197-2.jpg) You can now enjoy reading the longer web page on your Kobo e-reader! ### Where does the word "benchmark" come from? URL: https://rob-hoeijmakers.ghost.io/where-does-the-word-benchmark-come-from/ Last updated: 2026-05-10T08:54:18.000Z As a digital professional, it is always important to keep up-to-date of the latest technologies and trends. But certainly it is also valuable to look back at the history of technology to understand how we got to where we are today. In my daily practice, the word "benchmarking" certainly comes up a few times. In this article, I explain how the literal concept of "benchmarking" originated and figuratively incorporated into standard language. ## Landmeters In the past, a benchmark was a strut for a stake, marked to help surveyors determine the height of the terrain. This was a crucial tool for building roads, railways and other infrastructure. The hoof-shaped iron that supported the stake was called the "bench" and placing the stake on the "bench" was called "benchmarking". ![Benchmarks, comparing relative heights.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/05/DraggedImage.png) Benchmarks, comparing relative heights. So in short, there was a mark of relative height, in it a bench was mounted to set the staff on for further comparison. Here you can see such a bench: ![Strut with a bench on top. This thing was fixed in the carved marker. And thus formed the "benchmark".](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/07/DraggedImage-1.jpg) Strut with a bench on top. This thing was fixed in the carved marker. And thus formed the "benchmark". And an example of a marking: ![Marking where the bench could be attached directly.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/05/DraggedImage-2.png) Marking where the bench could be attached directly. ## Modern use of the term benchmark Today, benchmarking is a common term in the world of digital technology. Benchmarking refers to **the process of comparing performance between competing products or processes**. Or against your own past performance. So by setting a relative bar as a height, as a benchmark, you as a company can improve your products and services. ![Benchmark template from PowerSlides](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/05/Benchmark-Template-4.png) Benchmark template from PowerSlides Benchmarking is also useful in evaluating the performance of digital tools and technologies. For example, comparing the speed of different computer processors, the quality of cameras on smartphones or the usability of software applications. Useful interactive benchmarks where you can interactively compare your own product with others are very useful. For example, this site where you can benchmark your browser on a number of points: [Speedometer is a browser benchmark](https://browserbench.org/Speedometer2.0/) So while the original meaning of benchmarking was in the physical world, today it is a valuable practice for digital professionals. **Where absolute numbers do not always properly describe quality, benchmarks come in handy**. It can help improve the performance of organisations, products and processes, and drive innovation and competition. ⚠️ ****"Benchmarking Against Competitors? Think Again."** [Chief Financial Officer Luca Maestri of Apple](https://www.threads.net/@businessblurbb/post/C%5FLHglxOUwo?xmt=AQGzsscdz4cDtrJd356WcLZrTefWWzw8hwwU4RVC4uLlZg) suggests that benchmarking can lead to "bad ideas" by encouraging companies to merely follow industry norms rather than innovate. Instead of comparing yourself to others, focus on your unique strengths, foster creativity, and drive forward with strategies that truly set you apart. Don't just follow the industry—lead it. --- ### Further reading - [From Blueprints to Bytes: How Original Designs Shaped Modern Technology](https://rob-hoeijmakers.ghost.io/blueprints/) - [What We Mean When We Say “Flywheel”](https://rob-hoeijmakers.ghost.io/flywheel/) - [From Benchmarks to Evals: How We Measure AI and Why It Matters](https://rob-hoeijmakers.ghost.io/ai-benchmarks-and-evals/) ### Easy sharing of your content with the Web Share API URL: https://rob-hoeijmakers.ghost.io/increase-the-shareability-of-your-content-with-the-web-share-api/ Last updated: 2026-04-25T13:16:07.000Z Looking for ways to make your content more easily shareable by users? Learn how to use the Web Share API to add share buttons to your website or PWA so users can easily share your content on their favourite social media platforms, send emails and more! ## What is the Web Share API? As a digital professional, you are probably familiar with the Web Share API, a handy tool that allows users to easily share content from a web application to other apps on their device. But did you know that the Web Share API actually moves content sharing from the browser to the operating system (OS)? ## How does the Web Share API work? Traditionally, content sharing was a task performed by the browser. When a user clicked a share button on a website, the browser opened a share menu with options for sharing content on social media, sending emails or copying a link. But with the Web Share API, content sharing is now moved from the browser to the operating system (OS). This means that when a user clicks on a share button, the browser no longer manages the share menu itself, but uses the device's operating system to open the share menu. ![Example of using the Web Share API on iOS.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/04/DraggedImage-3.png) Example of using the Web Share API on iOS. ## What are the advantages of the Web Share API? This has a number of advantages. First, it means that the share menu looks the same as the share menu used in other apps on the device, creating a consistent user experience. Second, users can now use all available sharing options on their device, not just those available in the browser. Third, it makes for a better experience because the share button takes up much less space. This means that if, for example, a user uses a social media platform that is not available as a share option in the browser, but is available as an app on the device, it can now easily be used to share content from a web application using the Web Share API. The disadvantage is that support is not yet complete. For example, Chrome for MacOS does not support it and you either have to omit it or come up with an alternative ![Web Share API on MacOS](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/02/Share-API-MacOS-2.jpg) Web Share API on MacOS ## Main point The main point to remember is that the **Web Share API moves content sharing from the browser to the operating system (OS)**, giving users a seamless and consistent experience when sharing content from web applications. It provides a standardised way to add share buttons to websites and web applications, making sharing content much easier for users and improving the overall user experience. 💡 I created a share button on my [Dutch language WordPress blog](https://webstrategieblog.nl/verhoog-de-deelbaarheid-van-jouw-content-met-de-web-share-api/) and it uses the Web Share API. It was created via a WordPress Plugin: Super Web Share ### Twitter Lite: Exploring Progressive Web Apps URL: https://rob-hoeijmakers.ghost.io/progressive-web-apps-explained-using-twitter-lite/ Last updated: 2026-05-17T13:39:12.000Z A Progressive Web App (PWA) is a website that combines the functionality and user experience of a native mobile app with the accessibility of a web page. A PWA is built with web technologies such as HTML, CSS and JavaScript, but provides the user with an experience similar to a native mobile app. A PWA works on all devices and operating systems and can be installed on the home screen of a device. ![Twitter Lite installed on iOS.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/09/twitter-x-lite.jpg) Twitter Lite installed on iOS. ## Twitter Lite as an example of a successful PWA Twitter Lite, now X lit, is a good example of a successful PWA. It is designed to improve the experience of Twitter on mobile devices and reduce data consumption. Twitter Lite is optimized for use on slow or limited internet connections and is designed to load quickly, even on older devices. Twitter Lite offers many of the same features as the native Twitter app, such as the ability to post tweets, reply and retweet, send and receive direct messages, and view trending topics. The difference is that Twitter Lite is designed as a PWA, so it can be used as a web page and can be installed on the home screen of a device. Another advantage of Twitter Lite as a PWA is the small size of the app. The native Twitter app is often large and takes up a lot of storage space on a device. Twitter Lite, on the other hand, is designed to take up only a small amount of storage space, making it easier to install and use on older devices with limited storage space. ## Installing Twitter Lite on iOS Twitter Lite can be easily installed on any device with a supported web browser. Follow these steps to install Twitter Lite on your device: 1. Open the web browser on your device and go to [mobile.x.com](https://mobile.x.com). 2. Tap the "Share" button in the browser (usually found in the menu) 3. Choose "Add to home screen" or a similar option 4. Customise the app name if desired and tap "Add" or "Save" ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/07/IMG_5279-1.png) The steps to install PWA on iPhone Twitter Lite is now installed on your home screen and ready to use. ![Dialogue after selecting share on iOS](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/04/image-1.png) after selecting share on iOS ## My conclusion A Progressive Web App is a website that provides the user experience of a native mobile app and is accessible on all devices and operating systems. Twitter Lite is a prime example of a successful PWA because it is designed to load quickly, works on slow or limited internet connections, can work offline and takes up only a small amount of storage space. Twitter Lite offers many of the same features as the native Twitter app, but provides a better experience for users who want to use Twitter on their mobile devices. --- [*Subscribe for free*](https://rob-hoeijmakers.ghost.io/newsletter/) *and you receive the latest posts. For instance if* [*X Premium is worth the money*](https://rob-hoeijmakers.ghost.io/paid-subscription-on-x/)*.* ### Streamline your online meetings with AI note taking URL: https://rob-hoeijmakers.ghost.io/streamline-your-virtual-meetings-with-ai-note-taking/ Last updated: 2026-05-10T08:54:18.000Z As digital professionals, online meetings have become an integral part of our work routine. However, it can be challenging to keep up with the fast-paced nature of these meetings and ensure that all the important information is captured accurately. That's where [Tactiq](https://tactiq.io) comes in - a tool that provides real-time transcription for online meetings on platforms such as Google Meet, Zoom, MS Teams, and Webex. I tested it on a couple of Google Meet meetings and was very impressed. Tactiq's real-time transcription feature is simply magical! It accurately captures every word spoken during the meeting, including the nuances and inflections of the speaker's voice. I was blown away by the precision and usefulness of the tool when I tested it on a 1:1 conversation. After the meeting, Tactiq provided me, through [ChatGPT](https://rob-hoeijmakers.ghost.io/try-out-chatgpt-yourself/), with a **summary, action items, and an agenda**, which saved me a lot of time and effort in creating meeting notes. Tactiq's real-time transcription is not only efficient but also a time-saver! As a digital professional, time is of the essence, and Tactiq saves me a lot of time in creating meeting notes. The tool's AI-based algorithm accurately identifies key points and highlights them in the summary, making it easy to quickly review the most important information discussed during the meeting. In my opinion Tactiq is a game-changer for online meetings. Its real-time transcription feature is accurate, versatile, and incredibly useful for improving productivity during online meetings. While it's important to keep privacy and security concerns in mind, Tactiq is definitely worth trying for anyone who wants to streamline their online meetings and focus on the conversation instead of taking notes. ## Tactiq speaks its languages Tactiq's AI-powered transcription technology is capable of accurately transcribing speech in multiple languages. This is particularly useful for organisations that operate in a global context and conduct online meetings in different languages. Tactiq currently supports several languages, including English, French, German, Spanish, Portuguese, Dutch, and Italian. As a multilingual transcription tool, Tactiq provides users with accurate and reliable transcriptions in their preferred language. This can be particularly useful for non-native speakers who may struggle to follow discussions in a foreign language. Tactiq's transcription accuracy is based on its advanced algorithms and natural language processing capabilities, which analyse and interpret spoken language in real-time. Based on user feedback, Tactiq's multilingual capabilities have received positive reviews, with many users reporting accurate transcriptions even when conducting meetings in languages other than English. With its support for multiple languages, Tactiq is a valuable tool for global organisations looking to streamline their virtual meetings and improve productivity across different languages and cultures. ![I tested Tactiq in Google Meet with the Dutch language.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/04/2023-04-18_15-18-15.png "I tested Tactiq in Dutch and it passed the test.") I tested Tactiq in Google Meet with the Dutch language. ## Pricing Here is a summary of Tactiq's pricing plans: - **Free**: Offers up to 10 meetings per month, 5 AI credits, time-stamped notes, basic integrations, speaker identification, highlights, in-call chat saving, PDF export, and basic team collaboration. - **Pro**: Costs $8/month and provides unlimited meetings, 10 AI credits per month, screenshots, video add-on, auto-highlights, in-call chat saving, meeting engagement, PDF export, basic team collaboration, and more. - **Team**: Costs $16/month and includes all the features of the Pro plan along with 100 AI credits per month, team folders, centralised billing, collaborative workspace, membership admin, advanced collaboration, bulk export, and team engagement. [I tested Tactiq with the free plan.](https://tactiq.io) ## Summary - Tactiq is a real-time transcription tool for virtual meetings on popular platforms like Google Meet, Zoom, MS Teams, and Webex. - Tactiq accurately captures every spoken word, including nuances and inflections, saving time and effort in creating meeting notes. - Tactiq's AI-based algorithm identifies key points and highlights them in the summary for easy review. - Tactiq supports multiple languages, providing accurate transcriptions for non-native speakers and global organisations. - Tactiq offers different pricing plans to suit different needs and budgets --- ### Further reading - [Fireflies.AI: Meeting Transcriptions & Automation](https://rob-hoeijmakers.ghost.io/fireflies-ai-transcriptions/) ### Apple iOS devices now support web push notifications URL: https://rob-hoeijmakers.ghost.io/apple-devices-web-push-notifications/ Last updated: 2026-04-25T13:46:06.000Z If you're a blog writer, you'll be happy to know that Apple iOS devices now support Web Push notifications, a standard notification service for time-sensitive and high-priority events. With the **release of Safari 16.4**, users on Apple iOS devices can receive Web Push notifications, making it easier for you to keep your readers informed about new articles and updates. Web Push notifications are short, instant notifications that appear on a user's desktop or mobile device, even when they are not actively using your website. They are a powerful tool for engaging readers and keeping them informed about new articles and updates. With the addition of Web Push support in Safari on Apple devices, blog writers can now use this powerful tool to reach even more readers. ![Example of Push Notification on iOS](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/04/IMG_0599.jpeg) Example of Push Notification on iOS ## Instant delivery One of the biggest advantages of Web Push notifications is their instant delivery. When you publish a new article, your readers will know about it immediately, no matter where they are or what they are doing. This can help to drive traffic to your website and increase engagement with your content. Web Push notifications are also highly visible, appearing directly on the user's device, making them hard to miss. And because they are opt-in, users can choose to receive them with just one click, without having to provide their email address. This makes them a more convenient and less intrusive alternative to email newsletters. ![Example of Web Push Notification on macOS](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/04/IMG_0598-1.jpeg) Example of Web Push Notification on macOS Another advantage of Web Push notifications is their customizability. You can customize the content and timing of your notifications to ensure that they are relevant and engaging for your readers. This can help to increase open rates and click-through rates, leading to more engagement with your content. With the addition of Web Push support in Safari on Apple devices, blog writers can now take advantage of all of these benefits to reach a wider audience and keep their readers informed about new articles and updates. So if you're a blog writer, be sure to explore the possibilities of Web Push notifications and see how they can help you to engage your audience and grow your readership. [How I Tuned My Social Notifications into a Single, Silent StreamDuring the holidays, I rethought notifications. The result: one place, no noise, no rush. A calmer way to stay connected without being pulled in.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-196.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_5148.jpeg)](https://rob-hoeijmakers.ghost.io/apple-notifications/) ## Overcoming iOS Challenges While the iOS 16.4 update has made web push notifications possible, it's important to note that the process of implementing them on Apple devices is not as straightforward as on other platforms. Due to Apple's unique approach to user privacy and security, integrating web push notifications on iOS requires additional steps, such as asking users to add your website as a Progressive Web App (PWA) to their home screen. Steps to take to receice Web Push notifcations on iOS. This extra layer of user interaction may impact the adoption rate of web push notifications for some websites, and it's crucial to take this into consideration when planning your engagement strategy. Nonetheless, the opportunity to reach iOS users with web push notifications is an exciting development that shouldn't be overlooked. 💡 How do you get notifications on iOS? Open "share" with [this blog](https://webstrategieblog.nl) in Safari and choose Put on home screen. This adds an app and when you open it, you can subscribe to push notifications through [****OneSignal**](https://onesignal.com/blog/ios-web-push-is-finally-coming/amp/?ref=hoeijmakers.net). Read more about [Progressive Web Apps (PWA's)](https://rob-hoeijmakers.ghost.io/progressive-web-apps-explained-using-twitter-lite/), also referred to as Home Screen apps. 🔴 ****Difference Safari Push Notifications and Web Push Notifications** Safari push notifications are primarily associated with websites and the Safari browser, while iOS web push notifications are associated with home screen apps (PWA) installed on Apple devices. Both require user opt-in and are sent by push providers, but the context and implementation differ based on the platform they are designed for. ### Web Push versus Email newsletters URL: https://rob-hoeijmakers.ghost.io/web-push-vs-email-newsletters/ Last updated: 2026-04-25T13:16:36.000Z As a website owner, you want to keep your readers informed about new articles and updates, but which method should you use: Web Push notifications or email newsletter subscriptions? Both methods have their advantages and disadvantages, so let's take a look at how they compare. ## Web Push Notifications Web Push notifications are short, instant notifications that appear on a user's desktop or mobile device, even when they are not actively using your website. Here are some of the advantages of using Web Push notifications for sending article updates: - **Instant delivery**: Web Push notifications are delivered instantly, so your readers will know about new articles as soon as they are published. - **High visibility**: Web Push notifications appear directly on the user's device, making them hard to miss. - **No email address required**: Users can subscribe to Web Push notifications with just one click, without having to provide their email address. - **Customizable**: You can customize the content and timing of your Web Push notifications, to ensure that they are relevant and engaging for your readers. However, there are also some disadvantages to using Web Push notifications: - **Limited message length**: Web Push notifications are limited in length, so you may not be able to provide as much information as you would in an email newsletter. - **Opt-in required**: Users must opt-in to receive Web Push notifications, and they can opt-out at any time, so you may have a smaller audience than with email newsletters. I will use the (free) service of [OneSignal](https://onesignal.com) to find subscribers and push notifications on updates. ![Push notification subscription on my blog.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/04/2023-04-15_08-41-47.png) Push notification subscription on my blog. ## Email Newsletter Subscriptions Email newsletters are regular email updates that are sent to subscribers who have opted-in to receive them. Here are some of the advantages of using email newsletters for sending article updates: - **Longer message length**: Email newsletters allow you to provide more detailed information about your articles, including images and links. - **Direct communication**: Email newsletters are sent directly to the user's email inbox, making them a more personal and direct form of communication. - **No device restrictions**: Email newsletters can be accessed on any device with an internet connection, making them accessible to a wider audience. However, there are also some disadvantages to using email newsletters: - **Delayed delivery**: Email newsletters may not be delivered instantly, so your readers may not know about new articles as soon as they are published. - **Spam filters**: Email newsletters can sometimes be caught by spam filters, causing them to be delivered to the user's spam folder or not delivered at all. - **Email fatigue**: Users may receive a large number of emails, leading to email fatigue and causing them to unsubscribe. ![Email subscription on my blog.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/09/2023-04-15_08-40-18.jpg) Email subscription on my blog. ## Experiment: Which method works better? Both Web Push notifications and email newsletter subscriptions have their advantages and disadvantages when it comes to sending article updates. I consider to use them both but it might become pushy and confusing very quickly. As an experiment, I will be **comparing the effectiveness** of Web Push notifications and email newsletter subscriptions in the context of sending article updates. I will be using Web Push notifications to notify subscribers about new articles on my website, and email newsletters to notify subscribers about the same articles. I will be tracking the number of subscribers who opt-in to each method, the open rates and click-through rates for each notification, and the overall engagement of each audience. I will then evaluate the effectiveness of each method and report back on my findings. Stay tuned ([subscribe](https://rob-hoeijmakers.ghost.io/#/portal/signup) or receive [push notifications](https://webstrategieblog.nl) 😉) for the results of this experiment, which will provide insights into the best way to reach and engage your audience with article updates. ### From blog to book, publishing in the digital age URL: https://rob-hoeijmakers.ghost.io/from-blog-to-book-publishing-in-the-digital-age/ Last updated: 2026-04-25T13:17:11.000Z Tim Urban, author of the popular blog Wait But Why, has released a new book, "What's our problem". What makes this book unique is not only its content, but also the way it is presented. Urban has chosen to release the book in a purely digital format and also offers a digital audiobook. A nice touch is that Urban herself narrates the audiobook. ## Landing page On the book's landing page, found on Wait But Why, potential readers can purchase the book through various online platforms such as Apple, Google, Kobo, Kindle and Spotify. Besides these online platforms, the book can also be bought and downloaded directly as Epub or MP3. One of the most striking aspects of the landing page is the preview available for different devices and screen sizes. This allows potential readers to already get an idea of how the book looks on their device. Also charming is the thumbnail overview of the entire book on the landing page, giving readers an idea of the book's structure and content. ## Publish ebooks yourself Urban himself seems to be the publisher of all this, showing that it is possible to publish a book as an individual without the intervention of a traditional publisher. This makes the book an interesting example for people who also want to self-publish a book. It proves that you can use digital technologies to create a unique reading experience and create your own professional and informative landing page to promote and sell your book. All in all, "What's our problem" by Tim Urban is not only worthwhile in terms of content, but also exciting from a digital publishing perspective. The book shows that it is possible to publish a book as an individual using digital technologies to provide a unique reading experience. Check out the landing page: [What’s Our Problem? A Self-Help Book For SocietiesAvailable today![](https://149909199.v2.pressablecdn.com/wp-content/themes/waitbutwhy/images/favicon.ico)Wait But Why![](https://wait-but-why-production.mystagingwebsite.com/wp-content/uploads/2023/02/Feature-FB.png)](https://waitbutwhy.com/whatsourproblem) I myself chose to order through Apple Books. I did the ordering through the computer and that was literally two clicks. Now I can read the book on the computer, tablet and/or phone. ![Ordering the book through Apple Books](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/04/2023-04-06_10-04-14.webp) Ordering the book through Apple Books ### ChatGPT and the art of asking questions URL: https://rob-hoeijmakers.ghost.io/the-art-of-asking-questions/ Last updated: 2026-04-25T13:17:10.000Z Asking questions is a fundamental aspect of human communication, and it is one of the most effective ways to deepen our understanding of the world around us. From learning in the classroom to driving innovation in the workplace, the ability to ask meaningful questions is a skill that is critical to success in every area of life. ## Sculpting the unknown But what exactly is it that we do when we formulate a question? At its core, the process of formulating a question involves a combination of cognitive, linguistic, and social skills. Just like a sculptor carving a statue out of stone, we start with a vague idea or a gap in our knowledge, and use our cognitive and linguistic skills to craft a well-formed question. Like a sculptor, we need to be patient and precise as we work to refine and perfect our question. We may need to consider different angles, revise our approach, and experiment with different phrasings in order to arrive at the best possible version of our question. And just as a sculpture is the result of a collaborative effort between the artist and the material they are working with, a well-formed question is the product of a collaboration between our thoughts and our language, and a reflection of our curiosity and desire for knowledge. ![Sculpture with funny hat.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/03/A253BA05-F22D-45F8-9CEA-2ED337B52283_1_105_c.jpeg) Sculpture with funny hat. ## Value of a well put question The value of a well put question is undeniable. It is an incredibly valuable tool in many different areas of life, including **learning, problem-solving, relationship-building, and driving innovation**. When we formulate a question, we typically start by identifying a knowledge gap or area of uncertainty that we want to explore. We then use language processing skills to select the appropriate words and grammar to convey our question in a clear and concise manner. 💡 There are these four ways of answering questions. Which four? There are questions that should be answered categorically \[straightforwardly yes, no, this, that\]. There are questions that should be answered with an analytical (qualified) answer \[defining or redefining the terms\]. There are questions that should be answered with a counter-question. There are questions that should be put aside. These are the four ways of answering questions. Next, we engage in **critical thinking** to ensure that our question is relevant, specific, and meaningful. We may consider alternative ways to phrase the question, or adjust the wording to better reflect our intentions. Once we have formulated the question, we communicate it to others, either through speaking or writing, with the goal of eliciting a response or answer. > The riddle does not exist. If a question can be put at all, then it can also be answered. > Ludwig Wittgenstein ## Questions in the era of ChatGPT By mastering the art of asking questions, we can enhance our learning, deepen our understanding, and drive innovation in every area of life. Questions are the keys that unlock doors to the unknown, and with the rapid advancement of AI-powered language models like [ChatGPT](https://rob-hoeijmakers.ghost.io/try-out-chatgpt-yourself/), the importance of asking the right questions has never been greater. By skilfully using questions to gather insights and drive innovation, we can expand our knowledge and unlock new opportunities in every field. The art of asking questions is a skill that is essential for success in every aspect of life. By honing our cognitive, linguistic, and social skills, we can become more effective communicators, problem-solvers, and innovators. Asking good questions is a skill that can be developed with practice and effort. By **cultivating curiosity, reading** you can become a skilled questioner who deepens their understanding of the world and drives innovation in every area of life. So, embrace the power of the question, keep learning, and keep asking! ### Review: This Is Marketing by Seth Godin URL: https://rob-hoeijmakers.ghost.io/book-review-this-is-marketing-by-seth-godin/ Last updated: 2026-05-14T21:01:32.000Z In this book, Godin shares his decades of experience in the industry and provides practical insights and strategies for marketers of all levels. One of the key themes of the book is the **importance of empathy in marketing**. Godin emphasizes that successful marketing requires understanding the needs, desires, and fears of your target audience. By putting yourself in their shoes and creating products and services that truly address their needs, you can build a loyal customer base and establish your brand as a trusted authority in your industry. ## Storytelling Another important lesson from the book is the power of storytelling in marketing. Godin argues that humans are hardwired to respond to stories, and that the best marketing campaigns are those that **tell a compelling narrative that resonates with their audience**. Whether it's through a blog post, a social media campaign, or a traditional ad, the key is to create a story that captures the attention and imagination of your audience. What sets "This Is Marketing" apart from other marketing books is Godin's playful and inspiring writing style. He uses anecdotes, metaphors, and humour to illustrate his points and keep the reader engaged. The book is also highly practical, with plenty of actionable advice and exercises that you can apply to your own marketing efforts. ## Status In his book "This Is Marketing," Seth Godin highlights the practical social phenomenon of marketing, which involves creating and spreading ideas, building relationships, and making a positive impact on people's lives. One key aspect of this phenomenon that Godin emphasizes is the concept of status and its role in shaping consumer behaviour. Status is a powerful motivator for many consumers, and Godin argues that understanding this dynamic is key to successful marketing. People are often drawn to products or services that convey a sense of status, whether it's through luxury branding, exclusive access, or social signalling. Godin notes that status is not just about material possessions, but rather it's **a social construct that helps people define their place in the world**. By understanding the different types of status that people seek, marketers can tailor their messaging and branding to appeal to these desires. For example, a luxury car brand may appeal to consumers who are seeking status through conspicuous consumption, while a wellness brand may appeal to consumers who are seeking status through health and fitness. By understanding these different motivations, marketers can create messaging that resonates with their target audience and encourages them to take action. However, Godin also notes that status is not the only motivator for consumers, and that focusing too heavily on status can backfire. In today's socially conscious climate, consumers are increasingly seeking brands that align with their values and make a positive impact on society. ![Word cloud that shows the content of the iconic book This is Marketing by Seth Godin](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/03/IMG_9986.jpeg) Word cloud that shows the content of the iconic book This is Marketing by Seth Godin ## Conclusion For me, "This Is Marketing" is a must-read for anyone looking to improve their marketing skills and stay ahead of the curve in today's ever-changing business landscape. Whether you're a seasoned marketer or just starting out, Godin's insights and strategies are sure to inspire and inform. And Godin is a great storyteller, he speaks to you, he is almost tangible. A remarkable talent and with it, he does what he writes about, he sells you the message with transparency and empathy. ### Start with Midjourney URL: https://rob-hoeijmakers.ghost.io/start-with-midjourney/ Last updated: 2026-04-25T13:16:35.000Z ## Introduction: Midjourney is an exciting new generative AI for images that has captured the attention of creatives and technophiles alike. The cutting-edge AI technology allows users to create stunning visuals with minimal effort. If you're interested in being one of the first to try out this revolutionary tool, you'll need to join the Midjourney Beta program through Discord. In this article, we'll walk you through the process step-by-step. ## Step 1: Install Discord Before you can join the Midjourney Beta program, you'll need to have Discord installed on your device. Discord is a popular chat platform for communities and friends to connect. Download the Discord app for your device (PC, Mac, iOS, or Android) from the official Discord website (https://discord.com/download) and create an account if you haven't already. ## Step 2: Locate the Midjourney Discord Server To join the beta, you'll need to find the official Midjourney Discord server. You can search for the server by visiting the Midjourney website or social media profiles, where they'll likely share an invite link. If you're having trouble finding the invite link, you can also try searching online forums or asking for help from fellow AI enthusiasts. [MidjourneyAn independent research lab exploring new mediums of thought and expanding the imaginative powers of the human species.![](https://www.midjourney.com/favicon-32x32.png)Midjourney](https://www.midjourney.com/) Midjourney website ## Step 3: Join the Midjourney Discord Server Once you've found the invite link, click on it to join the Midjourney Discord server. You'll be redirected to the server, and you may need to complete a verification process (such as solving a captcha) to gain full access to the server channels. ![Join the Midjourney Discord Server](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/09/midjourney.jpg) Join the Midjourney Discord Server ## Step 4: Locate Beta Program Information In the Midjourney Discord server, look for a channel or a pinned message that contains information about the beta program. The channel might be named something like "#beta-info," "#beta-signup," or "#announcements." Read through the provided information to familiarize yourself with the beta program's terms, conditions, and prerequisites. But basically, that is it, you are in! ## Step 5: Start Select any newbies-# channel visible in the left sidebar and give a prompt to start imagining: 1. Type /imagine prompt: or select the /imagine command from the slash commands pop-up. 2. Type a description of the image you want to create in the `prompt` field. 3. Click return to send your message. ## Conclusion: Joining the Midjourney Beta program through Discord is a straightforward process that allows you to be among the first to experience this innovative image generation AI. Follow these simple steps, and you'll be on your way to creating captivating visuals with Midjourney's AI-powered technology. If you find starting with Midjourney a step too far, I can recommend using [Playground AI](https://rob-hoeijmakers.ghost.io/start-with-midjourney/). ### WebP vs JPEG: Which Image Format is Best for Your Website? URL: https://rob-hoeijmakers.ghost.io/webp-versus-jpg/ Last updated: 2026-04-14T14:20:40.000Z While WebP is generally considered a superior image format due to its smaller file size and better compression compared to JPEG, there are still some cases where JPEG may be the better choice. Here are a few scenarios where JPEG may be preferred over WebP: - **Compatibility**: While most modern browsers support WebP, there may be some older browsers or devices that do not support the format. In such cases, it may be necessary to use JPEG to ensure that all users can view the images. - **Complexity**: WebP can offer better compression and smaller file sizes, but this also means that the compression process can be more complex and resource-intensive. In cases where time or computing resources are limited, JPEG may be a more practical choice. - **Quality**: While WebP can offer comparable or better quality than JPEG at smaller file sizes, there may be some cases where JPEG is still able to achieve higher quality at a given file size. This may be particularly true for **images with complex or fine details**, where the lossy compression used by WebP may result in noticeable artifacts or blurring. Overall, while [WebP](https://rob-hoeijmakers.ghost.io/webp-avif/) is generally the preferred format for web images due to its better compression and smaller file sizes, [JPEG can still be a viable option](https://rob-hoeijmakers.ghost.io/image-formats-for-the-web/) in certain cases where compatibility, complexity, or quality considerations are a factor. ### A good tool to visually check differences between WebP and is called Squoosh When optimizing images, it's important **to strike a balance between file size and image quality**. While reducing file size can help improve website performance and reduce load times, it's also important to ensure that images are still visually appealing and don't appear too compressed or pixelated. This is where the **slider function in Squoosh** can be especially useful. When optimizing an image in Squoosh, you can use the slider to adjust the quality of the image and see the effects of the compression in real-time. As you adjust the slider, the image will update to show the changes in quality, making it easy to find the optimal balance between file size and image quality. The slider function in Squoosh is especially useful because it allows you to test the image quality at different compression levels without needing to save and re-upload the image each time. This can save a lot of time and make it easier to find the optimal compression level for each image. Check out my article on [Squoosh](https://rob-hoeijmakers.ghost.io/crisp-and-fast-loading-images-with-squoosh/) here. ### Squoosh your images URL: https://rob-hoeijmakers.ghost.io/crisp-and-fast-loading-images-with-squoosh/ Last updated: 2026-04-21T09:12:58.000Z As a web professional, one of the most common questions you may get asked is how to optimize images for the web. With so many different file formats available, it can be difficult to know which one to use and how to ensure that images are optimized for performance without sacrificing quality. Fortunately, there are tools available that can help make this process easier. One such tool is [Squoosh](https://squoosh.app), a web-based image optimization tool that offers a user-friendly interface and a range of optimization options. In this article, we'll take a closer look at Squoosh and the many benefits it offers for optimizing images for the web. [WebP vs JPEG: Which Image Format is Best for Your Website?Find out why WebP might be a better choice than JPEG for your website. Detailed comparison and insights on performance and quality. Plus a nice online tool to make hands on comparisons.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/size/w256h256/2024/05/Logo-Rob-Hoeijmakers-Square-Dark-Glow-3-1.png)Web StrategiesRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/size/w1200/2023/03/IMG_0160.jpeg)](https://rob-hoeijmakers.ghost.io/webp-versus-jpg/) ### Squoosh Squoosh is a web-based image optimization tool that allows you to compress your images for use on the web. The app was developed by Google and is available for free on the web. Squoosh offers a range of features that help you [optimize your images](https://rob-hoeijmakers.ghost.io/measure-page-speeds-with-core-web-vitals/), including resizing, cropping, and converting to different file formats. With Squoosh, you can easily optimize your images without compromising on quality. ## Benefits of using Squoosh There are several benefits to using Squoosh for optimizing your images. Let's take a closer look at some of the key benefits: ### Web-based and free Squoosh is a web-based app, which means you don't need to download or install any software to use it. All you need is a web browser and an internet connection. This makes it very convenient and easy to use. Additionally, the app is completely free, so you don't need to worry about any costs associated with using it. ### User-friendly interface Squoosh has a user-friendly interface that is easy to navigate. The app is designed to be simple and straightforward, even for people who are not familiar with image optimization. The interface includes a range of tools that allow you to optimize your images with ease, including options for resizing, cropping, and converting to different file formats. ### Fast and efficient One of the biggest advantages of using Squoosh is its speed and efficiency. The app is optimized to work quickly and efficiently, allowing you to optimize your images in a matter of seconds. This is particularly useful if you need to optimize a large number of images or if you are working with images that have large file sizes. ### Range of optimization options Squoosh offers a range of optimization options that allow you to customize the optimization process to suit your needs. For example, you can choose to optimize your images for a specific device or screen size, or you can [optimize for a particular file format](https://rob-hoeijmakers.ghost.io/webp-versus-jpg/), such as WebP. The app also allows you to adjust the image quality and compression level to achieve the optimal balance between image quality and file size. ### Support for powerful image formats Squoosh supports a range of powerful image formats, including [WebP and AVIF](https://rob-hoeijmakers.ghost.io/webp-avif/). These formats offer superior image quality and smaller file sizes compared to traditional image formats like JPEG and PNG. By using these formats, you can improve the performance of your website or app by reducing the file size of your images without compromising on quality. ## My personal conclusion on Squoosh Squoosh is an excellent tool for optimizing your images for use on the web. It is web-based, free, and offers a user-friendly interface, making it easy to use for anyone. It is also fast and efficient, with a range of optimization options that allow you to customize the optimization process to suit your needs. Additionally, the app supports powerful image formats like WebP and AVIF, which can help improve the performance of your website or app. Overall, if you're looking for an easy-to-use and effective image optimization tool, [Squoosh](https://squoosh.app) is definitely worth checking out. ### A demo of a chatbot in WhatsApp using ChatGPT URL: https://rob-hoeijmakers.ghost.io/whatsapp-chatbot-chatgpt/ Last updated: 2026-05-26T19:36:35.000Z Chatbots are popular today and are used to provide customer service, answer questions and even become personal assistants. A chatbot is a computer programme that can automatically answer messages and respond to user queries. It works based on artificial intelligence (AI) and machine learning (ML) to learn and improve itself as it has more interactions. ## Language models There are several language models available for chatbots, including ChatGPT, GPT-3, BERT, and more. These language models are trained on huge datasets and can understand what the user is saying and provide a meaningful response to it. While these models can be used for different purposes, they are all aimed at improving the chatbot experience. In this article, we will focus on ChatGPT and how it can be used for WhatsApp chatbots. ## WhatsApp chatbot training ChatGPT is a language model trained based on the transformer model. It is built using a large dataset of text, making it capable of understanding and generating human language. It is one of the most advanced language models available and is used by many chatbot developers. To use ChatGPT for your [WhatsApp](https://rob-hoeijmakers.ghost.io/is-whatsapp-social-media/) chatbot, you must first train the language model with your own dataset. This is necessary because ChatGPT is designed to provide generic answers based on the huge dataset it is trained on. By training your own, you can ensure that your chatbot gives specific and relevant answers based on the information you want to share. To start training [ChatGPT](https://rob-hoeijmakers.ghost.io/try-out-chatgpt-yourself/) for your WhatsApp chatbot, you need to collect a dataset of questions your users might ask and the corresponding answers. This dataset can be used to train the language model so that it can generate specific answers based on the questions asked. Once you have collected your dataset, you need to convert the dataset into a specific format and then train it using the ChatGPT language model. You can use platforms like Hugging Face to train ChatGPT and integrate it into your chatbot. ## Keep training for quality After training ChatGPT and integrating it into your chatbot, you can test your chatbot to see if it works the way you want it to. It is important to remember that training a language model like ChatGPT is an iterative process and that you need to keep training and improving the model constantly to ensure it continues to generate the best responses. Besides ChatGPT, there are other language models available that you can use for your WhatsApp chatbot. One example is GPT-3, which is used by many developers because of its impressive performance. However, it is important to remember that each language model has its own strengths and weaknesses, and it is up to you to decide which language model best suits the needs of your chatbot. Finally, it is important to mention that the use of language models such as ChatGPT and GPT-3 for chatbots is still evolving. There are still challenges in terms of language understanding and generating coherent and relevant responses. Training the language model with a relevant dataset is therefore crucial to ensure the quality of responses. ## Demo Here I give a short demo of a chatbot in WhatsApp using ChatGPT4\. In the demo, I ask some questions about connecting a monitor to a Mac mini. In conclusion, ChatGPT and other language models offer many opportunities for improving the chatbot experience on WhatsApp and other platforms. By using your own dataset to train the language model, you can ensure that your chatbot gives specific and relevant answers based on the information you want to share. Remember that training a language model is an iterative process, so keep training and improving the model to get the best results. *Originally published in Dutch on* [*Chat voor Bedrijven, a blog on chat solutions for companies and organisations*](https://www.chatvoorbedrijven.nl/whatsapp-chatbots-met-chatgpt-een-demo/)*. Interested in setting up your own chatbot using GPT?* [*Reach out and we can discuss options.* ](https://rob-hoeijmakers.ghost.io/about-rob-hoeijmakers/) ### How chatbots understand us URL: https://rob-hoeijmakers.ghost.io/how-chatbots-understand-human-language/ Last updated: 2026-05-26T19:32:28.000Z In the context of question-answering systems and natural language processing, intent refers to the underlying purpose a user has when asking a question or making a statement. Essentially, intent is the reason why the user is communicating with the system or another person. For example, a user may ask the question "How is the weather today?" with the intention of getting information about the current weather conditions in their area. Alternatively, a user might ask the question "How do I fix my computer?" with the intention of getting instructions on how to solve a technical problem. Intention recognition is an important aspect of natural language processing because it allows the system to understand the user's needs and provide an appropriate answer. By identifying the intent of a user's question, a question-answering system can determine what kind of information the user is looking for and provide a relevant answer. ![ChatGPT asked a personal question about intents](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/03/2023-03-01_12-45-07.jpg) ChatGPT asked a personal question about intents ## Intents and a language model An AI language model continuously develops and improves the ability to understand the intent of questions. The training data includes a huge amount of natural language text, allowing patterns and relationships between words and phrases to be recognised. An AI language model also uses a combination of rule-based systems and machine learning algorithms to analyse the context, structure and wording of a question to determine its intent. This approach allows the system to accurately interpret a wide variety of queries in different fields, including science, technology and everyday life. However, there are still cases where the intent of a question cannot be accurately determined, especially when the question is ambiguous or the context is unclear. In such cases, users are asked to provide additional context or clarification to help better understand their question. ## Examples of intents There are countless different intentions someone may have when asking a question, as human language is incredibly complex and diverse. Some examples of common intentions when asking questions are: - **Informational**: the user is looking for specific information, such as the answer to a question or the explanation of a concept. - **Clarification**: the user wants to clarify something they do not understand or are unsure about. - **Instruction**: the user is looking for guidance or instructions to perform a particular task or solve a problem. - **Comparison**: the user wants to compare different options or scenarios in order to make a decision. - **Opinion**: the user seeks the opinion of others on a particular topic or problem. - **Confirmation**: the user seeks confirmation or verification of something he thinks is true. These are just a few examples of the different intentions that may underlie a user's query. As an AI language model, I have been trained to recognise and respond to a wide range of query intentions, and I am constantly learning and adapting to better understand the nuances of human language. ## Entities, building blocks for intents Entities are another important concept in natural language processing that is closely related to intents. While intents represent the general purpose of a user's input, **entities refer to specific pieces of information within the input that are relevant to the intent**. Entities are used to extract important information from the user's input to provide a more accurate and useful answer. For example, if the intent is to book a flight, the relevant entities might be the cities of departure and arrival, the date of travel and the number of passengers. Entities can be considered the building blocks of intents, as they provide the specific details needed to execute the user's request. They can be extracted from the input using techniques such as regular expressions, named entity recognition or machine learning. In general, entities and intents work together to enable chatbots and language models to understand and respond to user input in a way that is relevant and useful. By extracting the specific information needed to satisfy the user's request, chatbots can provide accurate and personalised responses tailored to the user's needs. ## Examples entities Examples quickly make things clearer. Here are some examples of entities that may be relevant in different contexts: - **Dates and times**: Used for scheduling or event-related queries - **Locations**: Used for mapping, navigation or searches for local businesses. - **Names**: Used for contact management or personalisation. - **Numbers**: Used for calculations or quantitative searches. - **Products**: Used for e-commerce or inventory management. - **Organisations**: Used for business searches or company information. - **Monetary amounts**: Used for financial or budget-related searches. - **Quantities**: Used for orders or inventory management. - **Dimensions**: Used for cooking or construction-related queries. - **URLs or e-mail addresses**: Used for web- or e-mail-related searches. These are just a few examples, and the specific entities that are relevant depend on the context and purpose of the chatbot or language model. Entities can be identified using various techniques, such as regular expressions, named entity recognition or machine learning. ## Recent advances in natural language processing Extracting intents and entities from queries can be challenging because natural language is complex and often ambiguous. Understanding the nuances of language requires not only knowledge of syntax and grammar, but also an understanding of context, idiom and cultural references. Moreover, the complexity of the task depends on the specific use case and the variety of inputs the chatbot or language model is expected to process. For example, a chatbot designed to provide customer support for a specific product or service may have a narrower range of possible inputs than a general language model that needs to process a wide variety of queries. However, thanks to advances in natural language processing and machine learning, it is becoming **increasingly possible to accurately extract intentions and entities from user input**. Using large data sets and advanced algorithms, modern chatbots and language models (such as ChatGPT) can recognise patterns in language and make predictions about the intentions and entities most likely to be present in a given input. ## Context of a conversation Although extracting intentions and entities from user input can be challenging, it is a crucial aspect of developing effective chatbots and language models that can provide accurate and personalised responses to user queries. Here, context, in particular, also plays an important role. In the context of chatbots, context refers to **the information or cues available to the system at any given point in a conversation, including previous user input, the current status of the conversation and relevant external data sources**. By using context, chatbots can respond more intelligently and personally to user input. For example, a chatbot designed to help users book flights can use context to remember the user's favourite airline, departure times and other relevant details from previous interactions, and use this information to make more targeted recommendations or offer more efficient booking options. Context can be modelled in different ways, depending on the specific application and available data sources. Some common approaches include using rule-based systems, machine learning algorithms, or a combination of both, to identify and exploit relevant patterns and features in conversational history and other contextual information. ## Summary intents, entities and context The **intents** of a question refers to the purpose or meaning of the question, and can be identified and used by chatbots and other natural language processing systems to provide better answers. **Entities** are specific objects, concepts or pieces of information that are referenced in a user's input and are relevant to the purpose of the input. Chatbots can use entities to extract specific information or parameters from user input and provide more accurate and personalised responses. **Context** is important for chatbots and refers to the information or cues available to the system at any point in a conversation. By using context, chatbots can provide more intelligent and personalised responses. The process of finding intentions and entities is partly programmed and partly based on machine learning models trained on big data. I will be working on this in the near future and as soon as I have something to show you I will make another article. At least that is the intention 😉 ### Web Strategy Explained: Essential Insights from a Web Strategist URL: https://rob-hoeijmakers.ghost.io/the-web-strategist-and-the-web-strategy/ Last updated: 2026-05-17T13:38:54.000Z I am a Web Strategist and in this article I give insight in what this means and what I deliver. This information will help you setting up your own strategy for the web. A Web Strategist is responsible for **developing and implementing a comprehensive strategy for a company's website**. This includes defining the website's goals and objectives, developing a [user experience ](https://rob-hoeijmakers.ghost.io/user-experience-ux-and-ux-design/)that meets the needs of the target audience, [managing content](https://rob-hoeijmakers.ghost.io/content-management-system-cms/) and updates, and measuring and analysing website performance to continuously improve and optimize the site. The Web Strategist works closely with other teams, such as marketing and design, to ensure that the website supports the company's overall goals and objectives. Some of the key skills and experience required for a Web Strategist include: - **Web design and development**: A Web Strategist should have a solid understanding of web design principles and should be familiar with web development technologies such as HTML, CSS, and JavaScript. - [**User experience**](https://rob-hoeijmakers.ghost.io/user-experience-ux-and-ux-design/): A Web Strategist should have a deep understanding of user experience design and should be able to create intuitive, user-friendly websites that meet the needs of the target audience. - **Analytics**: A Web Strategist should be able to measure and analyze website performance using tools such as Google Analytics, and should be able to use this data to make informed decisions about website improvements and optimizations. - **Marketing and branding**: A Web Strategist should have a strong understanding of marketing and branding principles, and should be able to ensure that the website supports the company's overall marketing and branding strategy. - **Project management**: A Web Strategist should be able to manage website projects from start to finish, including setting timelines, allocating resources, and tracking progress. Some related terms and job titles for a Web Strategist include Web Developer, Web Designer, Web Manager, Web Marketing Strategist, and Digital Marketing Manager. As a Web Strategist, I am responsible for developing and implementing a comprehensive strategy for a company's or organization’s website, with a focus on user experience, content management, and website optimization. My key skills for this role include web design and development, user experience design, analytics, marketing and branding, and project management. ## The web strategy As a Web Strategist, I would deliver a web strategy. A typical web strategy will cover a wide range of topics, including business goals and objectives, target audience, content strategy, SEO, user experience design, technical considerations, performance metrics, and ongoing maintenance and updates. A typical web strategy is a comprehensive plan for achieving a company's website goals and objectives. While the specific components of a web strategy will vary depending on the company's industry, target audience, and business goals, here are the typical elements that may be included: - **Business goals and objectives**: A web strategy should start with a clear understanding of the company's overall business goals and how the website can support those goals. - **Target audience**: The strategy should identify the target audience for the website and create a user experience that meets their needs and expectations. - [**Content strategy**](https://rob-hoeijmakers.ghost.io/the-tower-of-seo-success-pillar-pages/): A content strategy should be developed to guide the creation and distribution of content on the website, including blog posts, product pages, and other types of content. - [**Search engine optimization (SEO)**](https://rob-hoeijmakers.ghost.io/my-basis-for-seo/): The strategy should include an SEO plan to ensure that the website is optimized for search engines and that it ranks well for relevant keywords. - [**User experience design**](https://rob-hoeijmakers.ghost.io/user-experience-ux-and-ux-design/): The strategy should outline the user experience design principles that will be used to create an intuitive and user-friendly website that meets the needs of the target audience. - **Technical considerations**: The strategy should take into account technical considerations such as [website speed](https://rob-hoeijmakers.ghost.io/measure-page-speeds-with-core-web-vitals/), mobile responsiveness, and security. - **Performance metrics**: The strategy should identify the key performance metrics that will be used to measure the success of the website and track progress over time. - **Maintenance and updates**: The strategy should outline a plan for ongoing maintenance and updates to the website, including content updates, bug fixes, and software upgrades. The goal is to create a comprehensive plan that guides the development and management of the website and supports the company's overall business goals. ## *About the Author* *I am an* [*Amsterdam-based web strategist*](https://rob-hoeijmakers.ghost.io/about/) *with a passion for exploring the intersections of technology, marketing, and content realisation. Drop me a note if you miss anything here, you did not find when looking for It helps me a lot, thanks! Rob* ### Review: Steve Jobs by Walther Isaacson. URL: https://rob-hoeijmakers.ghost.io/review-steve-jobs-by-walther-isaacson/ Last updated: 2026-05-14T21:05:01.000Z The book "Steve Jobs" by Walter Isaacson is a fascinating biography that is well worth reading. It offers an in-depth look at the life of one of the most visionary and influential entrepreneurs of our time and provides insights into his unique approach and unparalleled creativity. The book not only covers Jobs' successes at Apple and Pixar, but also offers a glimpse into his personal life and his relationships with colleagues, friends and family members. It describes how he could sometimes be a challenging and demanding person to work with, but also how he was able to inspire and motivate people to be their best. > In most people's vocabularies, design means veneer," Jobs told Fortune shortly after retaking the reins at Apple. "But to me, nothing could be further from the meaning of design. Design is the fundamental soul of a man-made creation that ends up expressing itself in succesive outer layers. > Steve Jobs One of the most valuable lessons to emerge from the book is that success depends not only on skills and knowledge, but also on passion, dedication and an unwavering belief in your vision. Jobs' determination to pursue his dreams and push boundaries led to some of the most ground-breaking technological innovations of our time. All in all, "Steve Jobs" is a compelling biography that not only gives readers a better understanding of Steve Jobs' life, but also a deeper understanding of how innovation and creativity can lead to extraordinary achievements. I would highly recommend the book to anyone interested in technology, entrepreneurship or personal growth. ### Digital scheduling calendars compared and explained URL: https://rob-hoeijmakers.ghost.io/digital-diary-and-calendars/ Last updated: 2026-05-10T08:54:18.000Z The diary and scheduling calendar are tools that mankind has used for thousands of years, and to which divine powers have long been ascribed. Something of that divinity still radiates from them and for many, the diary is something magical and the hustle and bustle of it even a status symbol. It is also pre-eminently the place where the personal and the professional should mix like oil and water. It is also a tool that can bring order to chaos, but is also quickly derailed when participants (or available spaces) are missing due to illness, abandonment or mistakes. In this article, I will take you through the different forms digital diaries and scheduling calendars can take digitally. To plan and manage personal and professional appointments. The calendars themselves, the types of invitations, scheduling appointments and reservations, schedules in public calendars. ## Types of digital scheduling calendars The most basic building blocks, of course, are the digital calendars themselves. That which people use to book and manage appointments, via computer and mobile phone. And actually, you still have to distinguish the programme from the calendar server. You can view a Google Calendar through your Apple Calendar just fine. (Or Exchange, Office 365, iCloud). There are several digital calendar or scheduling apps that people commonly used to manage their calendars. Market shares for calendar apps can be difficult to determine as they depend on several factors such as platform, region and type of user. However, here are some rough estimates made with available data: - **Google Calendar**: As of January 2021, Google Calendar would be the most used scheduling calendar app worldwide, with more than 500 million monthly active users. - **Apple Calendar**: Apple Calendar is the pre-installed calendar app on Apple devices, including iPhones, iPads and Mac computers. By 2021, Apple had an estimated 1.65 billion active devices worldwide, indicating a significant user base for the Apple Calendar app. - **Microsoft Outlook**: Microsoft Outlook is a popular calendar app among business users and is often used in corporate environments. According to Microsoft's Q4 2021 earnings report, Microsoft 365 had 300 million paid copies, including access to Outlook's calendar feature. Of course, the main calendars are also available as a web page or as part of webmail. Let me start with explaining how calendar's and schedules can be filled. ![Google Calendar, Microsoft Outkook and Apple Calendar](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/08/image-3.png) Google Calendar, Microsoft Outkook and Apple Calendar ## Planning appointments and reservations Full diaries are a well-known problem. But how do digital diaries actually get filled? I see these ways: 1. Creating an item yourself 2. An item is shared 3. Items are in a public or shared calendar For item two, it could be joint scheduling (Doodle/Date picker) or subscribing to a calendar (Calendly). In both cases, you will eventually receive an item to add to your calendar. And if you are scheduled somewhere, for family, school or work, these items are usually in a fixed calendar or a shared calendar where you have a bit more freedom. ## Invitations shared So besides creating an appointment yourself, you can also get an invitation shared. For a 1-to-1 appointment or with a group. Invitations that are shared are an item sent to you or posted on a web page. The request or event contains details such as the date, time, place and purpose of the meeting, and the recipient(s) can choose to accept or decline the invitation. ![Typical flow for an invitation from mail to calendar on an iPhone.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/08/Calendar-invite-on-mobile-phone-2.jpg) Typical flow for an invitation from mail to calendar on an iPhone. You get these calendar items as soon as you participate in an event such as a concert, train or plane trip, course or meeting. For example, if you participate via Facebook, LinkedIn or Hopin. Applications such as Date Picker and Calendly will also share an item with you so that it will end up in your digital calendar. And increasingly, online events are sharing a calendar from which you can choose sessions. 💡 ****Online and offline invitations** There are two variants of such an invitation. The first is an ICS file that you open and save locally, and the second is an invitation to Google Calendar. They are hard to tell apart, I wrote an articles on explaining [the difference between ICS and Google invitations](https://www.hoeijmakers.net/understanding-digital-calendar-invitations/). ### Applications such as Doodle Scheduling an appointment will involve looking for a date after some consultation, for instance via e-mail or WhatsApp. For this, there are applications like Doodle or date picker that work with a poll or poll. Situations for which this is useful are: - Planning meetings - Planning events (parties, workshops or conferences) - Planning interviews (recruiters) - Arrange travel - Arrange consultations (lawyers, doctors or consultants) In general, tools like [Doodle](https://www.doodle.com) are useful in any scenario where planning and coordination are needed. They can save time and reduce frustration by simplifying finding a mutually agreeable time for meetings, events or other types of appointments. If the appointment is confirmed by everyone in the group (or the majority) then it can be saved in the digital calendar. [Apple’s New ‘Invites’ App: A First LookApple Invites is here! A new iOS event app that jazzes up invites but lacks true collaboration. Can it compete with for instance WhatsApp’s voting? Let’s find out.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-17.png)Rob Hoeijmakers (robhoeij)Rob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_0846.jpeg)](https://rob-hoeijmakers.ghost.io/apples-new-invites-app-a-first-look/) ### Scheduling software It is a type of software application designed to simplify the process of scheduling meetings or appointments by allowing users to easily view available time slots and book appointments or meetings with others. Examples are [Calendly](https://calendly.com), Microsoft Bookings, [Google Calendar](https://rob-hoeijmakers.ghost.io/from-calendly-to-scheduling-with-google-calendar/), [Koalender](https://koalendar.com/) and [Acuity](https://www.acuityscheduling.com) by [Squarespace](https://rob-hoeijmakers.ghost.io/squarespace-the-experiences-after-a-year/). Planning or appointment scheduling software has become increasingly popular in both personal and business life, with a significant increase in application in recent years. In personal life, scheduling software is often used to schedule appointments with doctors, dentists, consultants, coaches and other service providers. Good to know is that, on the back end, these tools are often linked to the digital calendar of the professional who offers. If this person is not available according to the calendar, the available time will also disappear. ![Calendly flow on iPhone. From blog to meeting.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/08/Calendly-flow-Rob.jpg) Flow in Calendly, from blog to 30 minute meeting choice By the way, by no means everyone is enamoured with these kinds of tools. A good proportion of people find it unseemly and rude to be saddled with finding a free slot in someone else's diary. Personally I think Calendly makes a lot of sense but you need to use it with care and don't push them up to people. I use it to [make appointments available to meet new people](https://rob-hoeijmakers.ghost.io/about/). ### Online events The way we interact with events has undergone a major shift. Online events are the standard, offering convenience, global reach and flexibility. These virtual meetings include everything from webinars and workshops to conferences and product launches. As more and more organisations embrace the virtual world, the need for efficient event management has never been greater. Event management software, such as Cvent, Goldcast and Eventbrite, forms the basis of these virtual experiences. One feature of this tool is the integration of event schedules with digital calendars. Participants receive formatted calendar invitations, complete with fixed event times and session details. This integration unites structured schedules and participant flexibility. Participants can explore the range of sessions on offer and choose those that match their interests and availability. Individual items then go into the digital calendar. ![Example of items from an online event included in the digital calendar](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/08/2023-08-25_08-45-29.png) Example of items from an online event included in the digital calendar. ## Shared and public agendas On the above forms, you still have some control. This is often not the case for shared calendars, schedules and public agendas. You can access these, but usually not edit them. To have these in your digital calendar, you usually have to take out some kind of subscription. ### Shared calendars Shared calendars are powerful tools that enable individuals and teams to collaborate seamlessly and plan effectively. These calendars allow multiple people to access the same calendar and jointly manage appointments, events and deadlines. This type of calendar is particularly useful when people collaborate on projects, organise joint events or simply want to coordinate their personal and business schedules. These shared calendars offer different levels of access rights. Some participants may only be able to view appointments, while others have the ability to add new items and edit existing ones. This flexible system makes it possible to share information without sacrificing security or control. Shared calendars can have a variety of uses. They are often used in business environments to manage team activities and deadlines. They are also valuable on a personal level, as in the case of partners who need to coordinate both personal and business commitments. The ability to see real-time changes and respond immediately makes shared calendars an essential tool for modern collaboration and planning. ### Schedules Scheduling software, also known as employee scheduling software, helps organisations manage employee schedules, shifts and time off. Especially companies with hourly or shift workers, such as shops, restaurants and healthcare facilities, use these tools. The software improves efficiency, reduces errors and saves time in employee scheduling. It automates manual processes, allowing companies to focus on other core aspects of their operations. Schools also use scheduling software to keep track of class schedules, teacher allocations to classes and availability. It manages substitute assignments and tracks teachers' professional development hours. Besides timetables, these programmes can also track student attendance and manage extracurricular activities. Some examples include [Blackboard](https://www.anthology.com/products/teaching-and-learning/learning-effectiveness/blackboard-learn), Moodle and PowerSchool. ### Family planners Digital family planners have revolutionised the way families manage their busy lives. These modern tools go beyond traditional calendars and shared diaries by being specifically designed to streamline the complex planning of family activities. Digital family planners allow all family members to keep track of their appointments, events and tasks in one place. What sets these planners apart are additional features tailored to the needs of families. They often offer colour-coding to identify each family member, notifications for approaching activities and the ability to share shopping lists and to-do's. Some platforms even offer integration with other apps and services, such as grocery delivery or location tracking, to further simplify family life. With the proliferation of work, school and leisure activities, coordinating family calendars has never been more crucial. Digital family planners meet this need by enabling families to work together in a streamlined and organised way. Whether planning parents' evenings, sports events or family outings, these tools contribute to a smooth and well-organised family life. Examples of family planners are: Cozi, FamilyWall, TimeTree. ![Examples of family planners are: Cozi, FamilyWall, TimeTree.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/08/Family-planners.jpg) Examples of family planners are: Cozi, FamilyWall, TimeTree. ### Public calendars Public calendars are calendars that anyone can view or open. These calendars can be used to share information about events, appointments, meetings and other types of planned activities. Here are some common use cases for public calendars: - Business calendars - Public events - Municipal calendars - Personal calendars In general, public calendars can be a useful tool for sharing planning information with others, promoting events and improving communication and collaboration within a group or community. I use them to calendar my contacts' birthdays, keep track of public holidays and my plane and hotel bookings. But calendars for football matches and TV broadcasts are also popular. Schedjoules' app includes dozens of calendars you can subscribe to: ![App van Schedjoules met tientallen kalenders](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/02/2023-02-23_17-54-17.jpg) Schedjoules app with dozens of calendars ## Calendars wrapped up Digital invitations, diaries and calendars come in multiple shapes and sizes. They provide much-needed structure and the opportunity to bring groups of people together efficiently and effectively. But they are also a genre in their own right, and not yet easy to deploy well. They are subject to social dynamics, rankings and the whims of humans. As a result, they can sometimes cause more work than they take away. Also, the many options sometimes make it so complicated that people fall back on pen and paper. And another thing, digital diaries fill up faster it seems, especially if others have direct access to them. ### Further reading - [When AI Makes the Calendar Click](https://rob-hoeijmakers.ghost.io/gemini-google-calendar/) - [Regular and online calendar invitations](https://rob-hoeijmakers.ghost.io/understanding-digital-calendar-invitations/) - [From Calendly to scheduling with Google Calendar](https://rob-hoeijmakers.ghost.io/from-calendly-to-scheduling-with-google-calendar/) ### Three ways to use WhatsApp via computer URL: https://rob-hoeijmakers.ghost.io/three-ways-to-whatsapp-via-computer/ Last updated: 2026-04-25T13:16:02.000Z #### It is important to always have the ability to communicate with your customers via WhatsApp, regardless of the device they use. Meanwhile, more people surf via mobile than via computer. Current [figures for Europe](https://gs.statcounter.com/platform-market-share/desktop-mobile-tablet/europe) indicate that 50% of internet traffic is via mobile, 47% via desktop and 3% via tablet. So there is certainly still a significant proportion of traffic that goes via computer. Therefore, it is useful to know that you can connect WhatsApp via computer in three ways! ## 1\. WhatsApp Desktop WhatsApp desktop is [a programme](https://www.whatsapp.com/download) that lets you use WhatsApp on your computer. You can send messages, make calls and exchange files. It runs on Windows, Linux and Mac computers. ![Screenshot of WhatsApp Desktop](https://www.chatvoorbedrijven.nl/content/images/2023/02/screenshot.webp) Screenshot of WhatsApp Desktop ## **2\. WhatsApp Web** WhatsApp Web app is [an app](https://web.whatsapp.com/) that lets you use WhatsApp on your computer's internet browser. If you don't want to install anything on your computer, this is a good alternative. ![WhatsApp Web](https://www.chatvoorbedrijven.nl/content/images/2023/02/WhatsApp-Web.jpg) WhatsApp Web ## **3\.** WhatsApp mobile via QR code And if someone has the app WhatsApp Mobile installed on their phone, they can scan a QR code with their camera and instantly connect with you via messages or calls - no need to download anything! ![WhatsApp via a QR code on the big screen](https://www.chatvoorbedrijven.nl/content/images/2023/02/WhatsApp-QR-tag.jpg) WhatsApp via a QR code on the big screen ## Conclusion If you want to use WhatsApp as your main chat channel on your website, it is a good idea to take the computer into account. Not everyone will be able to switch devices quickly, and I also definitely recommend being able to fall back on a classic message or e-mail, for example. 💬 If you are reading this article on a computer, you can test what happens when you click a WhatsApp link. This is an example: [WhatsApp link](https://api.whatsapp.com/send?phone=310202101648&text=Hi%2C%20I%20got%20your%20WhatsApp%20information%20from%20your%20website.) But that aside, if I assume the huge spread of WhatsApp, mobile browsing as the biggest channel and QR code familiarity, you can bet on WhatsApp as the primary channel in 2023. *Article was first published on* [*"Chat voor bedrijven", a Dutch blog on why chat and messaging are important for companies*](https://www.chatvoorbedrijven.nl/drie-manieren-om-te-whatsappen-via-de-computer/)*. On there, right bottom corner of the screen, you can also check out the switch from web to WhatsApp.* ### How internet memes spread and stick URL: https://rob-hoeijmakers.ghost.io/internet-memes-spread-and-stick/ Last updated: 2026-04-25T13:16:02.000Z Memes to me, as a medium, resemble cartoons and also propaganda posters. They are very visual, have a compact message and are often easily altered, shared and remembered. They are complex in the sense that you often need to be an insider to the references that are made. In this article, I stay on the surface and describe what a meme is and speculate why it is so contagious. ## Meme as a word The word "meme" was coined by Richard Dawkins, an evolutionary biologist, in his 1976 book "The Selfish Gene." In the book, Dawkins proposed the concept of the meme as a unit of cultural transmission, analogous to the way that genes transmit biological traits from one generation to the next. According to Dawkins, **a meme is an idea, behaviour, or style that spreads from person to person within a culture**. Memes can include anything from a catchy phrase to a fashion trend to a piece of technology. The word "meme" is derived from the Greek word "mimema," which means "**something imitated**." Since its inception, the concept of the meme has evolved and been adapted to describe a wide range of phenomena, including the spread of **Internet memes** and other forms of viral content. Today, the word "meme" is used colloquially to describe any idea, behaviour, or style that becomes widely popular or influential through repetition or imitation. ![Richard Dawkins, and his own meme meme.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/01/Dawkinsmeme-09a60a478f2849178939f9bfe701a7dd.jpg) Richard Dawkins, and his own meme meme. ## Why are internet memes so contagious? Memes are so contagious because they tap into **basic human desires and emotions**. Many memes are humorous and are designed to be easily understood and shared with others. They often use cultural references, jokes, and other forms of humour to make them more relatable and appealing to a wide audience. In addition, memes are often designed to be **easy to share**, either through social media platforms or through more traditional means such as text or email. This makes it easy for people to pass them along to their friends and family, leading to their rapid spread. Finally, memes often have **an element of surprise or novelty**, which can make them more memorable and more likely to be shared. All of these factors contribute to the contagious nature of memes and help to explain why they are so popular and widely shared. ![Meme, with a visual surprise.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/07/DraggedImage-2-2.jpg) Meme, with a visual surprise. ## Picture Superiority Effect Compared to other cultural expressions like tweets, a meme has an advantage because it is a picture and that can have a **superior effect on memory**. The Picture Superiority Effect is a phenomenon in psychology that refers to the idea that people are **more likely to remember and recall information when it is presented to them in the form of an image or picture**, rather than in the form of text. This is thought to be because pictures and images are more easily processed and encoded by the brain than text, and because they can provide contextual and visual cues that help to anchor the information in memory. Research has shown that the Picture Superiority Effect can lead to better learning outcomes, particularly when it comes to complex or abstract concepts. Internet memes are often designed to be visually appealing and to convey a message or concept using a combination of text and images. As such, it is possible that the Picture Superiority Effect could play a role in the effectiveness of internet memes in communicating ideas and information. ### Blue LED changed the world URL: https://rob-hoeijmakers.ghost.io/invention-of-blue-led/ Last updated: 2026-04-25T13:16:01.000Z There was a decade long search for a reasonably bright and affordable blue LED (Light Emitting Diode). Not to pimp game computers, but it was the missing colour in [Red, Green and Blue](https://rob-hoeijmakers.ghost.io/three-colors-of-light-and-a-drop-of-water/). Or RGB. With the combination of these three colours, you can produce all the colours imaginable. Red and green LEDs were already available for a while, but blue was harder to find. This is a story about a determined team in Japan that kept looking for something many had stopped looking for: a blue LED light. ## Short history of LED The first LED (Light Emitting Diode) was invented in the 1920s, but it wasn't until the 1960s that the first LED displays were created. These early displays were very basic and could only display simple text and numbers. ![Red LED](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/01/image.jpeg) Red LED It wasn't until the development of the blue LED in the 1990s that LED displays became more widespread and advanced. 💡 Why was blue LED important for producing white lights? In a white LED light bulb, a blue LED is covered by a phosphor material that converts some of the blue light into yellow light. When the blue light and the yellow light mix, they produce white light. This method of creating white light is more energy-efficient and longer-lasting than traditional incandescent light bulbs . ## The search for blue LED Shuji Nakamura was driven by a strong scientific curiosity and a desire to push the boundaries of what was thought possible. He was also motivated by the challenge of finding a solution to a problem that many other researchers had been unable to solve. Despite facing numerous obstacles and setbacks, Nakamura was persistent in his efforts and continued to experiment and refine his approach until he finally succeeded in producing blue LEDs. ![Shuji Nakamura](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/01/Shuji_Nakamura.jpg) Shuji Nakamura This work **revolutionized the field of lighting technology**, as blue LEDs made it possible to create white light sources that were much more energy efficient than traditional incandescent bulbs. In 2014, Nakamura, with fellow researchers Akasaki and Amano, was awarded the Nobel Prize in Physics for the invention of the blue LED. This was the first time that the Nobel Prize in Physics had been awarded for a discovery in the field of solid-state lighting. Related article: [The influence of blue light: how to protect your eyes and sleep?](https://rob-hoeijmakers.ghost.io/the-influence-of-blue-light-how-to-protect-your-eyes-and-sleep/) 🔵 The technical background. Shuji Nakamura discovered blue LEDs by using a process called molecular beam epitaxy to grow high-quality gallium nitride (GaN) crystals, which are capable of emitting blue light. He discovered that by adding small amounts of indium to the GaN crystals, he could produce blue light, which was previously believed to be impossible. How blue LEDs changed the world. ### WhatsApp Messenger and Business on a single iPhone URL: https://rob-hoeijmakers.ghost.io/whatsapp-messenger-and-business-app-on-the-same-iphone-with-a-virtual-number/ Last updated: 2026-04-25T08:32:38.000Z I heard that you can use both WhatsApp Messenger and WhatsApp Business on the same iPhone, and I saw it as a great way to separate personal and business messaging. But I came back from that idea after a pretty cold shower. It is not a good solution. I purchased a new number specifically for my business traffic. A number that also indicates that I am in Amsterdam, but transfers this landline to my mobile phone. A so-called **virtual number**. Another such virtual number can be linked to WhatsApp, you will get a call in which a six-digit code is transmitted. You can also use two mobile lines: [From physical SIM to eSIM on iPhoneGoing from a physical SIM to an embedded SIM (eSIM) on an iPhone. The steps to transfer and the first impressions of having two SIM’s active.![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-9.png)Rob Hoeijmakers (robhoeij)Rob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_9936-1.jpeg)](https://rob-hoeijmakers.ghost.io/from-physical-sim-to-esim-on-iphone/) Use WhatsApp on a single phone with two lines. ## WhatsApp Messenger and Business on the same iPhone So request number, install WhatsApp Business app, configure it and it worked. I was completely happy! But the misery started when I also wanted to **create separate address books**. WhatsApp's contacts are on the server, and so I thought I could delete my private addresses from WhatsApp Business app. But that deletion is very difficult, namely one by one and with three screens, but later I also understood why and that I should actually be happy. This is because **WhatsApp Messenger (personal) and WhatsApp Business (business) share the iPhone address book**. If you discard something from WhatsApp, you also discard it from the iPhone and from the other WhatsApp app. Sigh. ![My WhatsApp Business with virtual number.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/01/Whatsapp-Business-Virtuale-Number.jpg) My WhatsApp Business with virtual number. ## Segmenting contacts Segmenting your customers in this way does become very difficult, if not impossible. It very well shows the limitation of WhatsApp Business. It actually needs a completely standalone phone with its own address book. And therefore also with its own Apple ID. That doesn't work for me on a day-to-day basis, and I am now considering the possibilities of **WhatsApp Business API**. But the requirement is actually also there that I can manage the chats via the mobile phone, otherwise reachability becomes too difficult to fill in. ## For whom is having WhatsApp and Whatsapp Business on the same phone suitable? Actually, mainly for people where there is **a large overlap between personal and business contacts.** And where the customer base is not too large and where communication takes place entirely through one person. Multiple connections to a single WhatsApp account can be made, but no information about any conversations can be shared. ### Composable DXP explained URL: https://rob-hoeijmakers.ghost.io/what-is-a-composable-dxp/ Last updated: 2026-05-10T08:54:19.000Z Let me first define a [digital experience platform (DXP)](https://rob-hoeijmakers.ghost.io/dxp/): DXP is a platform that enables organizations to create, manage, and deliver personalized digital experiences by composing and assembling different components, such as text, images, videos, and interactive elements. A **composable DXP** allows teams to break down the digital experience into smaller, reusable components, which can be combined and assembled in different ways to create a wide range of experiences. This allows teams to be more efficient, as they can reuse components and avoid duplicating effort. It also enables teams to create and test new experiences more quickly and easily, by assembling and composing different components in real-time. In addition to the benefits of component-based design, a composable DXP also typically includes features for data-driven design and personalization, allowing teams to create tailored experiences that are optimized for different user groups and use cases. By analysing user data and behaviour, teams can create experiences that are tailored to the specific needs and preferences of different user groups, increasing engagement, conversion rates, and overall satisfaction with the digital experience. Overall, a **composable DXP is a platform that provides teams with the tools and components they need to create engaging and personalized digital experiences**, by composing and assembling different components in real-time. ## What makes a DXP composable? A Digital Experience Platform (DXP) being "composable" refers to its architecture and design principles that allow for flexibility, modularity, and interoperability. Here's what makes a DXP composable: 1. **Modularity**: A composable DXP is built with independent modules or components that can be assembled, rearranged, or replaced without affecting the overall system. This allows for customization and scalability. 2. **Interoperability**: Composable DXPs are designed to work seamlessly with various other tools, systems, and platforms. This ensures that different components can communicate and function together, enabling a more cohesive digital experience. 3. **API-First Approach**: By adopting an API-first approach, composable DXPs allow for easy integration with other systems. This means that different parts of the platform can communicate through well-defined APIs, making it easier to plug in new tools or swap out existing ones. 4. **Flexibility**: The composable nature of a DXP allows businesses to adapt to changing needs and technologies. You can add, remove, or modify components as needed without a complete overhaul of the system. 5. **Vendor Neutrality**: Composable DXPs often support a wide range of third-party tools and services, allowing businesses to choose the best solutions for their specific needs rather than being locked into a single vendor's ecosystem. 6. **Focus on User Experience**: By enabling a more tailored and flexible approach, composable DXPs can provide a more personalized and efficient user experience. This aligns with the need to create engaging and relevant digital experiences for users. 7. **Cost-Effective**: The ability to swap out or add components as needed can lead to cost savings, as businesses can invest in specific functionalities that align with their current needs rather than a monolithic system that may include unnecessary features. In essence, a composable DXP is all about providing the flexibility to create a platform that's tailored to a business's unique needs and can evolve with changing trends and technologies. It's a concept that likely aligns well with your work in [web strategy and technology](https://rob-hoeijmakers.ghost.io/the-web-strategist-and-the-web-strategy/). ## (Composable) DXP Vendors There are many vendors that offer digital experience platform (DXP) solutions. Some examples include: - Adobe Experience Platform - Acquia - Bloomreach - Episerver - Kentico - Sitecore - OpenText - Oracle - Salesforce - SAP These vendors offer a range of DXP solutions with different features, capabilities, and pricing models. Some **DXPs are designed for specific industries**, such as retail, financial services, or healthcare, while others are more general purpose and can be used in a variety of contexts. It's worth noting that DXP solutions can be complex and may require significant customisation and integration with other systems and technologies. Therefore, it's important for organizations to carefully consider their needs and goals when selecting a DXP vendor, and to carefully evaluate the capabilities and fit of the different options available. --- ### Further reading - [Intro to Digital Experience Platforms (DXP)](https://rob-hoeijmakers.ghost.io/dxp/) - [What is digital experience composition and why should marketers care?](https://uniform.dev/blogs/digital-experience-composition-dxc/explaining-dxc-and-its-benefits-to-marketers) ### Signal rivals WhatsApp URL: https://rob-hoeijmakers.ghost.io/signal-messenger-emphasises-privacy/ Last updated: 2026-05-10T08:54:19.000Z While WhatsApp does use end-to-end encryption for all its chats, there are a few reasons why [Signal](https://signal.org) may be considered more private than [WhatsApp](https://rob-hoeijmakers.ghost.io/is-whatsapp-social-media/). 1. Signal uses a **more advanced encryption protocol** called the Signal Protocol, which has been independently audited and is considered to be more secure than the encryption protocol used by WhatsApp. 2. Signal is an **open-source project**, which means that its code is publicly available for anyone to review and audit. This allows for a higher level of transparency and security, as opposed to WhatsApp's proprietary encryption protocol. 3. Signal does **not collect or store any metadata about your messages**, whereas WhatsApp collects some metadata, such as the phone numbers of the sender and recipient(s) and the time and date of the messages. ## Signal’s ownership Let’s follow the money and see who owns Signal. Signal is a non-profit organization, it is owned by the Signal Foundation which was founded by Moxie Marlinspike and Brian Acton in 2018. Brian Acton is a co-founder of WhatsApp, he left the company and later donated $50 million to Signal Foundation to help fund the development of the Signal app. The organization is focused on developing open-source, end-to-end encrypted messaging and voice call technology. Because Signal is a non-profit organization and is not owned by a larger corporation like WhatsApp which is owned by Facebook, this means that Signal is **less likely to be influenced by outside parties**. ## Signal’s growing user base It's difficult to compare the usage of Signal and WhatsApp as they are both widely used messaging apps, but Signal has a smaller user base compared to WhatsApp. According to data from SimilarWeb, WhatsApp is the most popular messaging app in the world, with over 2 billion monthly active users as of 2021\. In comparison, Signal has seen a significant increase in downloads and usage following some privacy controversies surrounding WhatsApp and other messaging apps, but still has a smaller user base. 💬 Big plus, there is also a ****desktop app** for Mac, Windows and Linux. Works really well. One drawback is that existing chats are, for privacy reasons, not copied over. It's important to note that Signal's user base is still growing, and its usage is increasing rapidly. And I do need to say that Signal has improved a lot in recent years. In the onboarding and also in the user experience. Both Signal and WhatsApp offer high-quality voice and video calling features and so feature wise they go neck to neck. Signal definitely is a big rival of WhatsApp if it finds ways or momentum to further increase the user base. --- ### Further reading - [Signal vs. WhatsApp](https://rob-hoeijmakers.ghost.io/signal-vs-whatsapp/) - [How Signal Walks the Line Between Anarchism and Pragmatism](https://www.wired.com/story/signal-politics-software-criticism/) ### Sleeping Kobo is not waking up URL: https://rob-hoeijmakers.ghost.io/sleeping-kobo-is-not-waking-up/ Last updated: 2026-04-14T13:21:09.000Z The book you stopped reading yesterday is still on the display, but when you press the on button, only the backlight of the reader goes on. What to do now to make the Kobo wake up? ## Put the Kobo on the charger Just now, I wanted or continue reading and had that exact issue. The Kobo was sleeping and when I wanted to wake it up only the backlight started giving light, but no book appeared. Pressing the on button did not help. The solution for a sleeping Kobo not waking up is: **Connect the e-reader to the charger and then press the on button for 15 seconds.** The Kobo e-reader woke up immediately, but it had to reboot the software. After that, the Kobo reader worked. Of course, I needed to keep the Kobo charging for a while. 💬 An e-reader does not need energy to display something on the screen. Displaying content is an analogue, physical process and once done no further energy is needed. This makes it a bit hard to detect that the battery actually is empty. I wrote an [article on how e-ink works](https://rob-hoeijmakers.ghost.io/electronic-paper-and-digital-ink/). ## Possible causes of crashes Two things to keep in mind is the battery and the system software. ### Battery I have noticed that when the Kobo Clara e-reader is frozen, there is often something wrong with the battery. Poorly charged, for instance with a backup battery. Also, a different charge load is indicated than the one indicated later, after startup. > Try to keep the Kobo charged. ### System software Another possible cause is outdated system software. It is good to keep this up to date. > Keep system software up to date ### Slay the multichannel dragon with a headless CMS URL: https://rob-hoeijmakers.ghost.io/headless-cms/ Last updated: 2026-04-25T13:17:09.000Z A headless CMS ([content management system](https://rob-hoeijmakers.ghost.io/content-management-system-cms/)) has no built-in front-end functionality and instead, it provides an API (application programming interface) for developers to access the content and display it on the front-end using their own code. This means that the front-end can be built using any language, framework, or platform that can connect to the API, making it more flexible and decoupled from the CMS. ## Headless The term "headless" comes from the fact that the CMS has no "head" (i.e. no built-in front-end) and instead, it focuses on providing an API for managing the content. Headless CMSs are becoming increasingly popular because they allow for more flexibility and scalability, as the front-end can be built using any technology, separate from the CMS. This also enables the use of multiple front-end channels like web, mobile, smart devices, and even voice assistants. It also allows for better performance and personalization, as the front-end can be optimized for specific devices or contexts without being constrained by the CMS. This is different from traditional or monolithic CMSs, which usually have a built-in front-end that is tightly integrated with the content management functionality. This can make it more difficult to customize the front-end or to use multiple front-end channels. > In a headless CMS, the backend and the frontend are connected, but they’re connected by an application programming interface (API) instead. Please keep in mind that Headless CMSs are not the best solution for all use cases, it is especially useful for those who want more flexibility and scalability in their web or mobile application and also for those who want to create personalized experiences. ## Headless CMS vendors and products Here is an overview of vendors of headless CMS products. - [**Agility CMS**](https://agilitycms.com): a flexible and powerful headless CMS that offers a wide range of features for managing and delivering content, including support for multiple languages and teams. - [**Caisy**](https://caisy.io): Next generation headless CMS that enables you to manage and create content for all devices. - [**Contentful**](https://www.contentful.com): a headless CMS that offers a wide range of tools for managing and delivering content across multiple platforms. - [**Kentico Konten**](https://www.kentico.com)**t**: Kentico Kontent is a headless CMS that provides a complete set of tools for managing and delivering content across multiple channels and platforms. - [**Kontent.ai**](https://kontent.ai): a content management system (CMS) that uses artificial intelligence (AI) and machine learning to help users create, manage, and optimize digital content. - [**NetlifyCMS**](https://www.netlify.com): Open source, git-based, headless CMS that works with static site generators. - [**Strapi**](https://strapi.io): A headless CMS that allows developers to create websites, mobile applications, eCommerce sites, and APIs without knowing any backend or database programming. It is an open-source, Node.js based, and free-to-use service for developers. Strapi is a content management system (CMS) that provides backend-only functionalities, making content accessible through a GraphQL or REST API and displayable on any device possible. - [**Wagtail**](https://wagtail.org): Wagtail CMS is an open-source content management system (CMS) built for the Django web framework and powered by Python. - [**Sanity.io**](https://www.sanity.io): A highly customizable, developer-focused headless CMS known for its real-time collaboration capabilities and flexible content modeling. Sanity.io allows teams to create bespoke content structures tailored to their unique needs and integrates seamlessly with popular frameworks like React and Next.js. Its powerful content studio provides an intuitive, user-friendly interface, while the globally distributed CDN ensures fast content delivery across all devices. ## Example: Ghost, a headless CMS This blog is build on [Ghost](https://rob-hoeijmakers.ghost.io/ghost-blog/). Ghost is an open-source, node.js based headless CMS that is designed to be easy to use, customizable and efficient. Ghost allows users to create, manage and organize content using its built-in editor, and it exposes the content through a RESTful API, which can be consumed by any front-end technology or framework. Ghost also uses markdown for formatting and it has built-in SEO features. Headless CMSs like Ghost allow developers to decouple the content management system from the front-end presentation layer, which can be beneficial in many situations. It can increase development speed, improve scalability, and make it easier to create personalized, dynamic and responsive user experiences. In my case I can decide to change the full layout and design directly. ![Ghost themes](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/01/Ghost-themes.jpg) Ghost themes ### Find the right CRM system URL: https://rob-hoeijmakers.ghost.io/customer-relationship-management-crm/ Last updated: 2026-04-24T18:36:15.000Z Let me explain where a CRM is used for in a business. This quick walk through will already help to see if your business actually could benefit from a CRM. I will then give options to make choosing more efficient, and I will conclude with an overview of providers. ## Where is a CRM used for? A CRM, or Customer Relationship Management, is a software tool that helps businesses manage and analyse customer interactions and data throughout the customer lifecycle. The main functions of a CRM system include: - **Storing and organizing customer data**: A CRM system allows businesses to store and organize customer data in one centralized location, making it easy to access and manage. - **Tracking customer interactions**: A CRM system allows businesses to track customer interactions, such as phone calls, emails, and meetings, which helps them to understand customer needs and preferences. - **Managing sales and marketing**: A CRM system can be used to manage sales and marketing activities, such as lead tracking, campaigns, and forecasting. - **Improving customer service**: A CRM system can help businesses to improve customer service by providing tools for managing customer inquiries and support requests. - **Analysing customer data**: A CRM system can be used to analyse customer data, such as purchase history and behaviour, to identify trends and patterns that can be used to improve business strategies. In summary, CRM systems provide businesses with valuable insights into customer behaviour and help them to manage and improve their interactions with customers, ultimately improving sales and customer retention. ## Choose the right CRM for your business Choosing the right CRM for your business can be a complex process, as there are many different options available, each with its own set of features and capabilities. Here are a few key factors to consider when choosing a CRM for your business: - **Identify your business needs**: Before you start looking at CRM options, it's important to have a clear understanding of what you need from a CRM system. Identify which specific business processes you need to manage, such as sales, marketing, or customer service. - **Evaluate scalability**: Consider whether the CRM system can scale to meet the needs of your business as it grows. Think about how many users will be accessing the system, and whether it can handle an increase in data and functionality over time. - **Look for ease of use**: A CRM system should be easy for your employees to use and navigate. Look for a system that is user-friendly and intuitive, with a simple and straightforward interface. - **Consider integration capabilities**: It's important to ensure that the CRM system can integrate with other tools and systems that you are currently using, such as your email marketing software or accounting system. - **Customer support and training**: Consider the level of customer support and training provided by the CRM vendor, as well as the resources available such as documentation, tutorials, and user forums. - **Cost**: CRM systems can vary widely in cost, so it's important to consider the features you need and the budget you have available. Some CRM are offered as a service (SaaS) and some are licensed, take into account the costs of implementation, maintenance, and upgrades. Ultimately, the best CRM for your business is the one that meets your specific needs and provides the features and capabilities that you require. It might be helpful to try a free trial or demo of several options before making a final decision. ## CRM software vendors Here is a list of some popular modern vendors of CRM software: - [**Salesforce**](https://www.salesforce.com/crm/): One of the most well-known and widely used CRM systems, Salesforce offers a wide range of features and capabilities for managing sales, marketing, and customer service. - [**Microsoft Dynamics 365**](https://dynamics.microsoft.com/): Microsoft Dynamics 365 is a cloud-based CRM system that offers a wide range of features for managing customer data and interactions, as well as powerful analytics and reporting capabilities. - [**Zoho CRM**](https://www.zoho.com/): Zoho CRM is a cloud-based CRM system that offers a wide range of features for managing sales, marketing, and customer service, as well as integration with other Zoho tools and apps. - [**Hubspot CRM**](https://www.hubspot.com/products/crm): Hubspot CRM is an easy-to-use CRM system that offers a wide range of features for managing customer data and interactions, as well as powerful marketing and sales tools. - [**Insightly**](https://www.insightly.com)**:** a CRM platform designed to help businesses convert quality leads into opportunities and manage the entire sales process from opportunity to close. The platform aims to facilitate faster business growth by enabling team collaboration on mission-critical details. - [**Pipedrive**](https://www.pipedrive.com): Pipedrive is a CRM system designed specifically for sales teams, offering a range of features for managing customer data and interactions, as well as powerful analytics and reporting capabilities. - [**SugarCRM**](https://www.sugarcrm.com/uk/?utm%5Fsource=google.com&utm%5Fmedium=organic): SugarCRM is a cloud-based CRM system that offers a wide range of features for managing sales, marketing, and customer service, as well as powerful analytics and reporting capabilities. - [**Freshsales**](https://www.freshworks.com/nl/crm/sales/): Freshsales is a CRM system that offers a wide range of features for managing customer data and interactions, as well as powerful analytics and reporting capabilities. ### CRM Market share As I dive into Customer Relationship Management (CRM), one crucial aspect that stands out is the market share of these platforms. Choosing the right CRM for your business is no easy task, and understanding how these solutions fare in the market can provide valuable insights. Here is a graphical overview of the market share of the top CRM platforms, enabling you to make informed decisions for your organization's CRM needs. From the established industry giants such as Salesforce and Microsoft Dynamics to the rising stars like Zoho CRM and Hubspot CRM. ## Built-in CRM in Squarepace, Wix or WordPress Several CMS platforms offer built-in CRM functionality or integrations with external CRM platforms to help users manage customer relationships and interactions. We won't find traditional CRM functionality in **Squarespace**, but the platform provides tools that allow small e-commerce businesses to manage customers, including collecting and managing contact details, all in one place. **Wix** has a similar approach with Wix Engage, an optional service that collects contact data, manages customer interactions and sets up email campaigns automatically. This is available at an extra cost on top of the regular Wix subscription. Several CRM plugins are available in **WordPress**, including HubSpot, [WP-CRM System](https://www.wp-crm.com/), [Jetpack CRM](https://jetpackcrm.com) and [UpiCRM](https://www.upicrm.com/). These plugins allow users to collect contact information, manage leads and track customer interactions within the WordPress dashboard. It is important to consider which features and pricing models best suit your needs when choosing a CMS with built-in CRM functionality or integrating an external CRM solution. ### The MACH Alliance URL: https://rob-hoeijmakers.ghost.io/mach-alliance/ Last updated: 2026-04-24T18:34:55.000Z A bit of a blunder to ask the name of the host, and I decided to make up for it right away and write an article to introduce the MACH alliance a bit more. The organization is relatively new, founded in 2020\. I explain where they stand for, how they are organised and why they promote, amongst others, the [Composable DXP](https://rob-hoeijmakers.ghost.io/what-is-a-composable-dxp/). ## Goals of the MACH Alliance MACH is short for: - Microservices - API-first - Cloud-native - Headless The MACH Alliance is an industry alliance that **promotes the use of modern software architecture practices for building and delivering digital experiences**. The alliance aims to promote the use of microservices, API-first, cloud-native and headless technologies as a way to build more flexible, scalable, and adaptable software. The alliance is made up of a group of technology vendors, software companies, and other organizations that are committed to promoting the use of these modern software architecture practices. They provide **resources, best practices, and guidance** to help organizations adopt these technologies and practices in order to deliver digital experiences that are faster, more flexible, and more responsive to changing customer needs. ## MACH Alliance organization MACH Alliance is a **non-profit organization** and is not affiliated with any specific technology vendor or product. The goal of the alliance is to help organizations adopt modern software architecture practices, regardless of the specific products or technologies they choose to use. The MACH Alliance is a relatively new organization, it was founded in 2020, but it's gaining momentum and more organizations are joining it. As the digital landscape is evolving and the needs of customers are changing, the adoption of modern software architecture practices is becoming increasingly important for organizations that want to stay competitive in the digital world. ## Mach Alliance and composable software Composable software is a type of software architecture that is based on the principles of MACH, it **enables businesses to build and deliver flexible, modular, and scalable software solutions**. The idea behind composable software is to build software using small, independent, and loosely coupled services (microservices) that can be easily composed and recombined to create new functionality. MACH Alliance and composable software are related because they both focus on building software using small, independent, and loosely coupled services (microservices) that can be easily composed and recombined to create new functionality. They both also promote the use of APIs and cloud-native architecture, which allows for greater flexibility, scalability, and agility in software development. So, I am confident now I will not repeat my blunder and if the story helped you place [the MACH Alliance](https://machalliance.org/?ref=hoeijmakers.net) in the software landscape, I am happy. ### The right CMS for your online content URL: https://rob-hoeijmakers.ghost.io/content-management-system-cms/ Last updated: 2026-04-25T13:17:11.000Z As someone deeply immersed in the digital world, I know that picking the right Content Management System (CMS) is a bit like choosing the perfect coffee blend. It's not just about the taste; it's about how it complements your morning routine. Similarly, a CMS should not only manage your content but also seamlessly integrate into your business workflow. ### The Evolution of CMS: A Personal Perspective I remember the early days of CMS, back when they were just a way to manage a few web pages. Now, they've evolved into sophisticated platforms that handle everything from content publishing to user engagement and data analytics. ### Understanding CMS Functionality Rather than diving into a dry list of functionalities, let's look at them through the eyes of real users. There's Anna, a freelance content creator. She loves [WordPress](https://rob-hoeijmakers.ghost.io/content-management-system-cms/) for its simplicity and rich community support. Then there's John, a seasoned web developer, who swears by Drupal for its robustness and flexibility. 💡 Select the type of platform ([CMS, Content Platform, or DXP](https://www.hoeijmakers.net/cms-content-platform-or-dxp/)) based on the complexity and objectives of your digital strategy. A CMS is great for basic content management, a [Content Platform](https://www.hoeijmakers.net/content-platforms/) for enhanced content optimization, and a [DXP](https://rob-hoeijmakers.ghost.io/dxp/) for a comprehensive digital experience, especially for larger enterprises. Your choice should align with the specific needs and goals of your online presence. ### Why So Many CMSs? Understanding the Diversity The variety of CMS options is reflective of the diverse needs they cater to. For instance, Joomla strikes a fine balance, offering ease of use with some degree of customization, ideal for small to medium-sized businesses. On the other end, Adobe Experience Manager is a powerhouse, packed with features for enterprise-level operations. ### Real-Life Success Stories: Learning from Others I've seen some impressive transformations. A local café went from a barely-there online presence to a thriving online community using Squarespace. On a larger scale, a global retailer optimized its digital marketing strategy with Sitecore, personalizing customer experiences like never before. ### Choosing Your CMS: Key Considerations When it comes to picking a CMS, here's what I tell my clients: - **Ease of Use**: Can you and your team easily navigate the system? - **Customization**: Does it allow you to tailor it to your business needs? - **Scalability**: Will it grow with your business? - **Security**: How well does it protect your data? - **Support and Community**: Is there a community or professional support available? ### Wrapping Up: Making the Right Choice Selecting the right CMS is crucial. It's about finding a system that aligns with your current needs and can adapt to future changes. Reflect on your specific requirements and learn from others' experiences to make an informed decision. ## What's Next? Have questions or need some guidance in choosing the right CMS for your business? Feel free to reach out to me directly. I'm here to help you navigate through the options and find a solution that perfectly aligns with your online professional needs. [Let's connect and get your digital presence on the right track!](https://rob-hoeijmakers.ghost.io/about/) --- ## Some modern and popular CMSs Here is a list of some modern and popular content management systems (CMS) currently available: 1. [**WordPress**](https://wordpress.org): One of the most widely used and popular CMSs, [WordPress](https://rob-hoeijmakers.ghost.io/how-about-wordpress/) is open-source and free to use. It's easy to set up and use, and has a large community of users and developers. 2. [**Drupal**](https://www.drupal.org): Another open-source CMS, Drupal is known for its flexibility and scalability. It's often used for large, complex websites, and has a strong community of developers. 3. [**Joomla**](https://www.joomla.org) is an open-source CMS that is similar to WordPress and Drupal. It's known for its ease of use and wide range of built-in features, and is often used for small and medium-sized websites. 4. [**Adobe Experience Manager**](https://business.adobe.com/blog/basics/adobe-experience-manager) is a proprietary, web-based CMS that is part of the Adobe Marketing Cloud. It's known for its scalability and rich feature set, and is often used by large enterprises and organizations. 5. [**Sitecore**](https://www.sitecore.com) is a proprietary, web-based CMS that is known for its scalability and flexibility. It is often used by large enterprise, and for personalization and omnichannel experiences 6. [**Kentico**](https://www.kentico.com) is a web content management system and customer experience management system that offers a rich set of features for creating and managing websites, online stores, and intranets. 7. [**CloudCannon**](https://cloudcannon.com) is a web content management system (CMS) known for its Git-based infrastructure, real-time in-browser editing, and support for static site generation. It offers collaboration tools, customizable templates, and form handling capabilities, making it a versatile choice for web developers and teams. 8. [**Squarespace**](https://rob-hoeijmakers.ghost.io/squarespace/) is a proprietary, web-based CMS that is known for its ease of use and design-focused templates. It's often used by small businesses, individuals, and creatives to create and manage websites and online stores. (See: [Using Squarespace for a year](https://rob-hoeijmakers.ghost.io/squarespace-the-experiences-after-a-year/)) 9. [Kunstmaan CMS](https://kunstmaancms.be) is a Symfony-based, open-source platform offering flexibility with its modular "building block" design. It supports multilingual content, versioning, workflow management, and SEO tools, making it ideal for customisable, user-friendly websites. Some other notable systems are [Wix](https://rob-hoeijmakers.ghost.io/why-choose-wix-to-manage-your-valuable-content/), [RWS Tridion](https://www.rws.com), [CrownPeak](https://www.crownpeak.com), [Wagtail](https://wagtail.org), [Ghost](https://rob-hoeijmakers.ghost.io/ghost-blog/) and [Hubspot](https://www.hubspot.com). Please note that this is not an exhaustive list, there are many other CMSs available in the market and new ones are being developed all the time. Some of them might be more suitable for your specific use case than others. I also wrote about the difference between [CMS, Content Platform and DXP](https://rob-hoeijmakers.ghost.io/cms-content-platform-or-dxp/). *Missing your own CMS in the list,* [*let me know*](https://rob-hoeijmakers.ghost.io/about/)*.* ### Connect Facebook to Google Looker Studio URL: https://rob-hoeijmakers.ghost.io/connecting-facebook-instagram-google-looker-studio/ Last updated: 2026-04-25T13:48:01.000Z It is possible to connect Facebook and Instagram to Google Looker Studio (formerly Data Studio). To do this, you will need to connect your Facebook and Instagram accounts to the Facebook Ads API, and then use the API to extract data from your accounts and import it into Google Looker Studio. You can then create custom reports and visualizations using the imported data. There are **no Google Looker Studio connectors** though and so you will need to opt for a **third party connector vendor,** like Supermetrics. In this article, I will explain the concepts and the technical skills required to get connections to work. ## What is the Facebook Ads API? The Facebook Ads API is a **programming interface that allows developers to access and manipulate data related to Facebook ads**. This includes data on ad campaigns, ad sets, ads, and ad targeting, as well as performance data such as ad spend, impressions, and conversions. The Facebook Ads API can be used to automate and scale ad management tasks, such as creating, editing, and monitoring ad campaigns, as well as retrieving performance data for reporting and analysis. To use the Facebook Ads API, developers must have an active ad account, and they need to apply for and be granted access to the API. Access is granted on a case-by-case basis, and the level of access granted depends on the developer's use case and the data they wish to access. ![Person logged in into Facebook Ads API, part of the Facebook Marketing API](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/01/Facebook-Ads-API.jpeg) Facebook Ads API, part of the Facebook Marketing API ## Does connecting require technical skills? Connecting Facebook Ads API with Google Looker Studio does **require some technical skills** (to be honest: messy and challenging). The process involves working with APIs, OAuth2 protocols, and data visualization tools, which can be complex and require a certain level of knowledge and experience. Here are some skills you would need to connect Facebook Ads API to Google Looker Studio: - Knowledge of APIs and how they work: You will need to understand the basic concepts of APIs and how to use them to access data from Facebook ads. - Familiarity with OAuth2 protocol: To access the Facebook Ads API, you will need to be familiar with the OAuth2 protocol, which is used to authenticate and authorize access to the API. - Knowledge of programming languages such as JavaScript, Python or PHP: Some of the connectors require coding, you would need to know at least one of these languages to use them. - Familiarity with data visualization tools: You will need to be familiar with Google Looker Studio and the connector or data visualization tool you choose to import data from Facebook Ads API. - Understanding of Facebook Ads: Understanding how Facebook Ads works, the data structure and the metrics used will help you to define the data you want to retrieve and how to use it. In general, if you are already familiar with these concepts and technologies, connecting Facebook Ads API to Google Looker Studio should be relatively straightforward. However, if you are new to these technologies, it may take some time and effort to learn them before you can connect the two platforms. ## What are the costs of connecting? There may be costs involved in connecting Facebook and Instagram to Google Looker Studio, depending on how you choose to do it. Using the Facebook Ads API does not have a direct cost, but to access it, you need to have an active ad account, and you may be subject to ad spend minimums, which can range from $100 to $1,000 per month depending on your region and ad account type. Additionally, if you are using a third-party connector or data visualization tool that charges a fee to connect to the Facebook Ads API, there will be additional costs associated with using that tool. It's important to check the pricing and terms of service of any third-party tools you plan to use to make sure you understand the costs involved. ![Screen showing Google Looker Studio connectors](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/07/Google-Looker-Studio.jpg) Google Looker Studio connectors ## Steps to connect Facebook Ads API to Google Looker Studio To connect Facebook Ads API to Google Looker Studio, you will need to follow these general steps: 1. Apply for and be granted access to the Facebook Ads API. 2. Create an App in Facebook Developer's platform, you will need the app id and secret to make the connection. 3. Obtain an access token with the appropriate permissions to access the data you need from the Facebook Ads API 4. Use a connector or data visualization tool that supports the Facebook Ads API to import data from your Facebook ads account into Google Looker Studio. 5. Once you have the data in Google Looker Studio, you can create custom reports and visualizations using the data. It is important to note that not all connectors or data visualization tools support the Facebook Ads API, so you may need to research and test different options to find one that works for your specific needs. Also, you should be familiar with the use of APIs and OAuth2 protocols, as well as how to use Google Looker Studio and the connector you choose. Some connectors require coding and some others have a more user-friendly interface. If after reading all this, you find connecting the data source too complicated and your use case does not fit this level of automation, you can also fall back on manual import. Especially with smaller data sets or one-time reports that might be quicker and easier. ### Use Canva and AI to put a bike riding panda on a mug URL: https://rob-hoeijmakers.ghost.io/canva-ai-stable-diffusion/ Last updated: 2026-04-25T13:16:39.000Z Canva is a graphic design platform that allows users to create visual content such as posters, infographics, social media graphics, and more using a drag-and-drop interface. You can now also use [Canva](https://www.canva.com/ai-image-generator/) to generate an image with AI and use it in a regular workflow. 💡 Canva wrote me that they recently launched an all-in-one page dedicated to Canva AI Image Generators. This new page offers users convenient access to a wide range of image generators, templates, and design tips. With this tool, users can effortlessly create personalized images to meet their unique graphic design needs. Discover the limitless possibilities and unleash your creativity with [Canva's AI Image Generator ](https://www.canva.com/ai-image-generator/)today! Let me show you with an example. I will generate an image from text and then use smart mockup to transfer the image onto a mug. Image generation ## What generative AI does Canva use? Canva's AI image generator is based on a technology called **Stable Diffusion**, an open source text-to-image model that is capable of generating photo-realistic images for whatever you want to create a picture of. The app is part of Canva's design suite, so you do need a Canva account to use it. If you are working with social media, presentations, or other visual tasks, I definitely recommend to check [Canva](https://rob-hoeijmakers.ghost.io/canva/) out. ### Review: Mind over Media, Renee Hobbs URL: https://rob-hoeijmakers.ghost.io/review-mind-over-media-renee-hobbs/ Last updated: 2026-05-14T21:05:13.000Z ## Media in earlier times Because media used to be something that was pretty well-defined. Media was for watching news on TV, reading in the newspaper or magazine. Carefully trained professionals provided news coverage there; it was a more manageable domain. Now we have become a source ourselves and we all participate in media. News and entertainment are partly interchangeable, and much of the information is shared in emotionally charged and intrusive social environments. Of course, I didn't see it that way before. Back then, I also thought the media was a complicated subject. And propaganda and influence were there then, of course. So in one sense, there isn't much change and the key is to embrace today's situation if you want to at least understand it. And after acceptance and adaptation follows interest and curiosity. ### Propaganda At the heart of the book is the theme of propaganda. Once a word with a fairly neutral sound, now a meaning of manipulated and one-sided information to influence the mind. Propaganda used to be something for the mass media, but has become very personal and direct due to social media. How do you deal with this and make young people critical in this force field? What techniques are used to influence opinion? > As a propagandist, when you are promoting a solution, you intensify it using repetition, positive feelings, and careful design and composition. You make that solution seem like the best thing in the world. When you have a problem on your hands, you downplay it through omission, diversion, and creating confusion. ### Summary of the Mind over Media Mind over Media is a book written by Renee Hobbs that discusses the role of media literacy in the digital age. The book aims to help readers develop critical thinking skills and the ability to evaluate and analyse media messages in order to make informed decisions about the media they consume. Hobbs argues that media literacy is essential for helping individuals navigate the complex and often biased media landscape and make informed decisions about the information they encounter. The book covers a wide range of topics, including the history and evolution of media literacy, the impact of media on individuals and society, and **strategies for teaching media literacy to students and adults**. Ultimately, Hobbs advocates for the importance of media literacy in promoting critical thinking, citizenship, and democracy in the digital age. ### Review I really liked the book and I definitely think it is worth reading. A bit academic for my personal taste, and so I read it in conjunction with the much more tangible and direct Ryan Holiday book called [Trust me, I am lying](https://rob-hoeijmakers.ghost.io/review-trust-me-ryan-holiday/). The two books together gave me a rational and emotional update on how the media has changed the last decade and how to adjust and adapt to the new order. ### How web builders handle digital assets URL: https://rob-hoeijmakers.ghost.io/wix-squarespace-ghost-digital-assets/ Last updated: 2026-04-14T14:20:51.000Z For larger sites with broader features like community, courses, payment and multilingual, I currently work with [Wix](https://rob-hoeijmakers.ghost.io/why-choose-wix-to-manage-your-valuable-content/) or [WordPress](https://rob-hoeijmakers.ghost.io/how-about-wordpress/). Does the client need a more straightforward but beautiful looking site, I tend to work with [Squarespace](https://rob-hoeijmakers.ghost.io/squarespace-the-experiences-after-a-year/). For pure blogging, I choose [Ghost](https://rob-hoeijmakers.ghost.io/ghost-blog/). I already lined out what a [Digital Asset Manager (DAM) ](https://rob-hoeijmakers.ghost.io/digital-asset-manager-dam/)does and in this blog post I briefly outline if Wix, Squarespace and Ghost work with a digital asset manager ro manage media files in a media library. ## Does Wix have a DAM? Yes, Wix does have a Digital Asset Manager (DAM) feature. The Digital Asset Manager is a tool that allows you to upload and organize all of your media files in one place, making it easy to find and use them on your Wix website or other projects. You can access the Digital Asset Manager, or better media library, by clicking on the "Media" tab in the left-hand panel of the Wix editor. From there, you can upload and manage your images, videos, audio files, and other media assets. You can also create folders to organize your assets and use tags to label and categorize them. The Digital Asset Manager also allows you to edit your media files, add alt text for SEO purposes, and access your assets from any device with an internet connection. ![Digital Asset Management (DAM) in Wix](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/12/DAM-in-WIX.jpg) Digital Asset Management (DAM) in Wix ## Does Squarespace have a DAM? Yes, [Squarespace](https://rob-hoeijmakers.ghost.io/squarespace-the-experiences-after-a-year/) has a Digital Asset Manager (DAM) feature that allows you to upload, organize, and manage your media assets in one place. You can access the Digital Asset Manager by going to the "Content" tab in the left-hand panel of the Squarespace editor and clicking on the "Files" option. ![Squarespace image library](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/12/Squarespace-DAM.jpg) Squarespace image library From there, you can upload and organize your images, videos, audio files, and other media assets by creating folders and using tags to label and categorize them. The Digital Asset Manager also allows you to edit your media files, add alt text for SEO purposes, and access your assets from any device with an internet connection. Additionally, Squarespace's DAM integrates with the rest of the platform, so you can easily use your media assets in your website's design and content. ## Does Ghost blogging have a DAM? Ghost is a popular open-source blogging platform that is designed for professional bloggers and online publishers. It does not have a built-in Digital Asset Manager (DAM) feature, but you can still manage and use media assets on your Ghost blog. To manage and use media assets on your [Ghost blog](https://rob-hoeijmakers.ghost.io/ghost-blog/), you can either upload them directly to your blog's [content management system (CMS)](https://rob-hoeijmakers.ghost.io/content-management-system-cms/) or use a third-party tool to store and manage them. Some options for storing and managing media assets include using a cloud storage service like Google Drive or Dropbox, or using a dedicated Digital Asset Management platform like [Cloudinary](https://www.cloudinary.com) or [Bynder](https://www.bynder.com). Once you have your media assets stored and organized, you can use Ghost's built-in image and media embedding features to insert them into your blog posts and pages. You can also use Ghost's built-in image editing and optimization tools to resize, crop, and optimize your images for the web. ## Does WordPress have a DAM? Digital asset management within [WordPress](https://rob-hoeijmakers.ghost.io/how-about-wordpress/) is a powerful tool that allows you to manage, organize and publish your website content. It can also be used to manage files other than photos and videos, so it's not just reserved for photographers. The digital asset manager allows you to upload files and link them to the pages on your site. You can then use it to create thumbnails, resize images, easily change file permissions and more. You can add new images, videos and audio files quickly to your library using an uploader or by dragging files from your computer onto the interface. With WordPress, you can also organize media files into categories and tags (e.g., “slideshows” or “videos”). You can also create custom fields for each image or video and assign them custom labels (e.g., “price”). This makes it easier to find specific types of content later on when creating galleries or slideshows. Finding specific items using search queries (e.g., “images containing the word *dolphin*”). Please note, that is searching through the metadata, not the images themselves. ![WordPress Media Library](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/07/WordPress-Media-Library-2.jpg) WordPress Media Library ### Digital Asset Management plugins for WordPress There are many options for digital asset management within WordPress. Some of them are free and some of them you have to pay for, but all of them can help you manage your images and other assets in a better way. Here's a list of some of the most popular digital asset managers: **WP Smush Pro** \- This is the best option if you have a lot of images that need resizing and compression. It also has other features for optimizing your site performance and speed. **UpdraftPlus Backup** \- A free plugin that lets you back up all your files including database, uploads and media files. It also offers an automatic backup schedule. **WP Rocket** \- An excellent caching plugin that will make your site superfast while adding less than 1 second to page load times. ### Review: Trust me, I'm lying. Ryan Holiday URL: https://rob-hoeijmakers.ghost.io/review-trust-me-ryan-holiday/ Last updated: 2026-05-14T21:01:21.000Z Normally you probably review newer books I guess, but this 2012 book I read just now and because Ryan Holiday took social media as a topic you might think it is outdated. It is not. The writer took a stance from a much more eternal and moral point which gives a clear view on a moment in time but also at the underlying and durable moral patterns. Let’s dig in. ## What is the book about? "Trust Me, I'm Lying" is a book by Ryan Holiday that explores **the inner workings of the media industry** and the role that media manipulation plays in shaping public opinion and driving news cycles. The book discusses how media companies and individuals use a variety of tactics, such as creating fake news stories, promoting sensational headlines, and using social media to spread misinformation, in order to grab attention and drive traffic to their websites. Holiday argues that this focus on clicks and page views has led to a decline in the quality and integrity of journalism, and has contributed to **a culture of distrust and cynicism towards the media**. In addition to examining the tactics used by media manipulators, the book also provides practical advice on how to identify and avoid being duped by fake news and media manipulation. It encourages readers to be more critical and sceptical of the information they consume, and to seek out reliable sources of information. Overall, "Trust Me, I'm Lying" is a thought-provoking and informative look at the state of the media industry and the impact that media manipulation has on society. > IN THE INTRODUCTION I EXPLAINED A SCAM I CALL “trading up the chain.” It’s a strategy I developed that manipulates the media through recursion. I can turn nothing into something by placing a story with a small blog that has very low standards, which then becomes the source for a story by a larger blog, and that, in turn, for a story by larger media outlets. I create, to use the words of one media scholar, a “self-reinforcing news wave.” People like me do this everyday. ## My review I never read a book about the media, and social media, that had such profound impact as this book. It is not just the facts that make the story so compelling, it is the personal and moral struggling that give it depth and point the punch. Why would you lye? For money and fame? How far should you slide into the mud before it is too late to crawl back out on your own. That dilemma, next to the detailed facts and exciting and exact writing, made **the book worth every minute I spend with it**. And, even after all these years, it explained to me a lot about the media world we live in and the crazy years that raced by since the book was released. **I recommend this book to everyone who has a professional interest in media, or who wants to be an informed media consumer.** ![There is also an e-book on Trust me, I'm lying. I enjoyed the book so much, I wanted to have a hard copy as well.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/12/D089FB34-8D6B-42EC-8753-62E4847F5A54.jpeg) There is also an e-book on Trust me, I'm lying. I enjoyed the book so much, I wanted to have a hard copy as well. ### Try out ChatGPT yourself URL: https://rob-hoeijmakers.ghost.io/try-out-chatgpt-yourself/ Last updated: 2026-04-25T13:17:12.000Z The Chat Generative Pre-trained Transformer, or ChatGPT, is a chatbot model developed by OpenAI. It is based on the GPT (Generative Pre-trained Transformer) language model, a type of machine learning model that is trained on a large dataset of text and can generate human-like text when given a prompt. An important aspect of ChatGPT is that it is designed to have a conversation with a human user by generating responses based on the prompts it receives. It can be used in various applications, such as customer service chatbots, virtual assistants and social media bots. Some possible day-to-day applications of ChatGPT include answering questions, providing information and engaging with users in real-time. It is important to note that ChatGPT is a machine learning model and not a real person, so it does not always generate answers that are completely accurate or appropriate. ## How can I try ChatGPT for myself? Go to the address below and create an account: Then you can immediately start chatting with ChatGPT. You can ask your questions in English or in any other language. 💡 To fully unlock the potential and benefit from all the cutting-edge features of OpenAI's ChatGPT, we highly recommend considering the ChatGPT Plus subscription. For just $20 a month (as of 14/10/2023), you gain access to a premium experience that includes priority access, faster response times, and priority access to new features and improvements. In my personal opinion, it's well worth the investment for those who want to make the most of ChatGPT's capabilities and enjoy a seamless, enhanced AI-powered experience. ### Will ChatGPT replace the content writer? URL: https://rob-hoeijmakers.ghost.io/chatgpt-content-writer/ Last updated: 2026-04-25T13:17:12.000Z Tough question, but my first answer would be negative. [ChatGPT](https://rob-hoeijmakers.ghost.io/try-out-chatgpt-yourself/) for sure will have impact on the role and type of activities, but replacing is unlikely. Let me list what a content writer actually does and what the importance of the work is. And then see what will change and how to adapt. ## What does a content writer actually do? As the definition of content in the digital marketing world continues to evolve, the role of a content writer can vary. Some content writers focus specifically on one aspect of content production, such as researching and writing blog posts that are optimized for the search engines. Other content writers may provide a wide variety of services related to the production of engaging, relevant content. The fundamental role of a content writer begins with **understanding the target audience** of the company they are writing for. The best content writers know how to adapt their tone of voice and marketing efforts to the expectations of their target audience. At the same time, these professionals can write content that is optimized for both **human audiences and search engine algorithms**. Some tasks a content writer may be responsible for: - Keyword research - Content strategy - Production of content - Proofreading and editing - Publishing and outreach 💬 **If you are interested in writing for the web check out my articles on* [**the 4 professional writing roles*](https://rob-hoeijmakers.ghost.io/professional-writer-roles-for-the-web/) *and the* [**difference between a copywriter and content writer*](https://rob-hoeijmakers.ghost.io/difference-between-copywriters-and-content-writers/)**.* ## What is the importance of content writing? In today's digital age, the term content writing applies to the art of planning, writing, and publishing web content. Content writers may also be responsible for promoting and editing the content they create. Content writing is a form of inbound marketing designed to **attract an audience** to a specific company, product or digital place. The goal is to share valuable information about a product or brand, gain customer trust, improve engagement, and create sales opportunities. Today, content writing practices are used in the creation of various forms of content, including: **Blogging**: Blogs are a central part of building a strong content strategy, they can convey ideas related to products and brand while highlighting content expertise. Blogs can also be useful for SEO purposes, to build organic traffic and increase brand awareness. **Email**: Writing content for emails can help convert customers, as well as generate better engagement for your business. Email marketing also helps to build loyalty and increase the likelihood of repeat purchases. **Social media**: Social channels require content writers to convey different ideas and campaigns as effectively as possible. **Product content**: Selling products and services requires a unique set of writing skills. It requires balancing sales copy, storytelling and SEO strategies. **Brand journalism**: PR news releases, customer and brand stories, and internal communications can all need the support and guidance of a content writer. **White papers and e-books**: Content writers can help produce highly technical and professional content that requires a focus on a specific topic. This helps to reinforce the authority of the brand. **Video scripts**: If scripts are needed to create videos or even podcasts for a company's branding purposes, a content writer can help with this. ## Where will ChatGPT have impact on Content Writing I do content writing, but I am not a full time content writer. Now with ChatGPT my focus would shift from planning and designing content, learn and train writing [good text prompting](https://medium.com/mlearning-ai/the-anatomy-of-an-ai-art-prompt-dcf7d124406d), to editorial activities on AI-generated content and making it more search engine optimized. ChatGPT is, with a difficult word, a [stochastic parrot](https://www.calcalistech.com/ctechnews/article/vhhk7xrni), which means it can, in a very eloquent way, say stupid, dangerous and harmful things. This means there is an important aspect of fact checking that will need to be done. And also, ChatGPT works on existing information, human made. You would think it also needs new input, which still will be human made. So there is a premium on news. What is harder to predict is how ChatGPT will change the whole context where content has its informative value. Will the web be flooded with texts, choke or partially replace search engines? Will the readers look for cues that content is human made, or will they flock to ChatGPT to read answers there first hand? Exciting times ahead. ### Difference between Copywriters and Content Writers URL: https://rob-hoeijmakers.ghost.io/difference-between-copywriters-and-content-writers/ Last updated: 2026-04-21T17:25:47.000Z I wrote an article where I listed and explained [four main writing roles for online](https://rob-hoeijmakers.ghost.io/professional-writer-roles-for-the-web/): - Copywriter - Content writer - UX-writer/Content designer - Technical writer. Reading a post on LinkedIn that deepened my understanding, it was spot on. In this LinkedIn post, Chief Creative Officer & Founder at [RUMBLE](https://readytorumble.com.au), Andy Greenaway, explains very clearly the difference between Copywriters and Content Writers. ## The AIDA Model He argues, in short, that Copywriters are trained in persuasion and work and have been taught classic techniques like **AIDA**, which stands for: A : Grab ATTENTION I : Arouse INTEREST D : Stimulate DESIRE A : Ask for ACTION ![The AIDA model](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/12/IMG_7180.png) The AIDA model ## 5W's Content writers on the other hand are taught to follow the **journalistic 5W's**. 1\. Who 2\. What 3\. Where 4\. When 5\. Why. Check out his post to see how these difference sometimes mess up the work in creative agencies. [Andy Greenaway on LinkedIn: #marketing #copywriting #contentwriting #performancemarketing | 150 commentsWHY CONTENT WRITERS ARE DIFFERENT FROM COPYWRITERS. I was talking to friend the other day. He’s a senior manager in a creative agency and he was complaining… | 150 comments on LinkedIn![](https://static.licdn.com/aero-v1/sc/h/al2o9zrvru7aqj8e1x2rzsrca)LinkedInAndy Greenaway![](https://media.licdn.com/dms/image/D5622AQFUxKcziGErwg/feedshare-shrink_2048_1536/0/1670886024249?e=1697673600&v=beta&t=MeToHRgAVsFqvFsT7YSQqqSZnL6Grc3HY_6KQBLQA6s)](https://www.linkedin.com/posts/andy-greenaway%5Fmarketing-copywriting-contentwriting-activity-7008203945005240321-Rmd1/?utm%5Fsource=share&utm%5Fmedium=member%5Fios) *If you are interested in writing for the web, check out my articles on [the 4 professional writing roles](https://rob-hoeijmakers.ghost.io/professional-writer-roles-for-the-web/) and the question if [ChatGPT will replace Content Writers](https://rob-hoeijmakers.ghost.io/chatgpt-content-writer/).* ### Passkeys are set to replace passwords URL: https://rob-hoeijmakers.ghost.io/passwords-passkeys/ Last updated: 2026-04-21T09:26:09.000Z In that latest update, Apple presents Passkeys, their application of a new industry standard that aims to make passwords obsolete ([FIDO2](https://fidoalliance.org/fido2/)). In this article, I explain what that new way of security is and what it looks like for you as a user. ![macOS Ventura launch notes](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/12/2022-10-27_20-23-41.png) macOS Ventura launch notes ## What is Passkeys? The name Apple gives to the new way of logging in their ecosystem is Passkeys. And Passkeys is a passwordless way of logging in with the use of face recognition or fingerprint reading. Now, I myself still use the built-in password manager, but passkeys is going to greatly reduce the hundreds of passwords you maintain and promises a more or less frictionless way of logging in. 🔑 A passkey is like a special key or code that allows you to unlock something, like a door or a secure area, but in the digital world. It's a secret combination of characters, usually letters, numbers, or symbols, that you use to prove that you have permission to access a computer, a website, an app, or some other online service. ### Make account Because what do you do with Passkeys? You create a new account and in doing so, you get a personal key. A string of digital characters that is stored in a personal device, such as computer or mobile, and synchronized via the cloud. The account also comes with a unique, public key. The personal key goes into a key box which only you can open. This also makes it tied to a device of yours, and the locker can only be opened with bionic information (fingerprint, for example) or your Mac password. ### Log in The personal key must always remain local, but can be shared encrypted, via iCloud, across multiple devices. If you then log in, the public key will give you a challenge, a puzzle so to speak, that only you can solve with the personal key. You send that back and you get access. To demo I recorded both flows, make an account and log in, on an iPhone and made this video. With a thanks to [passkeys.io](https://www.passkeys.io) for sharing a test website. Using passkeys to make an account and then log in. On iOS for iPhone. ## Key cabinet So if FIDO2 and Passkeys catch on, you will soon no longer need passwords. You will manage a key cabinet. The disadvantage is of course that you can't do much without the key cabinet, which is local to your devices, but this is partly solved by working with [QR codes](https://rob-hoeijmakers.ghost.io/qr-code-explained/). But if you then want to help a six- or eighty-year-old, I predict it will be a challenge. Anyway, we are not that far yet. I have not yet come across Passkeys in the wild, am curious when that will happen for the first time. But as a site owner I would prepare for it, maybe it can give you an edge over the competition for some time. --- *Recently I produced an article with video on [2 Factor Authentication (2FA) on Apple Keychain](https://rob-hoeijmakers.ghost.io/2fa-within-apple-keychain/). Use this and there is no need for other password management solutions.* ### Digital Asset Management (DAM) key features URL: https://rob-hoeijmakers.ghost.io/digital-asset-manager-dam/ Last updated: 2026-04-14T14:22:07.000Z A digital asset manager (DAM) is a software solution that helps organizations **organize, store, and manage digital assets, such as images, videos, audio files, documents**, and other types of media. DAMs typically provide a central repository for digital assets, as well as tools for organizing, tagging, and categorizing them, and for controlling access to and publishing them. DAMs are often used by organizations in a variety of industries, including media, marketing, and advertising, to store and manage the large volumes of digital content that they [create, use, and distribute](https://rob-hoeijmakers.ghost.io/unlocking-digital-asset-managements-versatility/). ## Key features of a DAM Some key features of a DAM include: - **Storage**: DAMs typically provide a central repository for digital assets, with options for storing assets on-premises or in the cloud. - **Organizing**: DAMs may include tools for organizing and categorizing digital assets, such as tags, metadata, and taxonomies, to make it easier to find and retrieve them. - **Collaboration**: DAMs may provide tools for enabling teams to work together on digital assets, such as sharing, commenting, and version control. - **Security**: DAMs may include security features such as access controls, authentication, and encryption to protect digital assets and ensure that they are only accessed by authorized users. - **Integration**: DAMs may provide tools for integrating with other systems and technologies, such as [content management systems (CMS)](https://rob-hoeijmakers.ghost.io/content-management-system-cms/), [customer relationship management (CRM)](https://rob-hoeijmakers.ghost.io/customer-relationship-management-crm/) systems, and marketing automation platforms. Overall, a DAM helps organizations **streamline the process of storing, organizing, and managing digital assets**, and enables them to more effectively use and share those assets in their business processes and marketing efforts. 💡 I confused Digital Asset Management with Image Optimisation Services. There is overlap, they serve in the same space but there are also differences. I wrote an article on it: [What are Image Optimization Services for?](https://www.hoeijmakers.net/what-are-image-optimization-services-for/) ## DAM vendors There are many vendors that offer digital asset management (DAM) solutions. These vendors offer a range of DAM solutions with different features, capabilities, and pricing models. Some DAMs are designed for specific industries, such as media, marketing, or advertising, while others are more general purpose and can be used in a variety of contexts. It's worth noting that DAM solutions can be complex and may require significant customization and integration with other systems and technologies. Therefore, it's important for organizations to carefully consider their needs and goals when selecting a DAM vendor, and to carefully evaluate the capabilities and fit of the different options available. Here is a comprehensive list of vendors and their DAM product: - **Adobe Experience Manager:** Adobe Experience Manager is a comprehensive DAM solution that helps organizations manage and deliver digital assets efficiently. It offers advanced features for asset organization, collaboration, and seamless integration with other Adobe products. - **Aprimo:** [Aprimo](https://www.aprimo.com/?ref=hoeijmakers.net) provides a DAM platform that enables businesses to centralize, organize, and distribute digital assets effectively. It offers features like metadata management, version control, and workflows to streamline asset management processes. - **Bynder:** [Bynder](https://www.bynder.com) offers a cloud-based DAM solution designed to simplify digital asset management. It includes features like asset storage, organization, sharing, and collaboration, along with customization options to align with branding requirements. - **Canto:** [Canto](https://www.canto.com/?ref=hoeijmakers.net) is a DAM vendor that provides a scalable and user-friendly platform for managing digital assets. Its product includes features such as asset cataloging, metadata management, rights management, and integrations with popular third-party tools. - **Cloudinary:** [Cloudinary](https://www.cloudinary.com/?ref=hoeijmakers.net) offers a cloud-based DAM platform that focuses on delivering optimized digital assets for web and mobile applications. It provides capabilities for image and video management, manipulation, optimization, and delivery through a global CDN. - **Extensis:** Extensis provides a DAM solution that helps organizations streamline asset management processes. Its product offers features such as asset organization, metadata management, version control, and integrations with creative tools like Adobe Creative Cloud. - **Picturepark:** Picturepark stores any content no matter the format: images, videos, audio, PDFs, Word documents and virtually any other file or content types. They are all stored on our secure cloud, converted into custom-defined formats on import or download, with handy content previews available. - **Libris:** [Libris](https://librisdigital.com/?ref=hoeijmakers.net) is a DAM platform designed for visual media management. It offers features like asset organization, metadata management, search capabilities, and integrations with popular creative tools, enabling efficient management and distribution of visual assets. - **Tenovos:** [Tenovos](https://tenovos.com/?ref=hoeijmakers.net) provides a DAM solution that helps businesses optimize their digital asset lifecycle. Its product includes features like asset ingestion, enrichment, distribution, analytics, and governance, enabling end-to-end management of digital assets. - **Widen:** Widen offers a DAM platform that focuses on centralizing and sharing digital assets across organizations. Its product includes features such as asset organization, metadata management, workflow automation, and integrations with other business systems for seamless asset management. - **WebDAM**: WebDAM provides a cloud-based DAM solution designed to simplify asset management and collaboration. Its product offers features like asset organization, metadata management, sharing options, and integrations with popular creative tools for efficient asset workflows. - **WoodWing**: WoodWing offers a DAM solution that enables organizations to manage and distribute digital assets effectively. Its product includes features like asset organization, metadata management, collaboration tools, and integrations with other content creation and publishing systems. There is [the software review from Gartner](https://www.gartner.com/reviews/market/digital-asset-management?ref=hoeijmakers.net) for further reference. Sometimes the concept of the DAM is integrated into the content tool (CMS or DXP). I wrote an article [how digital assets are managed in Web Builders like Wix, WordPress, Squarespace and Ghost](https://rob-hoeijmakers.ghost.io/wix-squarespace-ghost-digital-assets/). If you still have questions, just [drop me a note](https://rob-hoeijmakers.ghost.io/about-rob-hoeijmakers/). ### Typeface: to serif or not to serif URL: https://rob-hoeijmakers.ghost.io/typeface-to-serif-or-not-to-serif/ Last updated: 2026-05-10T08:54:19.000Z A serif is an old word for stroke. Letters can have these serifs, and this is a relic from when they were carved into stone. People would first make the serifs and then fill in the letters. It was safer without an eraser. > Text should be treated as an integral work of (visual) art. But a serif is also, as a stroke, part of some more modern fonts. The serifs. Very crudely said, if you strip a letter of all these strokes, you get a sans-serif. ![Body text of my blog in Serif and in a Sans-serif.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/05/Untitled-design.jpg) Body text of my blog in Serif and in a Sans-serif. ## Knowledge about sans-serifs applied In my blog's template, created in [Ghost](https://rob-hoeijmakers.ghost.io/ghost-blog/), you can switch the letters of the texts between a serif and a sans-serif. And that's precisely what I've done now. From a more classic serif to a more modern and minimal sans-serif. In the case of blog, it is the [Inter by Rasmus Andersson](https://rsms.me/inter/). Everything looks smoother, more businesslike and more readable. I am very happy with it. It's a matter of appearance, a choice, but also a necessary choice. I find the sans-serif to be really more readable on the smaller screens that work with light. And to stick up for the serif as well, it looks more classic, chic and a bit more human. Less minimal. The serif works fine on paper and fine on e-reader. 💬 ****Typeface versus font** Typeface is the shared design of a set of letters, numbers and symbols. Like Helvetica, Roboto or Times New Roman. A typeface typically has different fonts. For instance, ****Helvetica 12 points light** and ****Helvetica 14 points bold** are different fonts, but the same typeface. ## Easy choice I could not find conclusive scientific proof for the readability claim, so you will have to make the choice with your intuition. But once you see it, then you can't miss it. Now I am happy with the new font, I hope you as a reader like it better too. I wondered why I didn't do switch over to Sans-serif much earlier, but I didn't really know the difference very well. With this article, I have tried to clarify and simplify choosing the right font, but there will always be an element of taste between serif or sans-serif. And something of a fashionable interpretation. But in most cases, for display on a screen, I would go with **a sans-serif typeface**. > When choosing typefaces, there are two key considerations: how does this type make us feel and how does this type work? The emotional response to the shape of letterforms is a very personal experience, and when readers first see type, they react to it in an emotional way before anything else. It's a major part of why so much emphasis is placed on choosing type even when it's not technically a part of typography. Google Fonts. ### Further reading - [Typography insights](https://rob-hoeijmakers.ghost.io/typography-insights/) - [From Typewriter to the Matrix](https://rob-hoeijmakers.ghost.io/single-spaced-fonts/) - [Typography, in motion](https://rob-hoeijmakers.ghost.io/typography-in-motion/) ### Intro to Digital Experience Platforms (DXP) URL: https://rob-hoeijmakers.ghost.io/dxp/ Last updated: 2026-04-14T13:23:55.000Z A digital experience platform (DXP) is **a software solution that helps organizations create, manage, and deliver personalized digital experiences to their customers**, employees, or other stakeholders. DXPs typically provide a range of tools and capabilities that enable organizations to **design, build, and publish digital content and applications**, such as websites, mobile apps, and online portals, as well as to track and analyse how users interact with these experiences. 🔍 ****Dive Deeper into Composable DXP**: Curious about the future of DXPs? One emerging trend is the rise of Composable DXPs. Unlike traditional, monolithic DXPs, composable platforms allow you to pick and choose the best-of-breed services and integrate them seamlessly. This modular approach gives organizations the flexibility to adapt to changing technologies and customer expectations without overhauling their entire system. Want to know more? Check out my in-depth article on [Composable DXP](https://rob-hoeijmakers.ghost.io/what-is-a-composable-dxp/). ### Key features of a DXP Some key features of a DXP may include: - **Content management**: DXPs often provide a content management system (CMS) that enables users to create, edit, and publish digital content such as text, images, videos, and audio. - **Personalization**: DXPs may include features that allow organizations to tailor the digital experiences they deliver to individual users based on their preferences, behaviours, or other factors. - **Integration**: DXPs often provide tools that enable organizations to integrate their digital experiences with other systems and technologies, such as customer relationship management (CRM) systems, enterprise resource planning (ERP) systems, and marketing automation platforms. - **Analytics**: DXPs may include analytics and reporting tools that help organizations track and analyse how users interact with their digital experiences, so they can understand how effective they are and make improvements. DXP solutions are often used by organizations in a variety of industries, including retail, financial services, healthcare, and media, to improve customer engagement and loyalty, drive sales and revenue, and streamline business processes. > A Digital Experience Platform provides the **architectural foundation** and modular services for developers and practitioners to create, orchestrate, and optimize **digital journeys at scale** \- to drive loyalty and new commerce outcomes across owned and third-party channels. ## Difference between DXP and CMS I just explained the DXP, and it sounds similar to a traditional CMS. A [content management system (CMS)](https://rob-hoeijmakers.ghost.io/content-management-system-cms/) is a software application that enables users to create, edit, and publish digital content, such as text, images, videos, and audio, on the web or other platforms. A CMS typically provides a user-friendly interface and a set of tools for organizing and managing digital content, as well as for controlling access to and publishing that content. While a DXP may include a CMS as part of its suite of tools, it typically offers **a wider range of capabilities beyond just content management**. In addition to a CMS, a DXP may provide tools for personalization, integration, analytics, and other functions that are designed to help organizations deliver more effective and engaging digital experiences. Therefore, the main difference between a DXP and a CMS is that a DXP is a more comprehensive solution for creating, managing, and delivering digital experiences, while a CMS is primarily focused on managing and publishing digital content. ![Simple illustration showing the overlap and difference in meaning between DXP and CMS](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/12/2022-12-30_20-13-51.png) Simple illustration showing the overlap and difference in meaning between DXP and CMS 💡 ****The genesis of DXP.** The DXP’s have grown around technical solutions for digital content delivery, like content management systems, and out of solutions that unified Online Marketing solutions across different digital touchpoints, like Web Experience Management (WEM). And now they have merged more and more into a single category. Theoretically, DXP's combine “[Content Management](https://rob-hoeijmakers.ghost.io/content-management-system-cms/)” with “Engagement Management”, to allow teams the flexibility of creating omni-channel content, with the reliability of agile workflows, data-driven insights, and consistent user-journeys. ## DXP Vendors There are many vendors that offer digital experience platform (DXP) solutions. Some examples include: - Adobe Experience Platform - Acquia - Bloomreach - Episerver - Kentico - Sitecore - OpenText - Oracle - Salesforce - SAP These vendors offer a range of DXP solutions with different features, capabilities, and pricing models. Some **DXPs are designed for specific industries**, such as retail, financial services, or healthcare, while others are more general purpose and can be used in a variety of contexts. It's worth noting that DXP solutions can be complex and may require significant customisation and integration with other systems and technologies. Therefore, it's important for organizations to carefully consider their needs and goals when selecting a DXP vendor, and to carefully evaluate the capabilities and fit of the different options available. ### Is WhatsApp social media? URL: https://rob-hoeijmakers.ghost.io/is-whatsapp-social-media/ Last updated: 2026-05-10T08:54:19.000Z To research a bit more I did a poll on Twitter with this question, it turned out exactly half of the audience thought WhatsApp to be a social medium and the other half not! Almost exactly 50/50. > Is Whatsapp een vorm van Social Media? > > — Rob Hoeijmakers (@robhoeij) [October 23, 2022](https://twitter.com/robhoeij/status/1584122131662524416?ref%5Fsrc=twsrc%5Etfw) Survey on Twitter/X asking people if they consider WhatsApp is social media. Almost fifty/fifty says yes or no. ## Social networking versus Social media Now I read this interesting and thought-provoking article by **Ian Bogost** in The Atlantic titled [The Age of Social Media Is Ending](https://www.theatlantic.com/technology/archive/2022/11/twitter-facebook-social-media-decline/672074/). And next to a very critical view on social media, he makes a strong separation between Social Networking and Social Media. And now the misunderstanding on WhatsApp is clearer to me. It started out as Social Networking, but mainly because of groups it evolved to being more like Social Media. This is a really helpful and I flipped sides. I now also see WhatsApp as social media. The benefit of this knowledge? I can treat WhatsApp as such, change my etiquette and ignore messages in WhatsApp more. Maybe that works for you as well 😄 I attached the article, it is absolutely worth reading. --- ### Further reading - [The Age of Social Media Is Ending](https://www.theatlantic.com/technology/archive/2022/11/twitter-facebook-social-media-decline/672074/) ### Using markdown for blogging URL: https://rob-hoeijmakers.ghost.io/using-markdown-for-text/ Last updated: 2026-05-10T08:54:19.000Z Markdown is a **lightweight markup language** that is used to format text in a way that is easy to read and write. It is often used to format text in online forums, readme files, and documentation, and is designed to be simple and easy to use. One of the main benefits of Markdown is that it allows you to format text without the need for complex formatting tools or HTML coding. It uses a set of simple, easy-to-remember symbols to format text, such as asterisks for bold text and underscores for italic text. This makes it very fast and easy to apply formatting to your text, and it also makes it easier for others to read and understand your formatting. Another benefit of Markdown is that it is **widely supported** across a variety of platforms, including online forums, blogging platforms, and documentation tools. This means that you can use it to format text in a way that is consistent and easy to read, regardless of where you are publishing your content. In summary, the benefits of Markdown include its **simplicity, ease of use, and wide support across platforms**. It is a powerful tool for formatting text in a way that is easy to read and understand, and it is widely used in a variety of settings. ## Types of markdown There are several different versions or flavors of Markdown that have been developed over the years. Some of the most commonly used versions include: 1. **Standard Markdown**: This is the original version of Markdown, developed by John Gruber in 2004\. It is a simple and easy-to-use markup language that is designed to be easy to read and write. 2. **CommonMark**: This is a more formalized version of Markdown that is designed to be more standardized and easier to parse. It is based on the original version of Markdown, but adds additional features and clarifies some of the ambiguous parts of the original specification. 3. **GitHub Flavored Markdown (GFM)**: This is a version of Markdown that is used on the popular code hosting platform GitHub. It is based on the original version of Markdown, but adds additional features, such as support for task lists and tables. 4. **MultiMarkdown**: This is a version of Markdown that was developed by Fletcher Penney as an extension of the original Markdown specification. It adds additional features and functionality, such as support for footnotes, tables, and metadata. 5. **Pandoc Markdown**: This is a version of Markdown that is used by the Pandoc document converter. It is a very flexible and powerful version of Markdown that supports a wide range of formatting options, including tables, footnotes, and metadata. In summary, there are several different versions of Markdown that have been developed over the years, each with its own set of features and capabilities. The most commonly used versions are Standard Markdown, CommonMark, GitHub Flavored Markdown, MultiMarkdown, and Pandoc Markdown. I use MarkDown to write my own blog using the writing app Ulysses. You can check out [my own, lightweight workflow for writing ](https://rob-hoeijmakers.ghost.io/blog-writing-workflow/)here. --- ### Further reading - [Markdown in Google Docs: A Practical Step Forward](https://rob-hoeijmakers.ghost.io/markdown-google-docs/) ### Create images for free with Playground AI URL: https://rob-hoeijmakers.ghost.io/synthetic-images-playgroundai/ Last updated: 2026-05-10T08:54:20.000Z Maybe you've heard of AI generated images through software like DALL-E and Stable Diffusion, but don't yet know where to get started with this yourself? I already explained [what you can do with DALL-E](https://rob-hoeijmakers.ghost.io/playing-with-ai-system-dall-e/) but if you want easier access to this groundbreaking technology and create your own images within no time then [**Playground AI**](https://playgroundai.com)is an ideal destination. And right now, it's still free, so be sure to give it a try. ## Play with AI images in Playground AI **Playground AI** is a website where you can experiment with the generative artificial intelligence techniques of Stable Diffusion and DALL-E. With providing a textual prompt, **a prompt**, synthetic images are created. New images that you can then marvel at, share or use for your project. Until recently, it was difficult to get to know and experiment with this AI technology quickly, but with Playground AI you easily create an account (it uses Google) and you can get started. Now this service is still new, and free, so check it out while it still is. The interface is uncluttered and easy to use. In fact, you can enter a prompt right away and click generate. ![Playground AI allows you to switch between Stable Diffusion and DALL-E](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/02/download.webp) Playground AI allows you to switch between Stable Diffusion and DALL-E ## Hints for prompts The text prompt is important for the style and content of the image, but there are also a few other ways to achieve the desired result. Filters provide a style, an atmosphere. So are not the colour filters we know from photo programmes. You can also insert your image to give the algorithms some inspiration. Try to use meaningful and rich descriptions of the scene you work with. Here is an example of the prompt that generated the Gecko on the top of this page: > Fullbody headshot of vibrant adorable anthro gecko in space, gecko made of cosmic space dust, subsurface scattering, ambient, misty, cinematic photograph, incredible detail, ultra quality Have fun! --- ### Further reading - [Free AI image generator: Art, Social Media, Marketing | Playground](https://playground.com) ### Quick tip: Do screen recording on the iPhone URL: https://rob-hoeijmakers.ghost.io/quick-tip-do-screen-recording-on-the-iphone/ Last updated: 2026-04-25T13:17:08.000Z Recording your screen requires a one-off preparation and then some dexterity. ## Turn it on once To achieve this, turn on Screen Recording under the *Control Centre*. You can find it under *Settings > Control Centre*. There are two lists there. Tools that are active and tools that are inactive. Look for Screen Recording under the inactive tools and **activate it by clicking on the plus sign**. ![Activate screen recording](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/10/IMG_5768.jpg) Activate screen recording ## Screen recording Now that it is on in the Control Centre, you can make a recording. Open the Control Centre by swiping down in the right-hand corner of the screen. In the Control Centre there now is an icon for recording and when you press it, a countdown starts. Close the Control Centre and record what you wanted to record. ![The button for Screen Recording in the Control Centre](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/10/IMG_7E41A83EFE63-1-1.jpeg) The button for Screen Recording in the Control Centre While recording, a red icon will appear in the top-left corner. Once the recording is finished, press that icon, and you can stop it. The clip will be saved with your photos. ![Screen recording is active.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/10/DraggedImage.png) Screen recording is active. Have fun with it! ### Do dashboards protect against flying mud? URL: https://rob-hoeijmakers.ghost.io/why-dashboards/ Last updated: 2026-05-10T11:15:42.000Z I took a short course on AI in education. And one of the lessons was about using dashboards. I used to know that word only for the car control panel, now there are countless dashboards I use. Usually, they are applications that make underlying, dynamic data (observed actions) transparent. Statistics of website traffic, progress of campaigns, or just an overview of screen usage. ## Different functions of a dashboard In the course I learned that there are several functions of dashboards, and I was shown these four gradations A dashboard can: 1. Inform 2. Advise 3. Guide 4. Induce action To make the comparison with a car's dashboard. It informs you about the current speed, it advises you about future servicing, it guides you to find a route, and it forces you to put on your seatbelts (or suffer an increasing and constant beeping). ## Etymology of the dashboard Explaining where the word dashboard comes from requires this picture of a carriage. The screen on the left, where the lady is standing, that is the dashboard. It prevents mud, via the horses' hooves, from being dashed into the carriage. ![The lady on the left is dusting the dashboard.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/10/01a-Victorian-carriage-donated-to-National.jpg) The lady on the left is dusting the dashboard. You can mount clocks, buttons, and lights on such a board to give overview and control. To drive a car. ![Modern dashboard in a car.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/10/IMG_9134-2.jpeg) Modern dashboard in a car. Or to guide a company, organization, or group of children with the help of a computer screen. It's funny to see that a dashboard once was used to keep mud and dirt out. ### How useful is a dashboard? That is a up for debate 😂 Check out the PowerBI dashboard post on LinkedIn. Power BI is an interactive data visualization software product developed by Microsoft with a primary focus on business intelligence. It is part of the Microsoft Power Platform. But the below is probably a also partially true for many other dashboards. ### Further reading - [Typography, in motion](https://rob-hoeijmakers.ghost.io/typography-in-motion/) ### Lions on the loose in Prague! By Ernest “GPT-3” Hemingway URL: https://rob-hoeijmakers.ghost.io/lions-on-the-loose-in-prague-by-ernest-gpt-3-hemingway/ Last updated: 2026-04-25T13:17:07.000Z ## The request to the machine So here is the question I asked by presenting this prompt to GPT-3: > Write a blog article about lions on the loose in Prague in the style of Ernest Hemingway. ## The article it wrote Here is the article that the system wrote. *Can you believe it? Lions, of all things, roaming the streets of Prague!* *I was out for a walk early this morning when I saw something large and furry moving around in an alleyway. At first, I thought it was a dog. But then I realized that it was much too big to be a dog. And it had a mane!* *I couldn't believe my eyes. I was seeing a lion! And not just any lion, but a big male lion. He must have escaped from some circus or zoo.* *I was about to turn and run when I saw that there were two more lions with him. A male and a female. They were both looking right at me.* *I knew I should have been scared. But for some reason, I wasn't. I just stood there and looked at them.* *And then, to my amazement, the male lion started to speak to me. In perfect English!* *"Hello, there," he said. "My name is Leo. These are my friends, Lola and Max. We're just taking a little stroll around the city. Would you like to join us?"* *I didn't know what to say. I just stood there with my mouth open.* *But then I realized that this was my chance to do something really exciting. So I said yes.* *And that's how I ended up spending the morning walking around Prague with three lions!* *It was one of the most amazing experiences of my life. I'll never forget it.* ## Try for yourself Quite a nice story! As to how well the style succeeded, I'm going to ask an expert. A writer in America. For now, you will need to decide for yourself how much Ernest Hemingway shines through. Want to try for yourself? Go to Open AI, create an account and put some questions forward: [OpenAI PlatformExplore developer resources, tutorials, API docs, and dynamic examples to get the most out of OpenAI’s platform.![](https://openaiapi-site.azureedge.net/public-assets/d/fa971f6e97/favicon.svg)![](https://cdn.openai.com/API/images/opengraph.png)](https://beta.openai.com/playground) Here is my question to the AI oracle: ![The prompt to GPT-3](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/10/2022-10-12_20-20-04.png) The prompt to GPT-3 ### Internet tracking: how to put a stop to it URL: https://rob-hoeijmakers.ghost.io/advertisements-without-being-followed/ Last updated: 2026-04-25T13:17:07.000Z ## How does that tracking work? An ad network may leave a cookie on your computer when you visit a web page that makes you identifiable. That doesn't seem like a big problem at first glance. But it becomes a concern if it turns out that the network consists of all the websites you visit. Think of it this way, a cookie from such a network is like **having a note pinned to your back unnoticed that says: I'm crazy**. Everywhere you go you get a pitying or mocking look and you yourself have no idea why. Very annoying, almost bullying behaviour. That in itself is annoying, but it still seems like the network doesn't know that much about you. A note on the back and that was that. But because it is a network, all data from other visits are linked. And the networks also exchange data about you among themselves. Then a grimly precise picture of your activities emerges. ![Trackers on a major website.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/12/2022-10-05_11-32-49.jpeg) Trackers on a major website. ## And it gets personal And because you personally give out information within this network, for example by logging into Google or social media sites, or filling in forms. This quickly creates a more **complete picture of your online behaviour**. Then, because you also log in on your mobile, via site or app, your behaviour on mobile also becomes linkable. You can imagine that quite a lot becomes known about you. The note then no longer contains a silly text, but **a highly personal number**. The network's files then record your behaviour (psychographic) and additional data such as your name, email address, region or IP number. The information about you is then traded by data brokers to advertisers, who can target you (as a segment). For targeted offers made with characteristics or behaviour. But also political movements that analyse whether you belong to their target group based on your behaviour and expressions, as in the big Facebook scandal. ## A solution through the publisher The other day I read a report from a major website ([Tweakers](https://tweakers.net)) that had decided to stop offering ads through the networks and their trackers. The publisher chose to place the **ads based on the context of the page**. So, no more peeking by strangers within a network on that site. Everything goes through their domain, nothing is actually known about the visitor, and no information is passed on to third parties. A brave move. How does it work? Their ad platform places the ads itself. In this case, through a party called [Adhese](https://adhese.com/consent-cookie-less.html). It does still set a voluntary cookie that ensures you don't see the same banner too often. But if you already find that objectionable, you can also reject that cookie. In any case, for me as a visitor it provides a lightning-fast website where you can also assume that you will be received discreetly. A nice thought and a way that for conscious visitors actually counts as a commendation and a sign of trust. 🍪 Does my website track you? No. I use [Plausible for analytics ](https://www.hoeijmakers.net/cookie-free-analytics/)and that is a cookie free solution. It is very tempting to add free Google Analytics for tracking because that gives insight but also can be the basis for remarketing. So presenting an advert in Google's search engine or another site to someone. ## A solution for you as the visitor One solution is to prevent the website you visit from setting tracking cookies. There are quite many solutions for this, but who has time for that? Because it also has to be done on the many devices of the partner, the children, etc. And then another browser breaks down and needs to be reinstalled, or there are also multiple browsers. It's almost impossible to keep up. However, it is now built into Apple's Safari browser by default (Intelligent Tracking Prevention, ITP). And there, blocking is also on by default. If we would rather not be fooled, and spied on, it is to be hoped that this will soon become the standard setup. Google says they aim to make standard blocks part of their Chrome browser in 2023\. ![Safari blocking trackers track record.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/10/safari_privacy_report.png) Safari blocking trackers track record. ### Playing with AI-system DALL-E URL: https://rob-hoeijmakers.ghost.io/playing-with-ai-system-dall-e/ Last updated: 2026-04-25T13:16:38.000Z For quite some time I saw the synthetic images of DALL-E popping up, and I read a lot about it. It was only accessible to invited guests, but now anyone can play with it. And so, I played with this fascinating technology. ## The name DALL-E DALL-E is the name given by the company OpenAI to their AI system. The name DALL-E is a portmanteau of the names of Pixar character WALL-E and Spanish surrealist artist Salvador Dalí. A well-chosen name for the often surreal images created by the computer. ![Wall-e and Dalí on DALL-E.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/09/gr9rc3xtzk091.jpg) Wall-e and Dalí on DALL-E. ## What is DALL-E? For the user, it is software that lets you use a textual clue (prompt) or existing image to request a new image. This is then not searched, but created. A new, synthetic image. Making those images was not the big breakthrough, but the fact that by giving a textual command, you get an image. The engine behind this is GPT-3, a so-called Large Language Model. A model for everyday language processing based on massive sets of information. So first, the given text in natural language will be interpreted by GPT-3 and broken into bits of meaning. Then, using those words, the system creates a new image from snippets it has scraped from the internet. So actually, the images are a sort of collages. 💬 ChatGPT generates well stated texts, and it will change the web and content writing for sure. But where and [how will we feel the impact of generative text AI](https://rob-hoeijmakers.ghost.io/chatgpt-content-writer/)? ## An example of playing with DALL-E The underlying technology is difficult to understand in detail, but you can get an intuitive feel for it. Especially if you get to work with it yourself, you will learn a lot. To get started, go to OpenAI's website and first create an account: And basically, you can then get started right away. By entering an English text or your photo. Here is the example I work with: > **A childs drawing of a happy dog singing the blues** That got me these four funny children's drawings. The dog even has a blue tongue on two of them: ![A childs drawing of a happy dog singing the blues](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/07/2022-09-29_17-49-12.jpg) A childs drawing of a happy dog singing the blues Not bad for a first attempt. You are shown four variations each time and by choosing one image you can request further variations. On the left, the original and then the four images from the second round. ![Variations of: A childs drawing of a happy dog singing the blues](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/07/2022-09-29_17-50-59.jpg) Variations of: A childs drawing of a happy dog singing the blues The programme provides you with ample tips. Ask for 3D renders, or add your favourite painter or style. For example, Japanese woodblock art ukiyo-e. ![Peacocks and blossom in the style of ukiyo-e.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/07/DALL-E-2022-09-29-18.32.50---peacocks-and.jpg) Peacocks and blossom in the style of ukiyo-e. ## Playtime over But not immediately game over. You get a few free credits every month, but you can buy extra if needed. They cost $15 per 115 credits at the time of writing. I must honestly say that I could play for hours and some images go straight into my dreams. So scary and realistic. It's absolutely fascinating, and it's a game with language too. What can I describe in a way that something comes out that I wanted or that surprises me. Some assignments give extraordinary results, and then others are mediocre. What struck me is that all the results are square, which is not always what I require ([why you need portrait and landscape](https://rob-hoeijmakers.ghost.io/portrait-and-landscape-images-on-social-media/)). But actually, many images are perfectly usable. There are, of course, questions about copyright. There are also ethical questions about deepfakes, for example. Some images are deceptively real, like the hawk on the handlebars of a bicycle. DALL-E provides a spectacular new technology that will have influence that stretches far beyond the digital world. ![Red-tailed hawk on the handlebars of a bicycle.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/07/DALL-E-2022-09-30-07.03.43---photo-of-a-red-tailed-hawk-on-steering-wheel-of-a-bicycle.jpg) Red-tailed hawk on the handlebars of a bicycle. *Your own AI images promptly and through a text prompt. With [Playground AI ](https://rob-hoeijmakers.ghost.io/synthetic-images-playgroundai/)you create images using Stable Diffusion and DALL-E within no time. Playground AI is a free AI image generator.* ### Review: Influence, Robert Cialdini URL: https://rob-hoeijmakers.ghost.io/review-influence-robert-cialdini/ Last updated: 2026-05-17T13:35:22.000Z I love reading about technology and behaviour. Books that inspire me, that I can keep re-reading, I summarize and share them here, alongside reviews of current books. This book by Robert Cialdini gave me some great insights. It felt like a magician explaining to me how his tricks work. Ah, so that's how it happens! I see it happening before my eyes and I didn't understand it. The book clarifies some **six social-psychological principles that can influence people**. Now, in numerous texts, videos and interactions, I can see the underlying principles and respond faster and better to what is happening in myself and in the outside world. I also use the principles to some extent when writing content for sales websites or landing pages. A very clear, important book. For yourself in ordinary life, for understanding marketing and sales processes. I can definitely recommend it, although I found the book a bit long-winded at times. But then again, I had already read a summary (via Blinkist) and it hadn't got through to me or stuck with me enough anyway. It's good to work hard sometimes and turn a few pages unread. I read the book as an e-book and highlighted quotes and took notes ([on the Kobo](https://rob-hoeijmakers.ghost.io/kobo-highlights-and-notes-to-your-mac/)). ## Reciprocity ![A valuable but more or less free gift. I was tricked into the idea to need to return a lot.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/09/E61483CC-980E-459A-B8C6-E5BC2BC6D392_1_105_c-1.jpeg) A valuable but more or less free gift. I was tricked into the idea to need to return a lot. People generally feel obliged to return favours they are offered. Sly salespeople often capitalize on this trait by offering potential customers a small gift. Studies showed that even if the gift is unwanted, it will influence the recipient to reciprocate. > The noted archaeologist Richard Leakey ascribes the essence of what makes us human to the reciprocity system: "We are human because our ancestors learned to share their food and their skills in an honoured network of obligation, "he says. A variation on this theme, known as the "door-in-the-face technique", is to ask for a particularly large favour. When this is refused, a smaller favour is requested. This is likely to be successful because a concession from one party (toning down the favour) will be reciprocated by a concession from the other party (agreeing to the smaller favour). > The need to reciprocate had transcended great cultural differences, long distances, acute famine, and immediate self-interest. ## Commitment and consistency ![A valuable but more or less free gift. I was tricked into the idea to need to return a lot.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/09/3C0816C3-D1F3-4294-A6DE-5E469483A86D_1_105_c-1.jpeg) Once you belief in something there is a risk not to be open to change. People have a general desire to appear consistent in their behaviour. People also generally value consistency in others. > It is, quite simply, our nearly obsessive desire to be (and to appear) consistent with what we have already done. Once we have made a choice or taken a stand, we will encounter personal and interpersonal pressures to behave consistently with that commitment. Those pressures will cause us to respond in ways that justify our earlier decision. This desire for consistency can be exploited by getting someone to make an initial, often small, commitment, known as the "foot-in-the-door technique". Subsequently, requests can be made that are consistent with this initial commitment. People also have a strong desire to honour commitments made by further justifying and motivating them. Sellers exploit this with the so-called "low-ball" technique, where buyers agree to an attractive offer before it is modified to be less favourable to them and more profitable for the seller. This pattern of behaviour towards or resulting in a negative outcome is called escalation of commitment. ## Social proof ![See how others do it as well.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/09/7E744FE0-7BCC-488F-959C-0C7907CDCFE4_1_105_c-1.jpeg) See how others do it as well. People generally look at other people similar to themselves when making decisions. This is especially noticeable in situations of uncertainty or ambiguity. > Once again we can see that social proof is most powerful for those who feel unfamiliar or unsure in a specific situation and who, consequently, must look outside of themselves for evidence of how best to behave there. This trait has led salespeople to give false information about what others are doing. Examples include staged interviews in television advertisements or "infomercials". ## Liking ![Beauty can make you soft and pliable. ](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/09/749483D7-4DD5-4AE3-9F90-C34194430C39_1_102_a-1.jpeg) Beauty can make you soft and pliable. People are more likely to accept offers from people they like. There are several factors that can influence people to like some people more than others: - Physical attractiveness can give people a halo effect that makes others more likely to trust them and see them as smarter and more talented. - People tend to like people who are most like themselves. - People tend to like people who give them compliments. - People with whom they are forced to work together to achieve a common goal tend to form a trust with those people. - People tend to like people who make them laugh. For example, many lectures start with a joke. Each of the above methods may not help influence people, but in combination, their effects can be magnified. > The clearest illustration I know of the professional exploitation of the liking rule is the Tupperware party, which I consider the quintessential American compliance setting. ## Authority ![It looks impressive, it must be true. ](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/09/19EEFA76-C826-4B7E-95FC-B29F7F4F058C_1_105_c-1.jpeg) It looks impressive, it must be true. Stanley Milgram's Milgram experiment provided some of the most stunning insights about the influence authority can have on others. > Con artists, for example, drape themselves with the titles, clothes, and trappings of authority. They love nothing more than to emerge elegantly dressed from a fine automobile and to introduce themselves to their prospective “mark” as Doctor or Judge or Professor or Commissioner Someone. They understand that when they are so equipped, their chances for compliance are greatly increased. Each of these three types of symbols of authority has its own story and is worth a separate look. People often respond automatically to orders from authorities and even to symbols of authority (such as academic degrees, uniforms, expensive cars, etc.), even when their instincts suggest that the orders should not be followed. ## Scarcity ![The last one, hurry up!](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/09/B3250793-C222-45A1-9DB7-8133C90283CB_1_105_c-1.jpeg) The last one, hurry up! People tend to want things when they are less available. This has led advertisers to advertise goods as "limited availability" or "only for a short time". It has also been shown that when information is restricted (such as by censorship), people want the information more and value it more. Items also gain higher value when they were once available in large quantities, but have now become scarce. > The idea of potential loss plays a large role in human decision making. In fact, people seem to be more motivated by the thought of losing something than by the thought of gaining something of equal value. ### Get found in Google Discover URL: https://rob-hoeijmakers.ghost.io/get-found-in-google-discover/ Last updated: 2026-04-25T13:17:13.000Z In the Google Search Console reports, I suddenly saw visits from Discover, and I had no idea what that was. So, I explored, **Google Discover**. ## Google Search Console Suddenly, I noticed an additional section in the navigation of [Google Search Console](https://rob-hoeijmakers.ghost.io/google-search-console-gsc/): DISCOVER. The console is Google's search engine optimization (SEO) dashboard and an important indicator of how much traffic my blog receives. I was naturally curious and saw a report about traffic on Google Discover. It turned out to be about a page on my website about [Pixels per Inch (PPI)](https://rob-hoeijmakers.ghost.io/pixels-per-inch-ppi/), a unit of measurement for the resolution of digital screens. I was able to click through on the report and saw that the page is only shown in **India**. ![On the left the new navigation item, on the right the traffic report in Google Search Console.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/12/2022-09-14_14-38-49.webp) On the left the new navigation item, on the right the traffic report in Google Search Console. But then what is **Google Discover**? ## Google Discover explained The Google Discover feature, for me as an iPhone user, shows up only within Google's iOS app. Within Android, the mobile operating system, it has a more prominent role. Google Discover is a **stream of news** that can be seen in the app. There are numerous streams available based on meta information about the content, and a Google user is shown information **based on profile and behaviour**. You can also self-report on the quality of information in the stream. ![On the left the new navigation item, on the right the traffic report in Google Search Console.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/12/Google-Discover-on-iPhone.jpg) Screenshot of Google Discover in an iOS Google app. ## How to get into Google Discover? So now there is a page of mine in Google Discover in India. I requested some tips on how to display the page even better to get a higher Click-Through Rate (CTR) but I actually have no idea how the article became part of Discover. The selection must almost be **editorial work** from Google for sure, I would think, but my blog has never been logged in as a news source as far as I know. I also read that the effect would be very temporary because it is news. But the article is older again and has been appearing in Discover for days. So still many questions, but this gives you an idea what it is for the reader and how you can see it as a publicist. --- **Update September 2023:** Discover isn't editorial in the traditional sense. It's a feature that shows users personalized content based on their interests and preferences. While Google's automated systems curate the content that appears in Discover, it's not a manually curated editorial process. The content shown is based on signals and algorithms that aim to match the user's interests and provide them with relevant and engaging information. So, it's more about personalization and algorithmic selection rather than traditional editorial decision-making. To increase the likelihood of your content appearing in Discover, here are some tips: 1. Use descriptive page titles that capture your content's essence. 2. Include high-quality, compelling images in your content. 3. Avoid using misleading or exaggerated details in preview content. 4. Provide timely, engaging, and people-focused content. 5. Ensure your content is suitable for interest-based feeds and avoid shock or unexpected content. *Read more on Discover on Google's page on [Discover and your website](https://developers.google.com/search/docs/appearance/google-discover).* ### Export Kobo highlights and notes to your Mac URL: https://rob-hoeijmakers.ghost.io/kobo-highlights-and-notes-to-your-mac/ Last updated: 2026-04-25T08:26:00.000Z ## Difference between a highlight and a note On the Kobo, you can highlight text and with highlighted text, you can make a note. Together, these two are called annotations. ![Example of a highlight on a Kobo e-reader](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/09/DraggedImage.jpeg) Highlighting on Kobo, adding notes. Highlighting is done with the touch screen. I find it rather clumsy, you sometimes feel like a drunken fool when you skim over the screen, but eventually, it works. Put your finger on the beginning, swipe to the end, and then, if necessary, touch up the selection with the dots. Once the highlighting is done, you can make a note of that section. ## Get notes to the Mac But how do you get the highlighted text and associated notes to your Mac? The method I'm highlighting requires **a one-off tricky job**, after which you can export the annotations as required and read them on your Mac. 💡 A while after writing this article I was made aware there is a much easier way to this with the help of a browser based tool and I wrote it down: [A Better Way to Export Kobo Notes and Highlights](https://robhoeij.com/a-better-way-to-export-kobo-notes-and-highlights/) ## Exporting annotations Exporting annotations is an option which, don't ask me why, can only be enabled via a custom setting in a configuration file. How does that work on the Mac? You connect the Kobo to the Mac, open the corresponding folder. Then you make hidden files visible because the config file is hidden. Then open the file and add two lines at the end. Save, hide the files again, disconnect, and you're done. Simple, isn't it? I'll take you through it step by step. ## Six steps to enable export Here are the six steps in more detail. ### 1\. Pair up Kobo Connect via cable, press connect and open folder. ### 2\. Make hidden files visible This is a key combination. Make sure that the focus is on the right Kobo window and press: **Shift-Command-.** 💡 The combinations of keys above, ****Shift-Command-.** , need to be pressed together, but you can start from left to right:First press shift, with left hand, keep it pressed.Start pressing Command, with left hand, keep it pressed.Briefly push the key for dot.Done. ### 3\. Locate the configuration file Browse to .kobo and then to Kobo. Here you will find *Kobo eReader.conf*: ![Screenshot showing Filesystem with the Kobo eReader conf file](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/09/2022-09-06_12-50-03.png) Filesystem showing the Kobo eReader conf file ### 4\. Modify the configuration file Open this file with textEdit (or your favourite editor) and add these two lines to the end of the file: ```text [FeatureSettings] ExportHighlights=true ``` Save the file. ### 5\. Make hidden files invisible Just tidy up after yourself and make files invisible again. With the KOBOeReader screen in focus, press: **Shift-Command-.** ### 6\. Unplugging Kobo Eject the Kobo drive on the Mac. ## Export Annotations After making this adjustment, you can go to the book overview on your Kobo and choose export. Or browse to notes via the book. ![Photo of the Kobo e-reader showing the "export annotations" function](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/09/DraggedImage-1.jpeg) Export Annotations on your Kobo Then you will find the text file, after linking, on your Mac for further editing. ![Exported annotations from Kobo e-reader on the Mac](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/09/2022-09-06_12-58-55.png) Exported annotations from Kobo e-reader on the Mac ## One more thing Here is a screencast that shows the right settings on a Mac. The Kobo is attached and connected. Screencast of Kobo settings for export notes and highlights. Exporting highlights and notes from Kobo is complicated and I tried to give simple steps. I hope it worked for you. ### Made to stick, Chip Heath & Dan Heath URL: https://rob-hoeijmakers.ghost.io/made-to-stick-heath-brothers/ Last updated: 2026-06-07T20:34:50.000Z "Made to Stick" is a book written by Chip Heath and Dan Heath that explores why some ideas and messages thrive while others are quickly forgotten. The book provides a framework for creating ideas that are memorable, impactful, and able to generate action. The authors introduce six principles that make ideas "stick" and provide numerous examples and strategies for applying these principles in various contexts. Aspects of how to create sticky ideas and have SUCCESs are build around these six ideas: **Simple** Simplicity isn’t about dumbing down, it’s about prioritizing. (Southwest will be THE low-fare airline.) What’s the core of your message? Can you communicate it with an analogy or high-concept pitch? **Unexpected** To get attention, violate a schema. (The Nordie who ironed a shirt...) To hold attention, use curiosity gaps. (What are Saturn’s rings made of?) Before your message can stick, your audience has to want it. **Concrete** To be concrete, use sensory language. (Think Aesop’s fables.) Paint a mental picture. (“A man on the moon...”) Remember the Velcro theory of memory—try to hook into multiple types of memory. **Credible** Ideas can get credibility from outside (authorities or anti-authorities) or from within, using human-scale statistics or vivid details. Let people “try before they buy.” (Where’s the Beef?) **Emotional** People care about people, not numbers. (Remember Rokia.) Don’t forget the WIIFY (What’s In It For You). But identity appeals can often trump self-interest. (“Don’t Mess With Texas” spoke to Bubba’s identity.) **Stories** Stories drive action through simulation (what to do) and inspiration (the motivation to do it). Think Jared. Springboard stories (See Denning’s World Bank tale) help people see how an existing problem might change. ### My view on this book Throughout the book, the Heath brothers use real-life examples and case studies to illustrate these principles in action. They draw from a wide range of fields, including advertising, politics, education, and social issues, to demonstrate how sticky ideas can be crafted and communicated effectively. Overall, "Made to Stick" is a practical guide that offers insights and techniques for creating ideas that are memorable, persuasive, and able to drive change in various domains. It is a great read, a MUST read! > A credible idea makes people believe. An emotional idea makes people care. The right stories make people act. ### More on Made to Stick On the website from the Heath Brothers you can download the Succes model: And read more about the book: --- ### Further reading - [Made to Stick - Heath Brothers](https://heathbrothers.com/books/made-to-stick/) ### Gated content and paywalls URL: https://rob-hoeijmakers.ghost.io/gated-and-paid-content-an-overview/ Last updated: 2026-04-21T09:20:40.000Z If you are interested in **gated content and paywalls**, you have come to the right place. I explain the differences through daily examples from an internet user. I will take you on a quick tour and first list and explain content gating through paywalls. The soft, metered and hard paywall. And then I will conclude with gated content on so-called lead magnet pages, where personal data is exchanged for gated content. ## Content gating and paywalls: the different types of paid membership There are loads of websites where you are asked to pay for further reading or to take out a digital subscription. You often see this in newspapers, magazines, and blogs. The mechanism to charge for the content is called a paywall. There are several forms of paywalls and I will briefly explain the soft, the metered and the hard paywall. ### Soft paywall With a soft payment wall, it is possible to view content. Typically, a premium model is used where certain articles are free and others require a free registration or paid digital subscription. With a free membership, you view content and participate in the community, for example by commenting. It is advantageous for the publisher because segmentation can take place and readers can be approached via e-mail for offers or new content. Furthermore, non-paying members might see advertisements and paying members do not (freemium). ![Soft paywall, CNBC website. Free news and premium articles. ](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/12/Soft-paywall.jpg) Soft paywall, CNBC website. Free news and premium articles. ### Metered Paywall Other media give you some articles for free. So, for example, you can read five articles a month and if you purchase more, you will see the pay wall. The professional term for this is "metered paywall". This is a widely used form among news publishers and platforms. ![Metered paywall, Washington Post. Four free articles. ](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/12/Wp.jpg) Metered paywall, Washington Post. Four free articles. ### Hard paywall With a hard paywall, access is only granted after payment. Durable hard payment walls are rare and are usually used by publishers who already dominate their market, target a niche audience or add significant value with their content. ![The Wall Street Journal website has a hard pay wall.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/12/hard-paywall.jpg) The Wall Street Journal website has a hard pay wall. ## Gated content and search engines Protecting content by gating naturally has an impact on search engines. If the content is totally protected, it cannot be accessed by search engine crawlers and will not appear in search engines. What often happens, as a solution, is that content is visible to search engines, but that a kind of screen is erected in browsers that is difficult to circumvent. This also means that the screen can be bypassed with special online tools such as 12ft.io. ## Personal data for exclusive, gated content Gated content can be part of a [**lead magnet**](https://rob-hoeijmakers.ghost.io/lead-magnet/)**.** That is a page where you can read preview text and where you exchange personal data to read more. It is a one-time transaction and in return for your information you can get an e-book, extra chapter or access to the full article. After the exchange, you are usually on a mailing list. By the way, the [European privacy law (GDPR) prohibits this method of working](https://rob-hoeijmakers.ghost.io/gdpr-and-lead-magnets/), but not if you also give express permission to be on the mailing list in addition to the offer. ![Bonus chapter after providing personal data. https://bernardmarr.com/bonus/](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/08/2022-08-24_10-32-40.png) Bonus chapter after providing personal data. https://bernardmarr.com/bonus/ 📱 NFT gated content. A new development in "gated content" is that purchase and access are arranged through NFTs. ### The secrets behind the QR Code URL: https://rob-hoeijmakers.ghost.io/qr-code-explained/ Last updated: 2026-04-25T13:16:03.000Z Until recently, you had to install a specific programme to read the QR code from the QR tag. But after Apple also embraced it and built the reader into the camera's software, the QR code became an object for daily use. ## What is the QR code? However, the QR code also has a bad ring to it, also because of Corona. Because of its secrecy and unfamiliarity, for many people, it came to symbolize the all-pervading digital control that would permeate our society. So, it seemed like a good idea, even if it is a bit late, to take a closer look at the QR code and explain it. The [QR code](https://www.adobe.com/express/feature/image/qr-code-generator) is **an optical label for exchanging information readable by machines**. Just like the better known barcode. The blocks in the QR code are a way of recording information. Information about the QR code itself and information that will be exchanged: ![The technical secrets behind the QR-tag](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/12/QR_Code_Structure_Example_3.png) The technical secrets behind the QR-tag ## The QR code in a daily use How does it work again? For the user, it is actually simple. Point your camera at the QR-code (on screen or paper) and an option will appear what to do with the information the QR-code contains. This could be an internet address, for example, and the phone will open a web page. Or you may have a QR code on your phone to display, and it will be read by another party. Usually, the code will contain unique information that is then linked to a central file. In both cases, the device will recognize and read the QR code. ![The technical secrets behind the QR-tag](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/12/IMG_3028.jpg) QR code recognition on an iPhone ## Contents of QR code Via a QR code e-mail addresses, contact information (vCards), calendar appointments, GEO locations, text messages, telephone numbers and texts (maximum 7089 characters) can be offered. This is also the case with [NFC codes](https://rob-hoeijmakers.ghost.io/an-explanation-and-demo-of-app-clips/). These are small transmitters that are activated by the receiver. They are used, for example, when you use your phone to make payments at the cash register. 🔒 QR code can be a security risk if they are changed and pointing you int the wrong direction. Check out this public service announcement by the FBI for more information and helpful tips: [Cybercriminals Tampering with QR codes to Steal Victim Funds.](https://www.ic3.gov/Media/Y2022/PSA220118) ## Versions of the QR code There are several versions of the QR code in circulation. This is good to know because it explains the difference. The higher versions often contain more information and are therefore also larger and have a finer resolution. This also makes them somewhat slower and more error-sensitive when read. ![Higher version of QR code](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/12/lukas-qS7fr07bljM-unsplash.jpg) Higher version of QR code ## History The QR code was invented in 1994 by the Japanese company Denso Wave. QR stands for **Quick Response**. The Denso-Wave company decided not to exercise its patent rights, and this means that QR-codes can be made free by everyone. The inventor was Masahiro Hara and the board game Go gave him the idea that information can be displayed in two directions rather than left to right like the barcode. ![The game of Go as the inspiration for the QR code.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/12/elena-popova-xdXxY5C9PUo-unsplash.jpg) The game of Go as the inspiration for the QR code. ## Ethics of QR codes and tags The QR code is actually a tool with which information can be exchanged. But because applications use it to link to systems and data, there is an understandable sense of concern surrounding the application. This concern was much less with the corresponding barcode, for example. If you use a QR code at a restaurant, it refers you to a web page or an application for ordering. Later, you pay for this order and the order can be traced back to your private data. People also don't like having to carry a sign. There are very negative associations with the past where signs, on passports or clothing, were used to make a system of exclusion efficient. For me, knowing the workings of these tags, makes it more transparent and less suspicious. The tag is a token, it contains readable information that can be connected by the receiver to bigger data sources. For good or for bad. I hope you find the QR code less secretive after reading this article. *Apple App Clips uses both QR code and NFC tags to start up mobile workflows:* [*An explanation and demo of App Clips*](https://rob-hoeijmakers.ghost.io/an-explanation-and-demo-of-app-clips/?ss360SearchTerm=qr)*. I also wrote about* [*RFID*](https://rob-hoeijmakers.ghost.io/the-paper-rfid/)*.* Great Veritasium video on the QR Code. Why not build on by hand? 📶 QR codes for Wi-Fi encode network details using a standard format (WIFI:T:;S:;P:;;), enabling automatic connection without manual input. This method is compatible with both iPhone and Android, providing a universal solution for effortless Wi-Fi access. Scanning the QR code automatically interprets network type, SSID, and password, streamlining the process of joining Wi-Fi networks. ### Get copyright for web content URL: https://rob-hoeijmakers.ghost.io/get-copyright-for-web-content/ Last updated: 2026-05-10T08:54:20.000Z When you publish on the web, you soon come up against copyright. I will first answer the most common questions and then explain the most common forms of copyright. ## What is copyright? Copyright is a form of intellectual property that protects original works of authorship once an author records the work in a tangible form of expression. In copyright law, there are many types of works, including paintings, photographs, illustrations, musical compositions, sound recordings, computer programs, books, poems, blog posts, films, architectural works, plays, and much more! ## Is copyright the same everywhere? Today, national copyright laws are standardized to a certain extent by international and regional agreements such as the Berne Convention and the European Copyright Guidelines. The copyright laws of different countries have similarities, but each jurisdiction has its laws and regulations. Some jurisdictions also recognize moral rights of creators, such as the right to be credited for the work. ## Are all images copyrighted? All images on the web are copyrighted. ## What happens if I violate someone's copyright? If the infringement is noticed, you may receive a notice ordering you to cease and desist from any infringing activity. Every so often that is enough, but it can also lead to a lawsuit. This web page explains it in detail: ## Do the same copyright rules apply to social media? Yes, actually, they do. But it does not apply to sharing photos within a platform such as Twitter, Instagram or Facebook because you have given permission for that. Here is handy scheme: ![Scheme to check if you can use an image on Social Media.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/08/Welke-beelden-mag-je-gebruiken-voor-je-social.png) Scheme to check if you can use an image on Social Media. ## How do I find out who owns the copyright to a work I want to use? Typically it is mentioned, but otherwise you will have to search in the database or ask. If in doubt, or if you can't find a copyright holder, it's best to leave it at that. ## Three most common forms of copyright After these frequently asked questions, I'll briefly explain the three most common forms of copyright. It's not exhaustive, but as they are the most widely used, it should go some way to explaining them. ### Creative Commons (CC0) The use of a work under the CC0 licence has no restrictions for personal or commercial use. Basically, if a work is classified as CC0, you can post it, modify it, add to it, share it, and do whatever you want without having to mention the author's name or ask permission. It's as free and open as can be. ![Creative commons logo. Photo by Umberto](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/08/umberto-lhJrm1BRVV0-unsplash-2048x1365.jpeg) **Creative commons logo. Photo by* [**Umberto*](https://unsplash.com/@umby?utm%5Fsource=unsplash&utm%5Fmedium=referral&utm%5Fcontent=creditCopyText) [Creative Commons — CC0 1.0 Universal![](https://creativecommons.org/favicon.ico)cc.logo.white![](https://creativecommons.org/wp-content/themes/cc/images/cc.logo.white.svg)](https://creativecommons.org/publicdomain/zero/1.0/deed) ### Royalty-free Royalty-free (RF) material that is copyrighted or otherwise protected by intellectual property rights may be used without the need to pay royalties or license fees for: - any use - per copy sold - volume - a specified period of use or sale ### Rights managed Rights Managed, or RM for short, is a type of licence for the use of stock images, which can be purchased from a stock photography agency, such as Getty Images. This type of licence is sold on a one-off basis. It allows the buyer to use the photo in a certain specific way, and only in that way. This type of right can pay off if it is an image that is very decisive in your marketing. ## More guidance on copyrighted images The guide from Pixsy goes into more depth and also explains Fair Use, the Public Domain and copyright as an employee. --- ### Further reading - [A Beginner’s Guide To Using Copyrighted Images - Pixsy](https://www.pixsy.com/academy/image-user/using-copyrighted-images/) ### Review: The Founders by Jimmy Soni URL: https://rob-hoeijmakers.ghost.io/review-the-founders-jimmy-soni/ Last updated: 2026-05-10T08:54:20.000Z I read the book The Founders in a series of other books that deal with the emergence of large Internet companies and the associated investments. These stories give a nice picture of a time, it gives background to my own life as a knowledge worker and inspiration for my current work as a MarTech specialist and Content Producer. I must confess that it is also partly out of nostalgia and curiosity for individual protagonists. In the case of The Founders, the company is Paypal and the founders, from two different companies, included Elon Musk and Peter Thiel. Now, in my own country, the Netherlands, Paypal is not a very important brand but in many countries it was an indispensable catalyst for e-commerce. I actually finished reading The Founders in one go and was impressed by how the author managed to create a story that actually takes place at the office. The financial world is abstract and not really tangible. And how do you tell a story about something that, according to internet years, took place centuries ago, where a factual account surely counts, and you don't want to lower yourself to gossip and backbiting? ## The story The author tells the story chronologically with some smaller side trips, such as a visit to Elon Musk or in depth. The picture of the company begins to emerge, but what I like about it is that it is not immediately clear, or could have been, where exactly it is going. There is movement, but not immediately a linear direction. That sense of possibilities becoming so numerous that it tends towards chaos gives a good insight into the spirit of the times. Don't stand still, keep moving, personal styles and forces that continue to drive the movement. Not immediately by being right, but in any case by thinking big and operating on the edge of megalomania. The bigger, the better. > PayPal isn't just about people banding together to shape a product-it's about how banding together shaped the people themselves. ![The Founders, Jimmy Soni.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/08/IMG_0066.jpeg) The Founders, Jimmy Soni. ## About writing How do you portray a story that actually takes place in boring offices? The author handles this very well, the tension curve remains tense. What I find very clever is that he does this without abandoning the facts. There must have been an awful lot of meticulous research in this book. And Jimmy Soni is a very gifted and clever writer. The story requires a lot of concentration, and you can tell. When the tension is off and the complex story is told, he lets go completely and freely and tells the story about two prisoners who see the PayPal Mafia as an inspiration. Then you notice the great potential of his ability, and I at least understand better how much power and energy (five years of work) the book must have cost. Besides the story about PayPal, I enjoyed the backstory of the making of this book. In it, Soni tells us candidly about the struggle that writing this difficult book cost him. It gives an insight into how you write a book, but above all how you create it. How and with whom. The surrounding professionals that you need to write a book like this. The Founders is cleverly written, profound and exciting. For people who work professionally with products and teams, are starting a business, or are interested in the history of the Internet or financial services. And for people who want to know more about what drives entrepreneurs like Musk and Thiel. I definitely recommend the book. [The FoundersThe story of PayPal and the entrepreneurs who shaped Silicon Valley![](https://jimmysoni.com/assets/icons/apple-touch-icon.png?_cchid=4bf994cf86b1ac71bb460001e8fd3354)Jimmy SoniJimmy Soni![](https://jimmysoni.com/assets/images/books/the-founders-thumbnail.jpg)](https://jimmysoni.com/books/the-founders/) P.s. I am eager to read Jimmy Soni's biography on Claude Shannon: --- ### Further reading - [A Mind at Play](https://jimmysoni.com/books/a-mind-at-play/) ### Fix Kobo Clara crashes and freezes URL: https://rob-hoeijmakers.ghost.io/kobo-clara-crashes-and-freezes/ Last updated: 2026-04-14T13:20:43.000Z My e-reader has become more and more important, and I want to make sure I can rely on it to work. Sometimes, my Kobo Clara e-reader freezes. Sometimes while reading, often when turning it on and off or in sleep mode. The book you stopped reading yesterday is still on the display, but when you press the on button, only the backlight of the reader goes on. What to do now to make the Kobo wake up? How do I solve a frozen Kobo Clara? ## Put the Kobo on the charger The most important tip I can give you here is to **connect the Kobo to the power after it freezes**. ![Connect your Kobo Clara before restart](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/08/image-1.jpeg) Connect your Kobo Clara before restart Once the Kobo Clara is attached to the charger, **press and hold the power button (on button) for about 15 seconds**, and it will shut down. Once the device has shut down you can start it up again. Please note that this is a reboot and may take a while. ## Possible causes of crashes Most possible cause of a crash is the battery not loaded properly, for instance through a power bank. Sometime the system software might be out of date. ### Battery I have noticed that when the Kobo Clara e-reader is frozen, there is often something wrong with the battery. Poorly charged, for instance with a backup battery. Also, a different charge load is indicated than the one indicated later, after startup. > Try to keep the Kobo charged. ### System software Another possible cause is outdated system software. It is good to keep this up to date. > Keep system software up to date --- I also wrote posts about [how to share ebooks with your partner](https://rob-hoeijmakers.ghost.io/share-kobo-ebooks-with-your-partner/) and [how to sideload your ebooks on a Kobo e-reader using an iPhone](https://rob-hoeijmakers.ghost.io/load-free-ebooks-on-a-kobo-using-the-iphone/). Ideal for holidays. Talking about holidays, check out how I work with a [personal MiFi hotspot](https://rob-hoeijmakers.ghost.io/a-mifi-router-for-my-mountain-holidays/) that also works with Kobo e-reader. *Thanks for reading, I hope it solved your problem. Cheers,* [*Rob Hoeijmakers, Digital Strategist*](https://rob-hoeijmakers.ghost.io/about-rob-hoeijmakers/)*.* ### Playing in the interactive sandbox URL: https://rob-hoeijmakers.ghost.io/playing-in-the-interactive-sandbox/ Last updated: 2026-04-25T13:16:42.000Z It is summer, and we are taking it easy for a while. But I saw an interactive sandbox on LinkedIn and after some enquiries it turned out, there was such a sandbox at in a geographic theme park. I didn't just want to see it, I wanted to experience it. ## What is an interactive sandbox? The more formal name is Augmented REality Sandbox (ARES, also the god of war) and it came out of a project by the American army. As an improvement on the sand tables that have been used for centuries to simulate the battlefield. The sand on the sand table is enriched with a projection, for example with contour lines or relief colours. As the sand moves and is moved, a depth sensor (Kinect) and a computer adjust the projection to the changes. ![Sandbox, beamer, depth sensor and a computer. ](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/07/3CEE5D47-C05B-44E7-B05E-B46555384764_1_105_c.jpeg) Sandbox, beamer, depth sensor and a computer. On request, for example by holding your hand above it, you can make it rain or produce lava. You can then make that liquid or mass flow through the sand/landscape. 0:00 /0:33 1× Interactive sandbox with erupting volcano. ## Making things digitally tangible These interactive sand tables are fun to play with. In the fortress, where there was one, children, and adults alike crowded around it to play. This digitally enhanced table also plays a role in education. It requires no manual, no explanation, you can start working with it immediately. It is extremely intuitive, and you can get to work digitally, interactively, without even touching a computer or device. Personally, I was fascinated by the interaction. Not swiping across a screen, moving a mouse or tapping a keyboard, but just a table with sand as input. And unlike movements, there is a direct contact with a matter, with the sand. I had actually never experienced this before, and it had an almost magical effect and was more than worth the visit to the fortress. I recommend playing with it or taking the children to a place where they have such a device. ![Making digital tangible.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/07/C2F6E505-242E-4D2F-9305-A96B052D2BA6_1_105_c.jpeg) Making digital tangible. ### Four professional writer roles for the web URL: https://rob-hoeijmakers.ghost.io/professional-writer-roles-for-the-web/ Last updated: 2026-04-21T17:24:01.000Z Writing is writing, you might say. To be honest, I thought so too. Or should I say, I didn't know better. But refining this idea makes writing easier. So, I delved into the different roles and functions that exist for writing for the web. All the writing does produce text, but the process of creating it and placing it correctly in a digital product like Website or App is extremely diverse. Moreover, a text serves different purposes. In the Internet industry, several functions have thus arisen that deal with writing specific texts and giving and keeping an overview of the texts as a whole, a bit like an editor function. We can define four functions, which in smaller teams are also called roles. For roles, too, it helps to empathize with the expectations and the right texts as a writer and to give direction to the process. 1. Copywriter 2. Content writer 3. UX-writer/Content designer 4. Technical writer I briefly outline the roles. ## Copywriter It is a profession older than the web. A copywriter writes short texts for advertisements and ads. The texts are aimed at **increasing the awareness** of a brand and to let the reader have certain thoughts or perform a specific action. For writing on the web, a copywriter will often work within a digital marketing agency. The work of a copywriter **increasingly overlaps with that of a content writer** because the copywriter will also write social media ads, Google Ads, online landing pages and compelling email texts. This new wave of copywriting, born out of the digital age, has made the profession more accessible. But not without a downside, as globalization has devalued some copywriting jobs, as it is so easy to find skilled copywriters working at different rates. ## Content writer The term content writing has become more common over the years, as traditional forms of advertising have been increasingly relegated to the background. Companies from all sectors learned that success in the digital age requires more than traditional, outbound marketing strategies. Content writing allows companies to build an **emotional connection with their target audience**, a prerequisite for customer loyalty and brand engagement. Content writing often forms the basis of any **digital marketing strategy**, as emotional content is necessary for everything from blogging and SEO strategies to email and social media promotion. ![Alpine flower in autumn to lighten up the heavy inventory ;-)](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/09/85A90710-13DF-4D0C-B2C5-7830BC45FCF6_1_105_c.jpeg) Alpine flower in autumn to lighten up the heavy inventory ;-) 💬 Check out the [difference between a Copywriter and a Content writer](https://rob-hoeijmakers.ghost.io/difference-between-copywriters-and-content-writers/). ## UX-writer / content designer A **UX Writer** at heart is of course a writer, but also someone who helps form content and applications into a whole, a story. This requires more than the ability to write, it is also a function comparable to that of a **director**. The UX-writer works with website and app content, not only on a page by page basis, but especially also on more complete, **scenario based User Journeys**. For courses, community, membership, handling sales, etc. But also directing marketing and sales related content. A UX writer finds the words for menus, definitions, buttons, labels, chatbots, and error messages. The small pieces of writing, which are collectively called “[microcopy](https://rob-hoeijmakers.ghost.io/micro-content-big-impact/).” **Content designer** works with text-oriented websites, often aimed at sharing larger forms of information. Think of local and national government websites. An **UX-writer** also deals with more interactive experiences such as user flows on, for example, mobile applications. Most UX writing is done after a user has completed a purchase or signed up for a service or app. As a result, there is no need to try to sell something. Instead, the writer's goal is to make sure it is easy and enjoyable to use the product or service. Other titles you may come across: Content strategist, UX copywriter, product writer, UX content designer, and content editor jobs ## Technical writer One difference between [technical writing](https://rob-hoeijmakers.ghost.io/the-crucial-role-of-technical-writing-in-composable-platforms/) and UX writing is that technical writers typically work with **text-rich documentation** (online help files, instruction manuals, etc.) rather than user journeys in digital interfaces. As such, the two roles require quite unfamiliar ways of thinking. In today's digital products, there are only traces of typical technical writing - for example tooltips and flows that guide the user. The goal of many digital products is to make the User Interface so frictionless and user-friendly that supporting documentation is no longer necessary. And this is precisely one of the reasons why many technical writers are moving into UX writing and content design. In general, however, technical writing is aimed at informing and instructing the reader rather than convincing. *If you are interested in writing for the web, check out my articles on the [difference between a copywriter and a content writer](https://rob-hoeijmakers.ghost.io/difference-between-copywriters-and-content-writers/) and the question if [ChatGPT will replace Content Writers](https://rob-hoeijmakers.ghost.io/chatgpt-content-writer/).* ### Content creation is an art, marketing technology is a craft. URL: https://rob-hoeijmakers.ghost.io/marketing-technology-martech/ Last updated: 2026-04-24T18:30:40.000Z For a while now, I've been looking for a good description of what I do. What I am good at. I got a good clue when I [read about Von Humboldt](https://www.andreawulf.com/about-the-adventures-of-alexander-von-humboldt.html), the biologist and explorer. Also called the inventor of nature. A great man who linked scientifically rigid measurements and loads of data to a humanistic and romantic image of nature. He connected facts with a story. ![Chimborazo map by Alexander von Humboldt. Combining loads of data with a visual story. ](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/06/Humboldt-Tableau_resized-for-new-cms.jpg) Chimborazo map by Alexander von Humboldt. Combining loads of data with a visual story. ## The story as a drive for actions And it was precisely the story that had fascinated me for some time. I practise writing and immerse myself in the art of storytelling. I read about plots, look for examples in the profession and practise copywriting by publishing on Twitter. And yet, the story is nothing without a medium, a soil on which it can grow. ## The screen showing us magic Almost everything we see and read comes to us through a screen. And behind that screen, magical things happen. In the making, in the transmission and in the display. Digital magic behind the scenes. And this requires technology. A lot of technology in the form of software that processes data. Exactly that is my other fascination in my (professional) life. That of technology. ## Marketing technology (MarTech) This technology is partly about producing content, but for an equal part about marketing this content (products, services, events). In the form of digital campaigns. To convincingly get a message to the right audience, you need magic. Just like a magician or an artist, this only works if it is done quickly, flawlessly and with an eye for detail. Only then will the illusion work and people will believe in what you are telling them or showing them. So, I link Content Production for the web to **Marketing Technology (MarTech)**. Art to technology. This is how I have the most wonderful profession and business that I can imagine! ## Content of my blog After this insight, I immediately changed my blog and biography. It is no longer just about Web Content, but also about Marketing Technology. I'm confident that with this change, I can write even more focused articles that help you make the world on and behind our screens digitally tangible. I also created a tag for MarTech and found that I had many articles that fit better with MarTech than with Content. A perfect adjustment. ## Business aspect of MarTech It is also important to note that clients often see content as a cost item and marketing (because it increases sales) more as an investment. It is also true that the supply of content production is greater, especially if quality is irrelevant. This does not apply to Technical Marketing because, due to complexities, it involves much more specific skills and experience in this area therefore counts much more. ![MarTech Map](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/12/IMG_6958.jpeg) [MarTech Map](https://martechmap.com) ### Why choose Wix to manage your valuable content? URL: https://rob-hoeijmakers.ghost.io/why-choose-wix-to-manage-your-valuable-content/ Last updated: 2026-04-24T18:50:36.000Z We searched a long time for a good Website Builder. It's a serious connection you make by doing all your content, work, optimizations within a system that isn't yours. A website builder, with all your valuable content, can move in directions that are not conducive to your content and business. It can become slow, corrupted, or gather dust in a corner. It may not keep up with the times, or you may be financially milked because you are locked in. Or it may start to squeak in parts, such as payments, or they may take away a local payment system that is important to me. So there are a lot of bears on the road. We were also looking for a system that could cope with a certain growth of the company so that we would not run into any boundaries too quickly. A system that feels more as a partner than a provider. Another important aspect was search engine optimization, because we rely heavily on content as a form of marketing. I tell you how and why we ended up with **Wix**. ## Hobby As a hobby, my wife makes [artistic Christmas balls](https://aleide.nl), and we were looking for e-commerce software. At the end of November, we chose Wix and started to work like crazy. Within a few days, the shop was already taking shape, and we took the plunge by subscribing to Wix for a year. Everything we needed was there and it worked intuitively. Taking pictures of the products, making up names, writing texts, making up a logo, coming up with a domain and working on a template and style. Optimizing the content by adding meta-information, titles, descriptions etcetera. It was a steep learning curve, because Wix doesn't work in the same way as Squarespace, WordPress or Shopify. Especially, designing in the editor is so powerful that it reminded me of those very expensive sports cars that are driven into the crash barrier by inexperienced drivers. But the shop was ready on time, payments came in, and we soon learned that the postal handling could be linked to partners (Sendcloud). We also quickly succeeded in creating a link to Instagram and setting up a shop there. The website soon turned out to be well optimized and the SEO talk lists helped us. ## Our company For our company, a [full-service web content agency](https://www.schmuki.nl), we were working on an architecture that consisted of multiple systems. The best of what was available (and not the cheapest either). The idea was to offer online courses, and we looked at Kajabi. And because Kajabi doesn't support the local payment system (iDeal) we needed a separate payment system. To connect Kajabi and payment system, we used Zapier. For managing clients Hubspot and emailing leads and clients ConvertKit. A website we build in Squarespace and later in Leadpages. And a blog in Ghost. That was already adding up, becoming complex and hindering us in quickly executing new plans. But we still saw this as the best way to build scale. And online courses are a discipline in themselves. ## Change of course We started our own company in June 2021 and the idea had yet to crystallize, but the solution we had in place proved complex and expensive. So we kept looking and changing. We actually wanted to bundle our blogs and share the courses and workshops at the lowest possible price or even better for free. But then the fixed costs would have to come down. The content is a flywheel, we do what we preach about, we learn by writing, and we get the audience and the attention we seek. Giving it away is a much more stimulating set-up. And then we discovered the possibility in Wix that we didn't know yet: **online programs**. That turned out to be a perfect fit with the idea of pouring in online courses and making use of the powerful member application. Creating and managing a user account is a strong part of Wix. So through those accounts we could welcome visitors and offer the courses. For free or for a fee. And the other good news was that the **blog of Wix**, a well-thought-out and powerful solution, also offers the possibility to give multiple authors a stage next to each other. We had been looking for this for some time, but it turned out not to be that easy. ## The solution was Wix So now we had: - A blog for multiple authors - An environment to build a website - A system for members - Payment methods including iDeal and Apple Pay And all these features turned out to be well suited for search engine optimization, an important source of traffic. As icing on the cake, there also turned out to be a complete **Site Search in Wix**. We took a two-year contract and paid in advance to keep costs down. We cancelled or scaled back all other systems. And then a few weeks, we worked on a digital move. ## Strengths of Wix We have been working with content and content management systems for over 25 years. We have also developed them ourselves for large companies, and at some point you recognize a well-made product. Just like with brands like Apple or BMW. You notice that there is a solid plan, a philosophy behind it and even more important that they have thought about the details. Not just functionally, but to make things pleasant to work with. > "In most people's vocabularies, design means veneer," Jobs told Fortune shortly after retaking the reins at Apple. > "But to me, nothing could be further from the meaning of design. Design is the fundamental soul of a man-made creation that ends up expressing itself in successive outer layers." So the strong points in our view are: - Affordable - Versatile and generous - App for creators - App for customers/members - Ample possibilities for search engine optimization - Services to set up - Possibility for subscriptions - Excellent helpdesk - Integrated stock photos ## Weaknesses of Wix Minor issue, but annoying when you use it a lot: - Complex and sometimes slow editor - Only five forms and two automations ## Wix specialists I have gained so much experience with Wix that we have become specialists as well and therefore also offer Wix as a solution for my customers, next to Squarespace and Ghost. [Feel free to reach out to me if you want to learn more.](https://rob-hoeijmakers.ghost.io/contact/) ### What a silk map learned me about content URL: https://rob-hoeijmakers.ghost.io/what-is-content/ Last updated: 2026-04-25T13:17:12.000Z I have been playing with the concept of content for a while now. About 30 years ago, I started my work in the digital field as a webmaster. And throughout my career, I have been working with content. By writing, photographing, creating websites, developing interactive applications, using management systems, doing search engine optimization. But it is such a fluid concept, a fleeting concept, much like the word energy. ## Content The word literally means something has content. That which is contained. But as soon as you do something with content, arrange it, adjust the medium, you just created some new content. Digitally, this can be done quickly, easily and in large quantities. A film is content, but consists of separate parts which are also content. Like the music, for example. If you play live music with a film, new content is created again. By linking content through references, a new product is created. An article on Wikipedia is content, but Wikipedia is also content. An image is content, that image in a webpage is content and that webpage is a website and that is content. In short, the whole internet is content. ## Content sets the mind in motion Let me just take a sober and practical look at it and not make it too philosophical. Content is an abstraction and, nowadays, also data that can be brought to life by an instrument (device) and transformed through the senses. Content can be used to learn (information), to enjoy or, for example, to distribute and share. In this sense, content is a potential, something that can **set the mind in motion**. Like an energy. ## The silk map This somewhat romantic narrative to content helps me not to get bogged down in a complex story with categories. If content can be seen as potential energy, then I can look at it more creatively. That's precisely the insight I got when I read about a Dutch woman who was imprisoned in a prison camp. But after the war, to go back to Europe, she made a beautiful dress for herself with needle and thread. From silk maps. ![Dress made from Silk Maps. Rijksmuseum.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/06/NG-2000-5.jpg) Dress made from Silk Maps. Rijksmuseum. At first, I did not understand it and had to ask around a bit. The story was that during the Second World War, Allied airmen wore a silk scarf. On that scarf was a map of the area over which they were flying. If they crashed, they had a lightweight, waterproof, non-cracking map of the area. This conversion of content from a paper map, to a silk scarf, to a dress made me happy and gave me an insight. This dress is an excellent anecdote to explain the special nature of content. It is what you do with content that makes it what it is. And a map is a map when used to find the way. On parchment, paper or a silk cloth. But as a garment it is a pattern, a decoration. A map without an area. ![Another example. A map on a cushion in a boat.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/07/IMG_8303.jpg) Another example. A map on a cushion in a boat. ### Get good images for your website URL: https://rob-hoeijmakers.ghost.io/three-ways-to-enrich-your-website-with-top-photos/ Last updated: 2026-04-14T14:21:00.000Z I manage three websites of my own. This blog, my business site and an e-commerce site. I pay a lot of attention to substantive, independent stories about creating, producing and publishing good content. I start with text and then find striking and appealing images to go with it. Where can I find suitable images for a website, landing page or social media campaign? So captivating, original and affordable images for a project of your own. How to move forward? I see three obvious, not mutually exclusive ways to get pictures and cutting and pasting just any old material from Google is not one of them 😉 I'll take you on a quick tour of the possibilities: - stock photos, free or paid - your own photos and - commissioned photos. ## Stock photos The most common way is to use stock photo’s. The word stock means a pile, a storage and therefore a collection. This is a service with many photos that can be searched by keywords or arranged by theme. Many people are on a budget when it comes to photos, so I'll make a distinction between the free service and the paid service. ### Free stock photos The services that work with a CC0 licence are very much on the rise. This is a commons licence that gives you the right to use the images online without much obligation. You may edit and crop images for various purposes. An acknowledgement of authorship is not obligatory, but is appreciated. Recently, a number of photo stocks have started offering high-quality photos under a free and open licence (CC0). You may use and edit these photos online. They are free, but a monetary contribution or honorary reference is appreciated. I use these photos regularly and especially for temporary landing pages where for example SEO on the images is less important. In other words, if I need to make a page quickly that will not be online for long. Here is a brief overview of services that offer free stock photos: - [Unsplash](https://unsplash.com) - [Pexels](https://www.pexels.com) - [Pixabay](https://pixabay.com) An example photo on photography: ![Photo by Matt Hardy: https://www.pexels.com/photo/woman-using-slr-camera-2179205/](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/06/pexels-matt-hardy-2179205.jpg "Photo by Matt Hardy: https://www.pexels.com/photo/woman-using-slr-camera-2179205/") Photo by Matt Hardy: https://www.pexels.com/photo/woman-using-slr-camera-2179205/ ### Paid stock photos If you want more original or better executed photos on your website or campaign, there are paid services. In essence, there are two types of licence. One gives you the right to the photo in return for payment (Royalty-free) and the other gives you the exclusive right (Rights Managed). This prevents you from seeing the same photo on a similar website or even in the competitor's website. If you have been working with stock photos for a while, you will soon notice the difference in quality. In the strength of the service, but also in the photos themselves. Both technically (lighting, processing, resolution) and thematically they are more elaborate. Here is a short overview of examples: - [iStockphoto](https://www.istockphoto.com) - [Adobe stock](https://stock.adobe.com) - [Shutterstock](https://www.shutterstock.com/) - [Getty images](https://www.gettyimages.com) ### The photo stock built into web builder It depends of course on which tools you use to build your website, but web builders like [Squarespace](https://rob-hoeijmakers.ghost.io/squarespace/) and [Wix](https://rob-hoeijmakers.ghost.io/why-choose-wix-to-manage-your-valuable-content/) have one or more stocks built in. That makes finding and processing the images straightforward. Wix, for example, has its own stock, there is access to Unsplash and there is a premium service where you pay a fee per image. ![On the left, the three stock photo options in Wix. ](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/07/2022-06-03_13-07-40.jpg) On the left, the three stock photo options in Wix. ## Your own pictures A second option is to use your own photos. The advantage is that these photos can be more original and more to the point. In my case, I also just like to use them, it gives me a stage. The disadvantage is that they sometimes look a little haphazard, but I accept that. And I do my best to make the connection clear with a playful photo caption. I wrote an article about how to make photos suitable for the web, and make sure to [shoot your images in portrait AND landscape mode](https://rob-hoeijmakers.ghost.io/portrait-and-landscape/). It is worth noting that as soon as you put your own photos online, you retain the copyright. Even without mentioning copyright, although it can't hurt to mention it. ![Bridge in Bilbao at sunset. From my own stock of photos. ⓒ Rob Hoeijmakers ](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/06/72D8569B-0B91-4DA3-BE37-669EF81F7A34_1_105_c-1.jpeg "Bridge in Bilbao at sunset. From my own stock of photos. ⓒ Rob Hoeijmakers ") Bridge in Bilbao at sunset. From my own stock of photos. ⓒ Rob Hoeijmakers ## Commissioned photographs Commissioned photography is another story. We recently had this done by two different photographers, and the methods and results were very different. As with many things in life, cheap is often expensive. A good photographer thinks of much more than photography, but will empathize with you or your product. There will be a consultation and list of attributes that might be needed for content-rich photos. Especially if the photos are shot on location, a lot of attention will be paid to lighting and there will be requirements for the environment in which the photos are taken. The quality photographer will offer you a clear price and agreement. In our case, not all the photos she took were actually ours. She made a series available in low resolution, and we could choose from them. ## To conclude There are many ways to get photos, and it will take some trial and error to find what suits your goals. Once you know where you can find photos and at what price, it is a nice challenge to find the right photo. ### Plagiat or copyright infringement URL: https://rob-hoeijmakers.ghost.io/plagiarism-or-copyright-infringement/ Last updated: 2026-04-25T13:16:35.000Z The other day I was stupid enough to take a glossary from a book and publish it as supporting material to an article on my blog. I had started adapting and adding to it to make it my own work, but I got a slap on the wrist from the copyright holder and deleted the content and apologized. And I looked deeper into plagiarism and copyright again. ## Difference of plagiat and copyright infringement The difference between plagiat and copyright infringement is that with plagiat, you pretend to be the author, and with [copyright infringement](https://rob-hoeijmakers.ghost.io/get-copyright-for-web-content/), you unjustifiably multiply the work of others. If I plagiarize I am an imposter, if I infringe on copyright I am a thief. So, when I wrote the glossary, where I did not state the author or the source, I was committing plagiarism and I infringed copyright. If I had mentioned the name and source, it would only have been a copyright infringement. Suppose I publish an old work in the public domain (free of copyright) under my own name, then I am committing plagiat, but I am not infringing copyright. Then it is not punishable, but morally reprehensible. ## Fair use Was I copying to steal? No, that was not my intention. I wanted to use the glossary, adapt it, expand it and use it for teaching purposes. My website is not a commercial work, it is a hobby and I do not earn anything from it. I didn't consider the glossary to be my own work, but as a common thing. After all, how else are you supposed to start a list? I thought it would be a fair custom. I went to find out more about fair use (sort of like the right to quote). But fair use is an elastic concept, and when in doubt it can be taken to a judge. The law requires that the legality of the use can be compared to the following four guidelines: 1. The purpose and character of the use, including whether the use is commercial or educational and not-for-profit 2. The nature of the protected work 3. The scope and importance of the copied portion relating to the protected work as a whole 4. The effect of the use on the (potential) market for or value of the protected work Compared to my article, the word list was long enough to be unbalanced and, on reflection, it would not be fair use. I could have thought of that too. ## How much change does a text need? When studying plagiarism and copyright, I was reminded of Theseus' ship. This is a thought experiment. In time, all parts of the ship are replaced by new ones and at a certain point, no part is original. The question then arises: is the ship still original? How many parts of a boat can you replace until the old ones no longer exist? ![Old or new 2CV? This car reminds me of Theseus Ship.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/06/7B9E98F3-CF9E-47D1-9283-AF46FFE76C98_1_105_c.jpeg) Old or new 2CV? This car reminds me of Theseus Ship. And how much text do you have to change or replace before it is no longer plagiarism or the copyright does not rule any more? That question is even more important now that smart algorithms can translate and re-translate. I even found a service that specializes in this: https://rewriteguru.com On that site, you can check texts for plagiarism, but also have them rewritten using algorithms. It's a bit of an awkward feeling because it may be legally permissible, but it still doesn't feel entirely right. As Warren Buffett tends to say: > I don't care if the law allows it. Can you explain it to your mother? Warren Buffett To a certain extent, paraphrasing and borrowing will be inevitable. Every so often it is honourable when your texts prove to be a form of inspiration, but occasionally, it will feel like stealing, especially if the texts are used where people are fishing in the same pond. ### Review: The Cryptopians by Laura Shin URL: https://rob-hoeijmakers.ghost.io/review-the-cryptopians/ Last updated: 2026-05-14T21:02:16.000Z In clear, non-technical language, Laura Shin takes you in her book **The Cryptopians** through the history of the legendary Ethereum blockchain. She describes the emergence of Ethereum chronologically, goes into the main players, the social dynamics and the milestones. ## Crypto Utopia A new world with new ways of doing business. No longer centrally managed and misled, that was the promise of Ethereum. Not only can you place a coin or token on it, but also a (financial) contract in the form of non-modifiable code (dApp). It has the power of a utopian vision, and the stories around the Mother Ship in Switzerland indeed reminded me of other utopias and communities such as Monte Verità in Ascona and Esalen in California. How they arose from idealism and fell apart due to greed and lies. Good title and subtitle for the book about the organization around Ethereum. ## Vitalek Buterin, creator of Ethereum Much also becomes clear about the great man behind Ethereum, its creator and director Vitalek Buterin. A driven and intelligent idealist, who was naive too much of the haggling around him. What it is like to sit in the eye of a hurricane and read how young he really is. Occasionally, he struggles to decide or to attach consequences of certain behaviour of others. It also became clear to me how important a central figure is around an idea that is still new and fragile. That the survival of an idea is largely due to the personal care of the inventor. > VITALIK BUTERIN WAS like a seed that had been blown far out of its habitat and struggled to take root in its new environment but was about to find the right soil in which to grow tall and flourish. Commonly described as “alien” by all manner of people—both strangers online and close collaborators, by those who meant it affectionately as well as those who didn’t—Vitalik exuded an awkwardness that practically manifested physically. ## The DAO A large part of the book is about the battle surrounding The DAO. This is a somewhat more complicated story if you are not familiar with The DAO. That way of organizing, as if it were a venture capital firm, was already innovative, but it also raised $150 million USD in a kind of crowdfunding. The Ethereum teams placed a contract on Ethereum that also partly regulated its governance. This DAO was actually a symbol for the deeper, idealistic, utopian idea of decentralizing property and power. And it was precisely with this symbol that things went terribly wrong. Through a hole in the Smart Contract, the value of this DAO got hacked and Ether (ETH) worth 60 million USD was stolen. So, there was an expensive error in the code, and eventually the team made a central intervention to fix the error with a "fork". A branch of the blockchain code to a moment before the error occurred. For some people, this was a kind of mortal sin against the basic idea, the philosophy. For others, it was a moment of maturity, a necessary intervention. Shin describes this struggle as a factual account, but with great attention to the moral aspects of the protagonists. It transcends the facts and becomes a struggle between good and evil. A delightful read. > Similarly, Griff felt the DAO attacker’s life would have been “way better” if he had informed the community. “He’s early Solidity 2016\. He would be a hero,” said Griff. “He really screwed the pooch… Reputation is way more valuable than money.” ## Ethereum blockchain What makes Ethereum such a legendary blockchain? In short, writing and executing code on the blockchain. So in addition to currency and token, on Ethereum you can execute code in the form of a Smart Contract. The book helped me to understand that Ethereum requires a network of computers (clients). Clients on which actions can be carried out and which can be made in different computer languages, such as Go or C++. ## A story for understanding I had been working for months to better understand the crypto world and especially Ethereum. I looked at it factually and went on Wikipedia, forums, Twitter and did a course in Solidity (programming language for Ethereum contracts). But it only really came to life when I read this book. It unfailingly showed me the power of a story and that knowledge, without real experience, is difficult to acquire. Because you are taken by the hand, the images of the development process, the places, the people, their emotions, this abstract world really comes alive for you and Ethereum suddenly becomes clearer and more tangible. I have read this book breathlessly, and it has thoroughly changed my image of the crypto world and its technology. Laura Shin makes you relive (recent) history with her book, and it makes you feel like you were there. As if you experienced yourself how Ethereum came into being, overcame difficulties. What I also found great is that the author takes you along in her story to the power within a company or organization. I worked in a corporate environment for years and never really understood the parallel world of power that often seems to be at odds with reason, productivity, or empathy. Reading this book, it became clearer to me why people in startups work long hours for little money. And how important it is then to have certain claims on later. Do I recommend this book? As a reader I experienced the rise of crypto and Ethereum, you read an exciting book with a powerful story, and you gain insight into social relations and human nature. If you are interested in the world of computers and crypto, yes, I recommend The Cryptonians without any doubt. [The Book - Laura ShinThe story of the idealists, technologists, and opportunists fighting to bring cryptocurrency to the masses. Order your copy of The Cryptopians today.![](https://laurashin.com/wp-content/uploads/2022/01/favicon.png)Laura Shin![](https://laurashin.com/wp-content/uploads/2022/01/Cryptopians_Instagram_01a.jpeg)](https://www.laurashin.com/book/) ## Talking about recommending I actually recommended the book to my 26-year-old daughter, and she was just as enthusiastic as I was. We read it together and that was even more fun because it is more her world than mine. ![My daughter reading the book](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/05/67352994848__B2F20B5F-3E0F-4FC4-85DC-F718718FF157-Large.jpeg) My daughter reading the book ### Lead Magnets and GDPR URL: https://rob-hoeijmakers.ghost.io/gdpr-and-lead-magnets/ Last updated: 2026-04-24T13:21:42.000Z [Lead Magnets](https://rob-hoeijmakers.ghost.io/lead-magnet/) are giveaways for obtaining e-mail addresses of interested parties. What are the rules for Lead Magnets within the [General Data Protection Regulation (GDPR)](https://rob-hoeijmakers.ghost.io/gdpr/)? ## Lead Magnets and newsletter signup GDPR requires specific, freely given and clear consent. Consent should not be a condition for receiving a lead magnet (gift, webinar, e-book, or workbook). So if GDPR applies to you, then you can **no longer add someone to the general marketing or newsletter list** when they sign up or subscribe to a lead magnet. It also means you can't add a newsletter tick to the opt-in and make it mandatory. ## Specific permission for newsletter sign up If someone wants to be added to your general marketing list, **they must specifically agree to join the list**, separate from any other offer. This means you can add a newsletter checkbox to the opt-in and make it optional. This way, if someone enters their email AND checks the checkbox, they can be added to your newsletter list. If they only enter their email address, they'll only get the lead magnet and its follow-up. ## Options to comply with GDPR Here are options to offer a lead magnet and stay GDPR-compliant: - Add a voluntary (unchecked) checkbox to your opt-in, asking if they also want to receive your newsletter. - Create a "middle page or sandwich page" between the opt-in and the “Thank You” page asking if they wish to subscribe. - Include a call to action in the email that gives access to the lead magnet. - Include promotions for your newsletter in your lead magnet. - Present the benefits of your newsletter and ask for specific permission in the email sequence for a lead magnet. *Also read: [Introduction to GDPR](https://rob-hoeijmakers.ghost.io/gdpr/)* ### Review: The Power Law by Sebastian Mallaby URL: https://rob-hoeijmakers.ghost.io/the-power-law-by-sebastian-mallaby/ Last updated: 2026-05-10T08:54:20.000Z Although I've been working professionally with the web since 1993, I've never had much insight into the financial side of it. I produce content and deal with scaling websites and online marketing. But now I wanted to know more about the forces of money in the world of Big Tech and Startups. To better understand my professional Werdegang. And also to see what the possibilities are if I want to further fund or invest in another company or startup. And to be able to join the discussion. ## The book The Power Law Through a [podcast of Azeem Azar](https://hbr.org/podcast/2022/02/how-venture-capital-made-the-modern-world-with-sebastian-mallaby), I heard of Sebastian Mallaby and his book The Power Law. That book is so full of names, facts, and stories that there is little point in recounting or summarizing it. If you are interested in the world of Venture Capital, Start-Ups and Silicon Valley, you simply must read it. A well-written book with a very thorough foundation. Mallaby does **a masterly job** of narrating Venture Capital from 1948 to the present day. He recounts the stormy developments in 14 chapters. Some histories are analytical but many are narrative and tell of the characteristics of individual players such as Thiel, Andreessen, Son, or Milner. What drives them, what are the strengths or weaknesses of their personalities. What is forceful about the book is how you feel that times are changing, when the public internet will be big in the mid-nineties. It's like there are boats, and you know a hurricane is going to hit. Who will survive it and how? Very excitingly written. ![Blossom time in "the valley of heart's delight", Santa Clara County, California.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/11/twitter_d994f4c8350ac5c5f37e11afef486c1e.jpeg) Blossom time in "the valley of heart's delight", Santa Clara County, California. ## Do VC's make a substantive contribution? Much of the book, in addition to the stories, revolves around whether Venture Capital makes a substantive contribution or whether they are a coincidental factor in the successes. Mallaby writes about this: *The undoubted success of VC-backed companies is often qualified with a question: Did the VCs create the success, or did they merely show up for it? But, as we have seen already, another line of research shows how startups that have the benefit of venture guidance fare better than their peers, and this book has related multiple cases of VCs positively impacting portfolio companies. Besides, even if it were the case that VC skill lay entirely in deal selection, and not in mentoring startups, that skill would still be valuable. Intelligent deal selection increases the chances that the most deserving startups will get the capital they need. It ensures that society’s savings are allocated productively.* *This book has pushed back against the randomness thesis, emphasizing instead the skill in venture capital. It has done so for four reasons.* - *First, the existence of path dependency does not actually prove that skill is absent.* - *The second reason to believe in skill lies in the origin story of some partnerships* - *Third, the idea that venture capitalists get into deals on the strength of their brands can be exaggerated.* - *Fourth, the anti-skill thesis underplays venture capitalists’ contributions to portfolio companies.* *This book has also made a second argument. Whatever the skills of particular venture partnerships or individual VCs, venture capitalists as a group have a positive effect on economies and societies.* The book The Power Law by Sebastian Mallaby is a cleverly written, exciting and informative book about the revolutionary developments in information technology. A book worth reading for a better understanding of the financial dynamics behind this industry. [The Power Law by Sebastian Mallaby: 9780525559993 | PenguinRandomHouse.com: BooksShortlisted for the Financial Times Business Book of the Year Named a Best Book of 2022 by The Economist “A gripping fly-on-the-wall story of the rise of this unique and important industry…![](https://www.penguinrandomhouse.com/favicon.ico)PenguinRandomhouse.comSebastian Mallaby![](https://images3.penguinrandomhouse.com/cover/9780525559993)](https://www.penguinrandomhouse.com/books/580120/the-power-law-by-sebastian-mallaby/) ### Further reading - [My Venture into Venture Capital: What I Learned About DeepTech Investing](https://rob-hoeijmakers.ghost.io/venture-capital-deeptech-investing/) ### Tools for creating landing pages and lead magnets URL: https://rob-hoeijmakers.ghost.io/tools-for-creating-landing-pages-and-lead-magnets/ Last updated: 2026-04-24T18:46:18.000Z There are a number of online tools for [landing pages](https://rob-hoeijmakers.ghost.io/leads-generating-landing-page/) and lead magnets. Sometimes they are completely dedicated to this task, sometimes they are modules within an existing system for marketing. The ones below I know from my own experience (and will continue to add to them): 1. Leadpages 2. Wix 3. HubSpot 4. ConvertKit 5. Squarespace 6. Linktree 7. WordPress ## Leadpages [LeadPages](https://www.leadpages.com/?ref=hoeijmakers.net), for example, is entirely focused on being able to quickly create landing pages yourself and store the leads. A lot of the value is also in their extensive set of templates, and these can help you get started quickly. A template like this quickly makes it clear what can and cannot be done. As a bonus, they also let you host a website. ## Wix There is a pretty extensive and well-supported set of [Landing Page templates in Wix](https://www.wix.com/website/templates/html/landing-pages?ref=hoeijmakers.net). ![Landing pages in Wix.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/07/Landing-pages-Wix.jpg) Landing pages in Wix. ## HubSpot The marketing system [HubSpot has a module for landing pages](https://www.hubspot.com/products/marketing/landing-pages?ref=hoeijmakers.net). There is a set of templates with landing pages and lead magnets. The data from users is nicely integrated into the CRM. ![HubSpot landing pages](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/07/Hubspot-Landing-Pages-2.jpg) HubSpot landing pages ## ConvertKit The marketing tool ConvertKit offers a great looking set of Landing Pages that integrates nicely. These landing pages can be used to trigger an e-mail sequence. ![ConvertKit Landing Page templates.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/03/2022-03-18_06-29-20.jpg) ConvertKit Landing Page templates. ## Squarespace Within a website builder like [Squarespace](https://rob-hoeijmakers.ghost.io/squarespace/), for example, you quickly build a landing page yourself. The navigation can be hidden, and then you build the landing page. ## Linktree Handy service to create a [Micro Landing Page ](https://rob-hoeijmakers.ghost.io/your-bio-site/)to use with Instagram or Twitter profile. ## WordPress In [WordPress](https://rob-hoeijmakers.ghost.io/how-about-wordpress/), you can also create a compelling lead magnet with an easy-to-follow roadmap for WordPress users. Increase your conversions and attract valuable leads using these effective strategies. Click here for [the full roadmap from WPForms](https://wpforms.com/how-to-create-a-simple-lead-magnet-optin-form-in-wordpress/). 💡 Creating a compelling landing page is crucial for converting visitors into leads, but it doesn't end there. Ensuring your landing pages are ****optimised for search engines** is equally important to drive consistent traffic. Discover the best practices for crafting landing pages that not only engage but also rank highly in search results. Learn more about the synergy between effective landing page design and SEO strategies in this valuable resource: [Landing Page SEO Guide](https://t-ranks.com/content/landing-page-seo/). ### English editor correcting the American translator URL: https://rob-hoeijmakers.ghost.io/english-editor-correcting-the-american-translator-and-vice-versa/ Last updated: 2026-05-10T08:54:21.000Z There is a difference between the vocabulary and spelling of British English and American English. And the difference is minimal, but big enough to cause annoyance and even drama. A story about how spelling correction caused a minor argument between my translator and editor, and how spelling played a role in the run-up to the Iraq war. ## Writing in one's own language I start all the articles on this blog in my own language, Dutch. That works better and feels more creative, freer. Many a reader will recognise this, no matter how good and thorough the English. After a first draft, or a few drafts, I develop an article. Then I first edit the own language via LanguageTool and then translate the article. The translation goes via [DeepL](https://www.deepl.com/translator). The quality of this is convincing. I copy my text and paste it into the translation application. The translated text goes into a newly created file. I read it carefully and again correct and edit it with the help of LanguageTool, the language assistant. ## Quarrelling Until recently, there were quarrels, the editor was picking a fight with the translator. And vice versa. One programme corrected the other, and it made me uncertain whether my articles were published without errors. And then I dug deep and found out that [LanguageTool](https://languagetool.org) (included in my writing programme [Ulysses](https://rob-hoeijmakers.ghost.io/blog-writing-workflow/)) takes over the computer's default settings and that DeepL has a setting per translation. ![DeepL gives you an option between translating to English the American or British way. ](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/04/deepl-american-british-english.jpg) DeepL gives you an option between translating to English the American or British way. Then I did [a poll on Twitter](https://twitter.com/robhoeij/status/1504168561765494791?s=21) and 80% of my readers think British English is the standard for the world language and not American English. So I cut the knot and conformed to that standard. Now the computer and therefore the editorial staff are set to British English and I use that as the setting for my translations. Now I can concentrate again on the content. ![The language system setting on a Mac.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/04/image.png) The language system setting on a Mac. ## War in Iraq In the run-up to the Iraq war, a newspaper published secret, American documents showing that the UK was illegally withdrawing from the war. An article appeared in the newspaper, which also reproduced the original document. The article caused a storm of attention, but soon there were reports that it was a hoax. Someone had found out that the secret American document used the British spelling. Later, it turned out that a well-meaning intern had spelt-corrected the document. 📱 To make sure this worklow is also aligned on the iPhone you need to adjust not the phone language but the [keyboard language](https://discussions.apple.com/thread/2738789). ### Stephen Hawking's American voice Stephen Hawking's switched from standard English to American English when he found a machine to synthesize his words. Weird, but his computer voice, and accent, is much better known than the personal voice he lost. To the point that people, like me, would think he was American. "It's American!" "Is that a problem?" --- ### Further reading - [‘You’ve caused an international incident’: how my work mistake came back to haunt me](https://www.theguardian.com/film/2019/jul/27/international-incident-work-mistake-official-secrets-film) - [American versus British English spelling | Cambridge English](https://www.cambridge.org/elt/blog/2017/09/08/american-british-english-spelling/) ### Make your own bio site URL: https://rob-hoeijmakers.ghost.io/your-bio-site/ Last updated: 2026-05-10T08:54:21.000Z Platforms are a curse and a blessing for the digital, creative producer. They have a large audience and to reach it you put yourself at their service. But you are actually maintaining a rented house. So, I choose the golden mean where I start with someone else's product, to get going quickly, and then later bring it to myself. An example is Linktree. If you're active on Instagram, you've probably had to deal with such a [Landing Page](https://rob-hoeijmakers.ghost.io/tools-for-creating-landing-pages-and-lead-magnets/). ## Why use a bio landing page? It's a small but important node on the web and requires some explanation. The bio landing page is an addition mainly to Instagram and Twitter profiles. This single page is a simply designed web page, aimed for mobile views, with the essential links to your website, web pages and social media accounts. All the important and up-to-date pages accessible via a single address. Extra significant for Instagram and Twitter, because there you only have one reference for all your posts. An example of this link you can see here on my Instagram profile: [![Example of profile link on Instagram](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/12/IMG_136A8BF38698-1.jpeg)](https://www.instagram.com/robhoeij/) Example of profile link on Instagram One obstacle to the flow of traffic is therefore this profile page. You do a post on Instagram and in the post you invite people to view a product or a related social account like Mastodont. It is impossible to include a reference to a web page there, unpaid. What you can do then is call on the interested reader to click on the link in the bio of the profile. And so that you don't have to adjust this link all the time, or you want to make multiple adjustments, you can also create a dedicated page with only references and link it. ![Example flow from Instagram profile to own online property](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/04/Instagram-linktree-website.jpg) Example flow from Instagram profile to own online property 🌳 The company Linktree has jumped into this gap, and they offer a great service. You create an account and then by filling in forms you link your pages and social media. It immediately looks pretty professional, a must: ## Manage this linking page yourself Eventually, there are disadvantages to an external service, and it is important to be and remain independent. It is an essential step for every producer to consider eventually where you want to remain your own boss. There are limitations to free products in particular. The **three biggest advantages of making your own Linktree bio page**: - Better for SEO - Control over your brand and domain - Adding tracking tags It is better to create your own list of links and place it on a domain of your own. That's precisely what I did last week. One on a stand-alone domain or preferably one on a subdomain of the main domain. If you like it, below I shared my way of working. > "The best time to plant a tree is twenty years ago. The second-best time is now." ## Better buy than borrow In short: Linktree is a nice, quick way to link Instagram to your other online assets. But, if you are a producer who wants to make serious use of social media traffic and use these channels for good brand recognition, more followers, more traffic, make sure you have your page. The idea of putting the most recent and indispensable links there is a great starting point and also very useful for Twitter profile. It's a key page, so it's better to buy it than borrow it eventually. 🌳 Create your own linktree page. I found a template and adapted it to my liking. In HTML and with an editor. When I was done, I created a repository on GIT on the server of my web hosting. Every time I make a change on my computer and push it via GIT, the web page changes. I found [the instructions on GitHub](https://gist.github.com/cjsteel/5bdab49c97ecacb67904056ccdcb956d). [![Link to bio micro page on X (Twitter)](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/12/2022-12-18_22-04-58.png)](https://rob.hoeijmakers.net/) Link to bio micro page on X (Twitter) --- ### Further reading - [Leveraging a Viral Photo: A Journey in SEO and Digital Strategy](https://rob-hoeijmakers.ghost.io/viral-photo-seo-digital-strategy/) ### The outside of a web page URL: https://rob-hoeijmakers.ghost.io/the-outside-of-a-web-page/ Last updated: 2026-04-25T13:17:13.000Z ## The outside The technical name for this outside is Open Graph (og). This is a technical term for a collection of information that is invisibly inserted into a web page. And it is this information that is used when you view the page from the outside. Like a label. ## Like a label or packaging On social media, when you share a page, this label is highly visible. So for example on Facebook, LinkedIn, Pinterest or Twitter. But also if you forward a page via WhatsApp, Messenger or iMessage. A good label makes a huge difference to the presentation on the channels, of course. Especially the image. And the application of such an image requires some extra care and work. This image does not fit into a page in the same way as the texts do, but has to be placed separately (just like normal images) and then referenced. 💬 The outside of a web page is made up of metadata. Check out [the function and importance of metadata](https://rob-hoeijmakers.ghost.io/metadata/). ## Open Graph image How you work with the image when creating a page depends very much on the tool you use to create it. Usually it is an extra option next to the page where you can fill in fields and add the image. I will give a short example to give you an idea, and then indicate what is a good size for the image. ## Example of adding an image As an example, I'll take a landing page that I made for an online course Introduction Squarespace. I made that landing page in Leadpages. On that page, under Page Settings, I find a tab called Social and there is the option to add an image, title, and description. ![Open Graph configuration example](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/04/Sociale-Graph-Image-example.jpg) Open Graph configuration example When published, this does not produce a visible result, but in [the code](https://www.opengraph.xyz/url/https%3A%2F%2Fmaya.coach%2Fcursus-introductie-squarespace%2F/) you can find the text and the reference to the image (as [metadata](https://rob-hoeijmakers.ghost.io/metadata/)). Open Graph information is also a factor in the best possible find-ability ([SEO](https://rob-hoeijmakers.ghost.io/my-basis-for-seo/)). ## Visible when sharing If I now share the page, the information will become visible. I will give examples. Starting with this reference here: And when sharing on Facebook, it looks like this: ![Facebook newsfeed Open Graph example](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/04/Facebook-Social-Image.jpg) Facebook newsfeed Open Graph example And iMessage: ![iMessage Open Graph example](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/04/2022-04-01_07-32-38.png) iMessage Open Graph example ## A technically correct image I use Facebook's instructions because they also invented Open Graph, and use this format: - Landscape, wide but narrow - 1200 pixels by 630 pixels ([aspect ratio](https://rob-hoeijmakers.ghost.io/aspect-ratio/) 19:10) - Not larger than 8 MB - Jpeg or PNG Before posting, you can also try out the Open Graph with the [Sharing Debugger](https://developers.facebook.com/tools/debug/?q=https%3A%2F%2Fmaya.coach%2Fcursus-introductie-squarespace%2F&ref=hoeijmakers.net). ## Important outside Now that we know how it works with web pages, you can focus on the content with love and care. When I choose the main image, I consider its complete use, also for use on the outside. I select an image that I can make landscape, portrait, or square. The outside is important, it's the first acquaintance with a web page and I better make a good first impression. 👾 Preview and generate Open Graph meta tags for your own site? ## Links to Open Graph resources The best start for further research is to check out the inventors of Open Graph, Facebook. As you can see, they don't particular care about the packaging 😉 - [Images in Link Shares](https://developers.facebook.com/docs/sharing/webmasters/images/?ref=hoeijmakers.net) - [The Easiest Way to Preview and Generate Open Graph Meta Tags](https://www.opengraph.xyz/?ref=hoeijmakers.net) ### Shoot in both Portrait and Landscape URL: https://rob-hoeijmakers.ghost.io/portrait-and-landscape/ Last updated: 2026-04-25T17:07:45.000Z A short post about what I have learned about images that need to fill multiple sizes and slots. I will show you by a personal example: I'm teaching a web copywriting workshop and for this I have a landing page with social image, a checkout page and a campaign on Facebook/Instagram. Of course I will also want to share my landing page on Twitter and LinkedIn. ## Find the words The workshop is about how - once you have a strategy, plan and schedule - you can overcome your fear of the blank page. So, I take the people with me through keywords gathering, copy writer briefing, and a series of very tangible and concrete tips. And that message of finding the words needed to include an image. You want to stand out but also have it functionally fit with the subject. I quickly made a pre-selection and ended up with the Scrabble letters. This is a strong idea because we need to find words. Unfortunately, the image was in English, so I decided to do produce on myself. ## Creating my own pictures I looked for a scrabble board, came up with a text, set the lighting and took a photo with the iPhone. What I learned soon after, especially since there was text on it, is to take [landscape and portrait](https://rob-hoeijmakers.ghost.io/portrait-and-landscape-images-on-social-media/) shots with plenty of room to crop. To be able to crop to a suitable ratio between height and width. Let me show you why. ## Formats The campaign is running on Facebook and Instagram, in the news feed and stories. The images there are generally standing or square: ![Facebook and Instagram ads previews](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/03/2022-03-22_11-33-01.png) Facebook and Instagram ads previews ![Instagram, Facebook and Stories preview](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/07/2022-03-22_11-33-12.jpg) Instagram, Facebook and Stories preview Then there's the landing page of the workshop and there I can choose, landscape or portrait: ![Workshop landing page in LeadPages](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/09/2022-03-22_11-35-46.jpg) Workshop landing page in LeadPages ## Open Graph image But the accompanying social media image for the landing page, Open Graph Image, is again landscape oriented with a different aspect ratio than the iPhone standard (4:3). I made that image landscape and then cropped it back to 16:9\. Otherwise, the letters are partly out of the picture. Now they look really nice and readable when the page is shared: ![Landscape image with 16:9 ration use in iOS iMessage.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/09/IMG_AE1E4FFD08E0-1.jpeg) Landscape image with 16:9 ration use in iOS iMessage. ## Lessons learned In conclusion, the ability to shoot images in both portrait and landscape formats is an indispensable asset for an online professional. As demonstrated through our exploration of various platforms and their distinct image requirements, the versatility of having both formats at your disposal is paramount. Whether it's Facebook and Instagram, your workshop's landing page, or Open Graph images, the adaptability of your visuals across different contexts can significantly impact the success of your campaigns. Remember to leave room for cropping and adjustments to ensure your images shine in every setting. As you navigate the ever-evolving landscape of online marketing and content creation, keep in mind that it's not just about finding the right words but also selecting the right image formats. These seemingly subtle choices can wield immense influence over how your message resonates with your audience and spreads throughout the digital realm. So, embrace the power of versatility in image formats, and let it be your ally in crafting compelling, shareable content that makes a lasting impression. ### Pixels per Inch (PPI). Measured as a line or a square? URL: https://rob-hoeijmakers.ghost.io/pixels-per-inch-ppi/ Last updated: 2026-04-25T13:13:56.000Z For years, I tried to figure out screen resolution in relation to the **physical properties of a screen**. Only now I understood that the physical properties of a screen, that is the number of pixels, are not measured on a surface but as a line. So when it says 300 pixels per inch, these are on **a line an inch long**. And there are many more on a square inch. There are 90,000 physical pixels on a square inch of such a screen. Of course, assuming the pixels are squares. A bit one dimensional to mix up lines and surfaces when defining resolution 😉 ## Calculations To have this figured out made me happy. This new knowledge gave me insight and input for calculations. The PPI of a display can be derived from the **diagonal size of a screen in inches** and the resolution (horizontal and vertical) in pixels. The calculation is done in two steps. ![Example of Pythagorean Theorem for Pixels per inch](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/03/Screen-PPI0-1.png) Good old Pythagorean Theorem for Pixels per inch 1. Calculate the diagonal resolution in pixels, relying on the Pythagorean Theorem: dₒ = √(w² + h²) w = width and h = height 2. Calculation of the PPI is the total numbers of pixels divided by the diagonal size: PPI = dₒ/dᵢ With these two simple formulas, I can plug in the numbers. ## Outcome for my e-reader So for the screen of six inches diagonal and resolution of 1072 by 1448 it is: dₒ = 1801.6 It is six inches, and so I divide by six. PPI = 300.27\. ![1800 pixels in diagonal divided by the diagonal in inches gives a PPI of 300 pixels per inch.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/03/Screen-PPI-ereader0.png) 1800 pixels in diagonal divided by the diagonal in inches gives a PPI of 300 pixels per inch. Close! Pretty, pretty close. And now the theory is clear, I also can use and understand this calculator: [Pixels Per Inch PPI Calculator](https://www.calculatorsoup.com/calculators/technology/ppi-calculator.php) ### ### Stay safe online with VPN URL: https://rob-hoeijmakers.ghost.io/virtual-private-network-vpn/ Last updated: 2026-04-25T13:17:11.000Z VPN is **one of the best ways to secure your internet traffic and protect your privacy**. It also gives you access to lots of protected information. Let me explain. ## Personal VPN for security and privacy Personal VPNs enhance internet security and privacy by creating an encrypted tunnel to a server, masking your identity. This is especially useful on untrusted networks, like public Wi-Fi, safeguarding against potential breach. Using **Personal VPN to secure your connection and have more privacy**. With a VPN connection like this, you make an encrypted connection (a tunnel) to a server, and then you continue on the web from there. No one can then see that it is you because the traffic then appears to come from that VPN server. This form is very practical if you are using a connection that you do not trust. For example, a free and open connection somewhere on the road. Then it is a very smart thing to use this VPN connection. Another important use of the VPN is if the government (or your employer) decides to shut down certain things. Something like this happened again recently in Russia. There they shut down Twitter and so you had to use a VPN connection. ![List of foreign connections optimized for streaming](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/03/Personal-VPN.jpeg) Personal VPN on iPhone. I am using CyberGhost. ## VPN connection for more access Another aspect of VPN, which is interesting for many people, is to use a Personal VPN to consume content which is actually limited to a certain geographical area. You then create a tunnel to a server in that protected area. That's how you get around the restriction and it is called geolocation spoofing. I myself also have such a service (CyberGhost) that I use for more security and sometimes to access the web from another country, mostly to watch geofenced German TV. That's also handy if you want to check Google for another area, for example, or stream content on Netflix in the US. ![List of foreign connections optimized for streaming](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/07/VPN-List-of-servers.jpg) List of foreign connections optimized for streaming ### Stay safe So there are basically two different sorts of VPN's and different goals for use. One to safely work with a local, non-internet network and a personal one to protect your data, be more secure and to be more independent of geographic location. I hope you learned that VPN will mean something different to a business owner or IT manager, to someone working with sensitive data like a journalist, and to the consumer who wants more or cheaper access to content. ### VPN Services review On [digital.com](https://digital.com) (review site on online tools for small business) you find a helpful VPN services review. They put in 40 hours of research on 120+ VPN companies from across the web. These reviews and the VPN guide help small businesses and startups find the best VPN service for their business. 💡 ****Google One VPN** The VPN feature included in the [Google One subscription](https://rob-hoeijmakers.ghost.io/choosing-google-one-to-use-google-meet/) primarily serves to enhance online security, especially useful when connecting to the internet via public Wi-Fi networks. However, it's important to note that this service does not support geolocation spoofing. This distinction makes the Google One VPN particularly suited for users primarily concerned with safeguarding their data on unsecured networks, rather than those looking to bypass geographic restrictions on content. ### Generate leads with Landing Pages URL: https://rob-hoeijmakers.ghost.io/leads-generating-landing-page/ Last updated: 2026-04-24T18:14:36.000Z Who still does anything with a website after launch? Some are very active with it and use a [web builder or CMS ](https://rob-hoeijmakers.ghost.io/content-management-system-cms/)to make regular changes. But I bet the majority are reluctant to do anything more to it or leave it to others. And that costs money. So what's the alternative if you do want to advertise a product or event? I use **Landing Pages**. ## What is a landing page? The short answer is a single page, with no distractions, focused on a single action. So no navigation and separate from the normal website. It must serve as an encouragement to take a next step in the process you use to attract customers. Maybe they subscribe, share an email address for more information, or they start the buying process. This next step is made with the Call to Action (CTA). ## Are landing pages difficult to create? In practice, there will often be a form attached to the page and a thank-you page. So some skill and technology is required. Perhaps, with [the right tools](https://rob-hoeijmakers.ghost.io/tools-for-creating-landing-pages-and-lead-magnets/), all you need to be is a bit digitally handy. ## Write for a landing page Perhaps many people still find this the biggest challenge. How do I get texts? Or how do I write them? Keep in mind, you are not “writing a landing page” but writing **for** a landing page. It is not a linear story but a composition with multiple text elements and images, layout etc. Especially when you work with templates, you fill in the blanks or replace dummy text. The key is to get directly in touch with the feelings of your audience. What is the problem, the pain and what kind of journey are you going on with them to make a transformation. How do we move from the unwanted A to the desired B? A -> B Think about who your dream customer is, make it completely clear. What does the customer want, how does the customer feel, what are their challenges. 💡 Unlock the [power of ChatGPT](https://www.hoeijmakers.net/try-out-chatgpt-yourself/) for crafting compelling landing pages! Whether you're new to web strategy or an experienced marketer, ChatGPT can provide expert insights to enhance your content realisation. From [SEO guidance to conversion optimization](https://www.hoeijmakers.net/my-basis-for-seo/), and the latest in marketing technology, ChatGPT can be your trusted partner. ## Some ingredients for a landing page It's like cooking food. There are recipes, but keep experimenting. In my experience, these are tasty ingredients for a good landing page: 1. Name potential objections 2. Provide evidence, social proof 3. Don't make it too long 4. Talk about the benefits (and bonus points for storytelling) 5. Take extra care with titles and subtitles 6. Call to Action 🤔 ****Difference between a landing page and a lead magnet?** Both landing page and [lead magnet](https://www.hoeijmakers.net/lead-magnet/) are aimed at getting contact information from a visitor and to turn them into a lead or customer. With a Lead Magnet the visitor can receive an ****incentive**, mostly in a digital format. An e-book, checklist, free course. The lead magnet can ****also be presented on a regular page on the website** as a sort of banner or be the purpose of a landing page. ## ### From Kobo to Twitter in 18 seconds URL: https://rob-hoeijmakers.ghost.io/from-kobo-to-twitter-in-18-seconds/ Last updated: 2026-04-14T13:21:19.000Z To make a quote from an e-reader like Kobo to another medium was always a pain. How do you get from reading something to making a reference on some social media like Twitter? ### Pain What I used to do was go to my computer instead, searched for the book, open it and find the text I was looking for. Once found the text I would do a copy and paste. In practice it did not happen and I even rather typed it over. ### Mobile workflow But with the new text feature ([Live Text](https://rob-hoeijmakers.ghost.io/iphone-live-text/)) on the iPhone I found a much quicker and easy to use workflow. It is even a bit magical. Like pointing at the thing you want to copy. You read a nice line and want to share it: 1. Point the iPhone camera on the e-reader page. 2. Select the text you want. 3. Use the Live Text select function. 4. Go straight to share and pick the Twitter app. Clean up, add text, add tags and GO. ### The Tweet > Unicorn ♣️ > > — Rob Hoeijmakers (@RobertHoeijmak1) [March 11, 2022](https://twitter.com/RobertHoeijmak1/status/1502189908164812800?ref%5Fsrc=twsrc%5Etfw) ### Blog writing workflow URL: https://rob-hoeijmakers.ghost.io/blog-writing-workflow/ Last updated: 2026-04-25T13:16:40.000Z A lot comes down to good tools. I'm picky about this and am happy to share my methods for writing and formatting the articles on this blog. Hopefully, it will give you some ideas. ## Starting an article I gather information first on paper and then as notes in Apple Notes. When the idea has taken some solid shape, I start a blank sheet in the writing program Ulysses. ## Writing in Ulysses I write all my articles in [Ulysses](https://ulysses.app), let me tell you why. For starters, I think it's beautiful. A quiet page layout and the ability to remove or leave out all the superfluous. It's quick and robust. So much for the declaration of love, back to the method for a moment. ### Writing assistance Ulysses has LanguageTool built in, and that tool knows just fine what language I'm writing in. It not only makes corrections for spelling errors, but also makes suggestions for grammar and even style. Thus, I have my own writing assistant. ![The writing assistant LanguageTool in Ulysses](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/03/2022-03-08_19-00-37-1.jpg) The writing assistant LanguageTool in Ulysses ### Searchable All texts I have written in Ulysses are searchable on the Mac as well as the iPhone. Both at the system level (Spotlight) but also through Ulysses itself. With this setup, I can find something quickly and also open it directly from search. ![Search through my texts from Mac or iPhone.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/09/IMG_FD4B33CE0870-1.jpeg) Search through my texts from Mac or iPhone. ### Cloud sync between Mac and iPhone The texts I write are instantly available both on the Mac, MacBook, and iPhone. They are kept in sync by the iCloud. So there is no need for extra accounts or anything. Often I kill some time when I have to wait to write or improve an article I'm working on at the time. ### Markdown The text format of the markup is [MarkDown](https://rob-hoeijmakers.ghost.io/using-markdown-for-text/). Sounds like a very technical thing, but it's actually quite the opposite. At best, it's a bit minimalist. Markdown is simple text with Markup for formatting, references and the like. Light, simple and interchangeable. I am the master of technology and not the other way around as with the terrible Word. ### Images Although MarkDown is used, you can easily work with images and also embed them in a single document, like PDF, for sharing. ![Working with images in Ulysses](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/03/2022-03-09_08-04-18.jpg) Working with images in Ulysses ### Coupling with Ghost A big advantage of Ulysses is that it can be linked with [blog platform Ghost](https://rob-hoeijmakers.ghost.io/ghost-blog/). I publish directly from Ulysses, as it were, without copying and pasting or anything like that. Output can also be HTML, Docx or PDF and that's easy if you quickly want to share something with someone. ### Other nice things A nice feature is that there are versions of each document and I can basically go back in time. Also, there is a nice overview of the structure of the document and work counting plus reading time indication. ## Clever Web Strategies ### Blog for the online professional. Subscribe Email sent! Check your inbox to complete your signup. No spam. Unsubscribe anytime. ## Language and translations I prefer to write the articles in my language, Dutch. Only when the first solid version is ready, I start translating it. Via the free version of the excellent translation program [DeepL](https://deepl.com). ![Translate with Deepl.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/03/2022-03-08_20-07-28.jpg) Translate with Deepl. The translation is excellent, but seen from the workflow, this is actually the least pleasant step. For translating, I make a copy and paste it into the website of DeepL. I seem to **lose markup** here, and then I have to copy the images manually and add links again. I will take the time to see if I can improve on this step, but for now, it is annoying. How nice it would be if translation were a better integrated part of the workflow. 🤔 I did some research on copy and pasting the text into the translator and noticed that links do not get lost when you export the written text as MarkDown. And then copy and paste, make a new markdown file to then import again. But then you still lose the pictures because they seem to be linked on the file level. A lot of extra work for limited improvement. I then edit this first English version again in Ulysses. ## Final edit and publication The first version goes from a writing program, as draft, to my blog. So from Ulysses to Ghost. Here's the final edit. Layout tidying up, [check outline of HTML headings](https://rob-hoeijmakers.ghost.io/html-headings/), writing summary,[ image descriptions](https://rob-hoeijmakers.ghost.io/micro-content-big-impact/) and adding [meta information](https://rob-hoeijmakers.ghost.io/metadata/) to page and images. Then, once I'm done with it and hit publish, I can send it out the door immediately or set it up for later. I also get the option to send the article as a message to my subscribers. ## Post-Production: Webmaster Check with Ahrefs After hitting the 'Publish' button, the work isn't quite done yet. To ensure that my articles are [SEO-friendly](https://rob-hoeijmakers.ghost.io/my-basis-for-seo/) and free from any website issues, I make it a point to regularly run a free webmaster Ahrefs report. This step is crucial for optimizing the 'health' of my blog. Ahrefs helps identify missing alt tags on images, too long or too short meta descriptions, broken links, and more. It serves as a diagnostic tool that points out the areas for improvement, helping me to maintain the quality and discoverability of my blog. This way, I not only focus on the writing and publishing aspects but also pay attention to the overall performance and searchability of my articles. ## Costs There are the annual [costs for Ulysses](https://ulysses.app/pricing/) and for [Ghost](https://ghost.org/pricing/). DeepL, I can use for free. ## Conclusion When I sum it all up like this, it seems like quite a story, but actually, I have never had such a pleasant workflow for publishing. 1. Gather notes 2. Start composing the article and do the writing 3. Create links and images 4. Translate 5. Publish draft 6. Finalize (text, micro texts, links, images) 7. Publish Except for the translation step, that one is a bit cumbersome. I will try to optimize that step, and [please let me know](https://rob-hoeijmakers.ghost.io/contact/) it you found a better way. For the rest, I can only recommend Ulysses with Ghost for blogging. It works from my office set up, it works when I am on the road. ### Using 2FA with Apple Keychain URL: https://rob-hoeijmakers.ghost.io/2fa-within-apple-keychain/ Last updated: 2026-04-21T09:26:55.000Z This article is for you when you want to improve and simplify your cybersecurity and you : 1. Use a Mac and an iPhone together 2. Use Apple Keychain for passwords 3. Sync via iCloud 4. Your default browser is Safari So this is a bit of an article for the fans. And off course the content here probably is outdated when the next update by Apple is done, but then still you can learn a bit about the clever setup of 2FA. Apple doesn't advertise it much, and I get that. With 2FA you add security but rarely ease of use. Also, the current flows are certainly not without bumps. I wrote another article about what [Two-Factor Authentication](https://rob-hoeijmakers.ghost.io/two-factor-authentication-2FA/) is, and why this is important. The process consists of two steps. Creating a token once and later entering codes in browser or app. ### Setting Up 2FA for a Account To enable 2FA, you'll follow a series of steps that might vary slightly depending on the service you're using. Behind the scenes, a unique token is generated and linked to your account, but you won't need to worry about that. Here's a general guide: - **Choose Your Device**: You can set up 2FA on your phone or computer, or even let them work together. The latter is surprisingly simple, and I'll use a Google Account to illustrate how it's done. - **Navigate to Security Settings**: Head over to your Google Account settings and look for the Two-Factor Authentication option under the security tab. - **Select an App for 2FA**: Indicate that you have an app for 2FA, and you'll be ready to begin. A QR tag will appear on the screen. ![Image showing the scanning of QR-tag with iPhone for native 2FA](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/03/2022-03-04_13-57-45.webp) Scanning QR-tag with iPhone for native 2FAx - **Scan the QR Tag**: Simply scan the QR tag with your camera. Clicking it will open your password repository with Google accounts, where you can manually link the request to the correct account. If you only have one account, it's even easier. ![Image showing how to connect 2FA token to account in Apple password manager.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/03/IMG_1384--1-.webp) Connect 2FA token to account in Apple password manager. - **Confirm the Link**: Finally, confirm the link by typing the first code, and voila! 2FA is set up. Your settings will be backed up in iCloud and shared between macOS and iOS, ensuring a seamless experience across your devices. This process sets up 2FA without requiring you to understand the technical details of token generation. It's designed to be user-friendly and adds an essential layer of security to your accounts. ## Keychain is my password manager of choice Keychain does everything I need from a password manager. It safely stores passwords across devices, helps me generate new passwords and I can add 2FA in a seamless flow. Also, I can exchange accounts easily with my partner by airdropping them when necessary. *New development on the horizon: [Passkeys which might replace passwords all together](https://rob-hoeijmakers.ghost.io/passwords-passkeys/).* ### Two-Factor Authentication (2FA) URL: https://rob-hoeijmakers.ghost.io/two-factor-authentication-2fa/ Last updated: 2026-04-21T09:27:13.000Z Our insecure digital times call for extra care and attention to your digital security. It is just like taking care of your health that good health is not a certainty, but there are some things you are better off leaving and some that it is important that you do. In matters of digital security, you need to be resilient because we are more and more depending on Cybersecurity. For security, I prefer not to use technology that I don't fully understand myself, and so I delved into **Two-Factory Authentication (2FA)** and I share what I learned. Key takeaway is to prefer an authenticator app above SMS. ## What is Two-Factor Authentication (2FA)? Passwords, by themselves, are only moderate security. They are non-personally identifiable information that is also often easy to guess or find out. The name of your oldest child, cat or dog name followed by a year. Sounds familiar? We almost all do it that way because how else do you remember it all. 2FA adds an extra step to logging in. A second **factor**, a second something. And that would be, for example, knowledge, a possession or something from your biology. With **knowledge**, it could be, for example, a PIN or character string or a question that only you know the answer to. With **possession,** you can think of an ATM card or a cell phone. And with **biology**, your fingerprint or shape of the face. With those extra factors in addition to a password, you can then use different types of 2FA. Sometimes it's hardware things, sometimes SMS, push notification, a fingerprint reader, your location, or you have to use an app that generates a number. ![2FA on Facebook](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/03/2FA-Facebook.png) 2FA on Facebook 👾 Next to two factor authentication there is ****Multi Factor Authentication (MFA)**. As the name suggests, 2FA challenges you to prove your identity twice, whereas MFA tests your identity with multiple factors like biometric, face recognition, time, location, and several others. ## Two-factor authentication on an app I was abroad the other day and wanted to log in to a 2FA secured site but unfortunately, no idea why, my SMS was no longer working. So I didn't get a number forwarded and then couldn't log in. I suddenly realized that I was now dependent on a working phone connection. That makes it pretty vulnerable. I made a note to look into 2FA a bit further and put my experiences into an article. I had used a 2FA app in the past, but never fully understood it. This made me feel insecure because if it doesn't work then you lock yourself out. Now I have looked at it in more detail and learned something important: The 2FA works by the server sharing a key once and with time based calculations. There is **no active connection** necessary for generating these numbers! What happens is that the server where you will use 2FA sends you once a string of characters (token) that you store in an app. That app and the clock on your mobile can then be used to generate a number that matches the calculation on the server. That gives you access. So the app is used for storing the tokens. This can then also be done for multiple accounts. Like a wallet with multiple cards. And the app then does the calculations based on time. ![The one time password (OTP) is a form of Two-Factor Authentication(2FA). On its turn, 2FA is a form of Multi-Factor Authentication (MFA).](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/09/MFA-2FA-OTP.png) The one time password (OTP) is a form of Two-Factor Authentication(2FA). On its turn, 2FA is a form of Multi-Factor Authentication (MFA). ## How to put this to work You need to decide on which app you want to use. And with new security comes new responsibilities. There is a risk of losing the tokens. So some providers give you a series of backup codes that you then have to store somewhere. Quite a messy arrangement if you do a lot with 2FA. Another approach is to have secure cloud storage for your tokens. By looking at 2FA more closely, I discovered that [Apple now offers a build in 2FA](https://rob-hoeijmakers.ghost.io/2fa-within-apple-keychain/) that is also stored in the cloud. Having this build into the browser gives convenience. But the popular app [Authy](https://authy.com) for Twilio can do the same, you need to be on Chrome and then there are plugins. Bottom line, whatever app you choose for storing the tokens and generating the one time passwords, it is more resilient and safe than using SMS. It is also more private because you don't have to sprinkle your mobile phone number around. And I truly hope that with this explanation you got a better understanding, and you feel more confident in using the technology. 💡 As a rule of thumb I secure accounts with 2FA that have a financial component or could harm me or my companies reputation. Some examples are Google account, LinkedIn, Facebook, Coinbase, Instagram and Twitter. ### Unraveling Apple's HEIC Format URL: https://rob-hoeijmakers.ghost.io/heic-a-blood-type-for-images/ Last updated: 2026-04-21T09:13:52.000Z An innocent question in a tweet that went viral: I'm sorry but what is .HEIC and why did Apple make it the most difficult photo format. The question got me thinking and after a night's sleep I got the insight that HEIC is like a blood type. Unique, invisible and unimportant most of the time, crucial when you want to exchange. I also recognized that question because if you work with Apple stuff and if you are a producer then sooner or later you are bound to run into this tricky problem and then you are out of rhythm and time is lost. Mainly, I also had no idea what I was doing or needed to do and usually fumbled my way through it. ## Standard within MacOS and iOS Any user of Apple stuff will have come across that format at some point but as a digital producer you definitely come across it. And the question, and the discussion that ensued, gave me **the insight that in Apple's world, the source type of images is always HEIC**. It's been the standard, the blood type, of all images since iOS11. ![Info on image in Apple Photo showing the HEIC file extension and the HEIF format.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/07/DraggedImage.jpg) Info on image in Apple Photo showing the HEIC file extension and the HEIF format. ## About HEIC and HEIF The abbreviation HEIC stands for High Efficiency Image Container. HEIC is the container (container for multiple files such as sound) and the image is stored as HEIF which stands for: **High Efficiency Image Format** It's a modern image file format designed to provide efficient and high-quality compression for images. HEIF is capable of storing not only images but also image sequences and even audio in a single container. It uses advanced compression techniques, often based on the High Efficiency Video Coding (HEVC) standard, to achieve better compression ratios while maintaining image quality. HEIF files can include multiple images with different exposures, making it useful for techniques like [High Dynamic Range (HDR)](https://rob-hoeijmakers.ghost.io/high-dynamic-range-hdr/) photography. This format has gained popularity for its ability to reduce file sizes without compromising image quality. ## Standard on web On the web however, still, the standard for image is the [time-honored JPG](https://rob-hoeijmakers.ghost.io/image-formats-for-the-web/). So if you are a digital producer and publish images or if you want to share images with people who use Windows or Android then somewhere a conversion is going to take place or something is going to go wrong. HEIC is not natively supported on these platforms. ## Conversion The most common place to store photos on the Mac and iPhone is the Apple Photos application. That means that when you share or publish photos on the web somewhere a conversion, a conversion from HEIF to JPG takes place. On the MAC I do that by an export but sometimes by dragging a photo from Apple Photo to the desktop or some other folder. ![Exporting images from Apple Photo to JPG](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/07/DraggedImage-1-1.jpg) Exporting images from Apple Photo to JPG ## iMessage When I get photos sent to me via iMessage they are sometimes coded as HEIC. Until the tweet and this story, I would have first gotten them into Apple Photo and then exported them. I also read about people who would email an image to themselves like that or even take a screenshot and use that further. For the Mac, I found an elegant solution and made a video of how to quickly go from HEIC to JPG. Convert images on Mac from HEIC to JPG 💡 This article made me realize that I should convert my iPhone photo's straight from HEIC to the modern [WebP](https://rob-hoeijmakers.ghost.io/webp-versus-jpg/). So I need to skip JPG all together ideally. ### Review: On writing, a memoir of the craft, Stephen King URL: https://rob-hoeijmakers.ghost.io/on-writing-a-memoir-of-the-craft-stephen-king/ Last updated: 2026-05-14T21:01:44.000Z Without a doubt, this was one of the most surprising and breathtaking books I ever read. It is totally unexpected and takes you down a personal tunnel that makes the messages about the writing stick even better. More I can't tell, otherwise I spoil the story. ## Truly inspiring and a personal favourite This book on writing is not directly related to marketing, but I read it when I was learning about **storytelling**. And I learned a lot from this book, but master King is a harsh teacher: > I am approaching the heart of this book with two theses, both simple. The first is that good writing consists of mastering the fundamentals (vocabulary, grammar, the elements of style) and then filling the third level of your toolbox with the right instruments. The second is that while it is impossible to make a competent writer out of a bad writer, and while it is equally impossible to make a great writer out of a good one, it is possible, with lots of hard work, dedication, and timely help, to make a good writer out of a merely competent one. What King does in this book is to not offer a manual, but a story. He also gives concrete tips and rules, but more valuable is that in his book he takes you along in his work, his personal history and the way he sees his own talent and craft. That I found truly inspiring, and I highly recommend this book to everyone interested in the craft of writing. And as mentioned this book will take you down a deep, surprising and to be honest a scary, painful journey. [The book, On Writing, on the website of Stephen King.](https://stephenking.com/works/nonfiction/on-writing-a-memoir-of-the-craft.html) ## Writing tips from Stephen King Does Stephen King then not give writing advice in this book? Yes, he does. And sometimes really straightforward, like rules and often as the implicit learning of a short story he tells. I found this infographic and this gives you a head start. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/11/IMG_7791.jpeg) 14 writing tips by Stephen King. *The book was originally published in 2000 and the word processor was already replacing the typewriter. King talks about single-spaced pages a couple of times, and he uses it as a measure for progress or volume of work done. It inspired me to research a bit deeper into* [*single-spaced fonts*](https://rob-hoeijmakers.ghost.io/single-spaced-fonts/)*.* ### No code automations URL: https://rob-hoeijmakers.ghost.io/no-code-automations/ Last updated: 2026-04-25T13:17:14.000Z You can buy or rent a website with everything on it as a whole system, but you can also manage it yourself in modules and then hook it together. And that hooking together requires automations because you have to transfer the data and such by hand between them, and that is of course not very scalable and slows down the processes requested by the customer. ## An example I want to hold a live webinar via Zoom and automate the payment, registration and invitation. To do this, I have created a landing page with an order button. ![Landing page](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/02/2022-02-23_11-25-18.jpg) Landing page This is followed by a payment process that asks for the customer's name and email address. After payment, the data goes to my accounting program for the official VAT invoice and the customer is credited in the [CRM system](https://rob-hoeijmakers.ghost.io/customer-relationship-management-crm/) and then added as an invitee to a Zoom meeting. The customer receives an invitation from this. So a landing page in [Leadpages](https://www.leadpages.com), a payment process via [Plug&Pay](https://www.leadpages.com), financial administration in [MoneyBird](https://www.moneybird.com), relation management in [Hubspot](https://www.hubspot.com) and invitation via [Zoom](https://zoom.us). ## Zapier To hook all of this together, we use [Zapier](https://zapier.com). That's a no-code solution that, through webbooks, APIs, filters and the like, and it allows us to connect these stand-alone programs and share the data. An API is access to the program, a webhook is the signal that goes out and offers the data to a different server. Here is the flow, in a Zap, that I described above: ![Automation flow in Zapier](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/02/2022-02-23_11-26-31.jpg) Automation flow in Zapier ## E-mail sequence This is where I will link to [ConvertKit](https://convertkit.com) later. We use this program for marketing and sending programmatic e-mails. These are messages that include conditions such as time or actions and from which the customer can also unsubscribe. Those emails are related to the purchased product and also to possibly other interesting products or events of our company ([graymail](https://rob-hoeijmakers.ghost.io/graymail-what-is-it/)). The e-mail sequence is easily added to the list of automation by extending the existing Zap. ![Adding ConvertKit email sequence in Zapier. Note that many fields are prepopulated.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/02/image-4.png) Adding ConvertKit email sequence in Zapier. Note that many fields are prepopulated. ## Is no-code automation easier? I would say yes, but it does have a learning curve and is not always easy to test. I also think scalability is limited, but time will tell. What it mainly offers us is creativity and a certain freedom when coming up with new products and especially campaigns. It also makes it easier to iterate. To see whether a campaign works or not and then adjust it. But a process also does break more quickly, you depend more on others, and then you have to open all kinds of drawers and boxes in Zapier of which you no longer have any idea. That is sometimes more laborious than writing, managing and documenting code, but it might be a lack of experience. So I will be playing around with no code and low code and keep you updated. Also, I will be on the lookout for agencies to whom you can outsource certain requests and explore to see if no code automations are scalable. ## Read more [Why Automation with Make.com Made Me So HappyMake.com transformed my workflow—what used to take 40 minutes now runs in seconds. Cloud automation has never felt this creative, flexible, and fun!![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-55.png)Rob HoeijmakersRob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_1194-1.jpeg)](https://rob-hoeijmakers.ghost.io/make-dot-com-automation/) [LLM-Driven Automation and Anthropic’s Computer UseAnthropic’s Computer Use, make.com, and Automator: See how LLMs simplify automation, analyze tasks, and redefine productivity. Watch examples and learn more!![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/icon/Circle-logo-2-13.png)Rob Hoeijmakers (robhoeij)Rob Hoeijmakers![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/thumbnail/IMG_9049-2.jpeg)](https://rob-hoeijmakers.ghost.io/llm-driven-automation-and-anthropics-computer-use/) Valuable video course on Webhooks. Webhooks follow the Hollywood principle: we call you, don't call us. Webhook for beginners ### Building a blog audience (2022) URL: https://rob-hoeijmakers.ghost.io/building-an-audience-for-my-blog/ Last updated: 2026-04-25T13:17:14.000Z For a little over a year, I have been writing this blog. When I started, it was more for myself. To organize my thoughts and gather and hone existing expert knowledge about production on the web. And to share my photos. Creating a website itself was also part of the process, and I had fun working on it. Later, the blog also became a tool to acquire and sharpen new knowledge about writing and online marketing. Fifteen years ago I also had a blog, in the wake (Kielzog) of daily life, and actually everything I wrote then ranked high in Google, and so I got quite an audience. Fleeting though because I didn't do anything about engaging those people. So no newsletter or anything of the kind. And it was in the days of the first generation of the web and social media was not there yet or much smaller presence. In that sense, this is kind of a "grandpa talks.” But I am looking for modern ways to reach and build a larger audience. ## And what about my audience now? Search engine optimization is still very relevant, and I pay a lot of attention to that. I also do "cross publishing" and have become much more present and active on social media. ### Search engine optimization (SEO) One of my biggest hurdles, and frankly annoyances, was that my newly chosen domain (robhoeijmakers.com) was somehow poorly or not picked up by Google. And this despite all the possible [optimizations](https://rob-hoeijmakers.ghost.io/my-basis-for-seo/) I know of. If you [searched for my name](https://www.google.com/search?q=rob+hoeijmakers&sxsrf=APq-WBsFjJK5ceRtzAkLI4LpSuPGit1Vlw%3A1645862450553&source=hp&ei=Mt4ZYtygH67jkgXkiYnICA&oq=&gs%5Flcp=ChFtb2JpbGUtZ3dzLXdpei1ocBABGAAyBwgjEOoCECcyBwgjEOoCECcyBwgjEOoCECcyBwgjEOoCECcyBwgjEOoCECcyBwgjEOoCECcyBwgjEOoCECcyBwgjEOoCECcyBwgjEOoCECcyBwgjEOoCECcyBwgjEOoCECcyBwgjEOoCECcyBwgjEOoCECcyBwgjEOoCECcyBwgjEOoCECdQAFgAYLQNaAFwAHgAgAEAiAEAkgEAmAEAsAEP&sclient=mobile-gws-wiz-hp) in Google, this domain structurally did not appear. I submitted the matter to Google, but didn't get a real answer. Someone told me to read the manual, and that hurt a bit. Then I picked up my old domain again, and it was a good grab. Now my site appears in the top 5 if you search for my name, and I'm very happy about that. I am positive more traffic will start to come out of Google. ![The SERP of my homepage.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/02/2022-02-20_11-53-10.jpg) The SERP of my homepage. ### Republishing That is publishing or cross publishing articles also on platforms like Medium, Substack and Hashnode. I do that too, though selectively. Whatever is most appropriate for that specific platform. That's where it gets read, but in my case not spectacularly. Having your articles on these platforms extends visibility, but it is work to do it properly. I will maintain doing this for the coming months, although it is not always on top of mind. My main goal is building [my own domain and content](https://rob-hoeijmakers.ghost.io/ghost-blog/), not to get dependent on a platform. ### Social media I also collect email addresses through my blog in Ghost and as soon as I publish an article the subscribers receive that article via email. But other than my benevolent family and friends, no one seems to be subscribing. (There are now 12 of you). Not even when I put out a post on my social media like [Twitter](https://twitter.com/RobertHoeijmak1) (1000+ followers), [LinkedIn](https://www.linkedin.com/in/rob-hoeijmakers/) (1000+) or [Facebook](https://www.facebook.com/hoeijmakers.rob). Either it doesn't stand out enough or people don't want it. The information spaces we are nowadays are of course crowded and loud, that finding an audience requires patience as well. ## Newest attempt: Revue I write English language tweets and to this account I have now linked Revue. That's a newsletter service from Twitter. Creating a newsletter or digest works great I must say and collecting addresses is very conveniently done by linking it to your Twitter profile. I now have two subscriptions and both are myself 😂 *.* But all the effort I already put in Twitter I hope to be able to use for my blog as well. Let's see. I decided not to add existing subscribers to that list because I want to get to know the system a little better first, but of course I would be delighted when you subscribe: ## Aim Where do I aim for with my blog? The main goal is to be curious and share what I can. In terms of numbers, I aim for **a minimum of 1000 monthly unique visitors** within a year from now. In a follow-up article, I will show a bit more of the features of Revue and let you know how Revue is performing for my goals. ### The Attraction of a Lead Magnet URL: https://rob-hoeijmakers.ghost.io/lead-magnet/ Last updated: 2026-04-25T14:49:38.000Z A "lead" is something that starts something. Take the lead. In online marketing, a lead is a person or organization interested in your activities. By calling you, emailing you or filling out a form on your website, for example. After that, the lead is registered on a list with a basic set of information like name and email address. Of course, the quality and degree of interest can vary greatly, but this is a lead. A starting point for business. [](https://ghost.org/resources/content/images/size/w1600/2023/09/64e2fbd9e35161c9de1fc1a7%5Fa38374-80d5-0dba-a83e-4d00040cb05%5F75d11486813d79be786ad91b2cda167d5fb116a2-1655x1543%5F1%5F.webp) ![Lead Magnets, engagement and ease of consumption](https://ghost.org/resources/content/images/size/w1600/2023/09/64e2fbd9e35161c9de1fc1a7_a38374-80d5-0dba-a83e-4d00040cb05_75d11486813d79be786ad91b2cda167d5fb116a2-1655x1543_1_.webp) [Source](https://ghost.org/resources/r/220f1fa5?m=357cb241-722d-4146-a852-ae88296f982b) ## How to attract? Once upon a time, the Internet was a place where people seemed to be actively looking to fill out a form. That has changed quite a bit. Now only bots do this for fun and I as a business owner and producer have to be much more active in looking for leads. I have to try to attract leads to me. Like with a magnet. Welcome to the stage: **The Lead Magnet**. ## Lead magnet This magnet is an attractive and free online offer that I make in exchange for some of the lead's information. A giveaway, a freebie. This can be an e-book, a survey, a song, a spoken meditation, a podcast. Something I usually advertise on the website. On a specific page or between the lines. Of course, I can then also advertise my lead magnet through a post on social media or make the freebie available directly within the posts. And if I'm looking for a lot of leads or my website and social media just don't have enough traffic, I can also buy ads for more traffic. > Lead magnets are not just about collecting emails. They’re about providing genuine value, building trust, and paving the way for creating a customer relationship with your audience. Remember - a lead magnet is often the very first interaction a potential customer has with you and your business. Don’t make it feel like you’re “stealing” their email address. Make it easy. Make it valuable. Make it for the long term. – Justin Welsh Practical example of a lead magnet Let me give my wife and business partner's website as an example. She runs a coaching practice for business clients and for visitors to the site, she gives away an e-book. The website was created in [WordPress](https://rob-hoeijmakers.ghost.io/how-about-wordpress/) and the Lead Magnet in ConvertKit. That programme is specifically aimed at converting a visitor to a lead. ![Example of a Lead Magnet in WordPress with ConvertKit.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/09/lead-magnet-2.jpg) Example of a Lead Magnet in WordPress with ConvertKit. The lead fills this out and then receives an e-mail with a download link to the e-book. And we get the data via email and in they are in then also in ConvertKit's system. The magnet has done its job. The lead has a wonderful, free gift and we have enough data for the next step. See also the overview of [landing page tools](https://rob-hoeijmakers.ghost.io/tools-for-creating-landing-pages-and-lead-magnets/) and lead magnets with lead magnet options within Wix, Squarespace, ConvertKit and Leadpages, among others. ![Behind the gate? The wine fields at the lake of Biel.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/09/136_file-2.jpg) Behind the gate? The wine fields at the lake of Biel. ### Electronic paper and digital ink explained URL: https://rob-hoeijmakers.ghost.io/electronic-paper-and-digital-ink/ Last updated: 2026-04-25T08:30:40.000Z I visited my mother-in-law the other day, and she is very sceptical of modern, digital technologies. Partly she distrusts the physical properties that may cause negative radiation, but also the hypnotic effects of many screens. But my e-reader fascinated her immensely. So compact, long battery life, versatile, but I could only moderately explain what makes the screen so special. A new attempt. ## Electronic paper The electronic paper of an e-reader does not emit light, but works just like a paper book. It receives light from its surroundings and reflects it. So the light and dark parts of a page are physically present and therefore so sharp and quiet for the eyes. No flickering caused by the frequencies of refreshment. But an e-reader also emits light, right? I myself like to read in the dark and so use the e-reader as a sort of torch. But that light does not come from the screen, but they are LED lights in the side of the screen. So no light comes out of the screen, but the screen is illuminated. 0:00 /0:09 1× E-reader in the sunlight ## Workings of digital ink The screen of an e-reader is a grid, just like a normal screen of a computer or mobile. But the small elements in the grid become black, gray or white. An electric current causes the pigments in the tiny sphere to swap places. ![E-ink - How it works](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/02/image.png) E-ink - How it works This is also the reason that after the page is built, the e-reader no longer uses power and the e-reader just shows a page or image even if the device is turned off. With a difficult word, the spheres are bistable. ## Benefits So the page of an e-reader is physically present and has a high contrast. With a magnifying glass, or better a microscope, you would be able to see the white and the black. That's also the reason you can read in sunlight and at a reasonably wide viewing angle. Next to that, electronic paper and digital ink are low power, very compact and durable. I myself have a Kobo Clara and the screen has a [resolution](https://rob-hoeijmakers.ghost.io/resolutions-for-screens-and-images/) of 1448 x 1072 pixels. That means that it has one and a half million pixels over a six-inch screen. In practice, it means you don't see the actuals pixels and the fonts really look smooth. ![Kobo Clara, magnification](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/07/eink-you.jpg) Kobo Clara, magnification Want to see more, check out this video explanation and tour: How E-Ink works ### Digging into Web3 URL: https://rob-hoeijmakers.ghost.io/web3/ Last updated: 2026-04-21T09:25:25.000Z Just a quick update. I'm busy going through a course called [Solidity](https://soliditylang.org) in addition to my work as a [producer](https://soulwebacademy.com) and [entrepreneur](https://soulweb.nl). That is a programming language that you use to create Smart Contracts. Such a contract is an algorithmic progression of agreements and is placed on a blockchain as a so-called DApp (decentralized app). Chances are, I've already lost you? ## What is Web3? Web3 is not so much a fact yet but more of a vision and perhaps best explained by looking at what Web1 and Web2 entailed. **Web1** was a generation of the web where we all happily created a website and learned about emailing. At its core, it was an almost entirely public good, comparable to clean air, schooling, or infrastructure investment. > Economists formally think of public goods as ones which are non-rivalrous (meaning we can both consume them) and non-excludable (meaning I can't stop you from using them). **Web2** is a generation of the web dominated by large, private companies like Google, Amazon, Twitter and Facebook with platforms of a much more closed nature. And the vast majority of value accrues to the platform itself rather than to the users. **Web3** is more or less a reaction to the problems of the current web. No privacy, limited openness. Web3 is shaped by decentralization and a crypto token-based network. ![History of the web. Image: moralis.io](https://moralis.io/wp-content/uploads/2021/06/Blog-Web1-Web2-Web3-1024x710.jpg) History of the web. Image: moralis.io ## Where Web3 is going These are still big chunks to me, and it is certainly not yet clear exactly what Web3 will mean. There's debate whether it's going to be a kind of libertine Wild West by design, or whether something of a social or idealistic system is actually going to be possible. For now, it's what you hope to see. > There will be centralized services in web3 just as there were in web1\. The key question in web3 is whether the network effects accrue as private goods (as they did in web2) or public goods (as they did in web1). > > — cdixon.eth (@cdixon) [January 23, 2022](https://twitter.com/cdixon/status/1485303907949727749?ref%5Fsrc=twsrc%5Etfw) > A number of people have asked me why I get so bothered by crypto/blockchain/NFTS/Web3 when no one is making me participate against my will. It has come up enough times that I thought it merited a response. A thread. > > — neil turkewitz (@neilturkewitz) [February 16, 2022](https://twitter.com/neilturkewitz/status/1493967940273770502?ref%5Fsrc=twsrc%5Etfw) I want to get a better handle on it in the near future, by working on it even more practically, and write about it with enthusiasm because I certainly find it exciting, and I hope you have or get interested in the idea of Web3 as well. I started on the wrong foot by confusing Web3 with Web 3.0, the [semantic web](https://rob-hoeijmakers.ghost.io/semantic-web/), and I am happy to be on the right track again. ### Review: Building a StoryBrand, Donald Miller URL: https://rob-hoeijmakers.ghost.io/review-building-a-story-brand-donald-miller/ Last updated: 2026-05-14T21:06:27.000Z Finding a good way to discuss your product or service can be challenging if you don’t have the experience. To then be guided with a framework can be really helpful. Is this book by Donald Miller a good guide? ## Let the book speak for itself Every day, most business leaders make a mistake that costs them money: they don't explain clearly what it is their company does. The **StoryBrand Framework** is a paradigm-shifting approach to connecting with customers that provides the ultimate competitive advantage, revealing the secret to helping customers understand the benefits of using your products, ideas, or services. Building a StoryBrand teaches: - the seven universal story points all humans respond to - the real reason customers make purchases - how to simplify a brand message, so people understand and act on it - how to create the most effective messaging for websites, brochures, and social media Building a StoryBrand will forever transform the way you discuss who you are, what you do, and the unique value you bring to your customers. > Clarify your message so customers will listen. ## What did StoryBrand do for me I read the book carefully and used the framework that goes with it. Used it successfully to craft a campaign for a client. It gave me good insights in what storytelling is, and Miller is generous with tips and advice. Because I was more or less a novice to storytelling, the framework also helped me get started and follow the right steps. One of my main takeaways was to see branding and the story around a product from the perspective of the prospect rather than my story. To see the customer as the hero and myself (or the client) as a guide. **Do I recommend the book?** I do. Especially for people or entrepreneurs who are new to storytelling and feel they need some guiding and protective rails to navigate the challenge of telling a convincing and exciting story. ## The SB7 framework In the book, Miller talks about the SB7 framework. Those are seven principles or elements you can identify: 1. A character 2. Has a problem 3. And meets a guide 4. Who gives them a plan 5. And Calls Them to Action 6. That Helps Them Avoid Failure 7. And ends in success ![Storybrand, SB7 Framework, example journey](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/10/IMG_5903.jpeg) Storybrand, SB7 Framework, example journey With these steps, you can start clarifying your message and work it into a story with the provided tool called the StoryBrand BrandScript. Once this stage is set, the book takes you by the hand and talks you through all the seven elements. Again, for me, it really worked, and it got me going quickly. ![Storybrand, SB7 Framework, downloadable framework](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/10/IMG_5904.jpeg) Storybrand, SB7 Framework, downloadable framework The framework can be downloaded (with providing data) for free here: [StoryBrand![](https://www.mystorybrand.com/favicon.ico)![](https://www.mystorybrand.com/fb-share-logo.jpg)](https://www.mystorybrand.com) ### Portrait versus landscape images on social media URL: https://rob-hoeijmakers.ghost.io/portrait-and-landscape-images-on-social-media/ Last updated: 2026-04-24T18:31:29.000Z **Updated with Threads, July 2024** Are you familiar with the expression a bed of Procrustes? Procrustes was the nickname of an innkeeper among the Ancient Greeks. Procrustes invited passing travellers to spend the night with him. If they accepted his invitation, and went to rest, Procrustes came to see if his guest would fit in the bed. Usually they didn't. Now if a guest was too short, he stretched out his limbs by force; if he was too tall, the host chopped off a piece. Typically, this treatment ended in death, allowing Procrustes to loot the traveller's possessions. It works exactly the same way with images that do **not fit into the requested format of an app**. In this article, I will go down the road of wisdom and provide some practical questions and answers to the question: **When to use what format on social media?** ![Examples. Legs and heads chopped off. Automatic resizing at work.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/07/image.jpg) Legs and heads chopped off. Automatic resizing at work. ## Topic or subject of image Sometimes the topic determines the best choice, and sometimes the medium. What I mean by this is that if you take your own photos with a cell phone, you actually have to make a choice every time how you are going to hold the device: portrait or landscape. ### Landscape photos Subjects that call for **landscape photos** are landscapes, cars, bridges, groups of people, soccer fields and other subject with horizontal, wide-angle images. The great thing is that there are exceptions to all of these rules. ### Portrait photos Subjects suitable for **portrait photos** are portraits, vases, buildings, a full person. Anything that works vertically. Again, there are exceptions that can also work out well artistically. ![Japanese artwork showing a landscape in portrait format](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/07/image-6.jpg) Japanese artwork showing a landscape in portrait format ## Device and Platform The question of portrait or landscape can also be answered by looking at the medium where the image will be placed. Will the image be on a landscape screen such as television, computer or on a portrait screen such as the mobile phone or e-reader? In recent years there has been a huge shift to mobile and therefore to the portrait format. Which social media platform is the image going to be in? Twitter, Facebook, Instagram and or Pinterest? Is the image on its own or in a series, a gallery with other images. And increasingly, does an image need to work well both in a portrait and landscape medium, like most websites today. I'll provide an answer for common social media channels because you often take those photos in passing. 🌅 I assume standard formats for landscape and portrait. If you want to learn more about the different size and relation of the sides, [check out my article on aspect ratio](https://rob-hoeijmakers.ghost.io/aspect-ratio/). ## Answers for Social Media Let quickly take you through the most common social media - Instagram, Facebook and Twitter - and tell you what happens to images with single pictures per post and multiple pictures per post. ### Instagram Instagram is actually a good start because immediately a special exception because most images on Instagram are square. With a single photo, you get the choice to volunteer your photo on Procrustes' bed. With the advantage that you can see what it will look like. After squaring up, you can then quickly adjust the crop and shift the focus. ![Making images square on Instagram](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/12/image-5.png) Preview making an image square with Instagram In a series of photos, you can choose landscape or portrait. This applies to the entire series. So there will be edits, and you can lose head or feet in the bed of Procrustes. In a series, you can still make adjustments and indicate what is best to be cut off. Head or feet. You do this by placing a finger on the image and then moving the image around. ### Facebook Also in this medium, there is a clear distinction between a single photo or a series, a gallery. With a single photo, either a portrait or landscape image can be used and nothing will be cut. If you attach multiple photos to a post, Facebook will turn it into a gallery with all square images. Again, some will be cut on the bed of Procrustes. In the standard gallery, the bottom part of a standing photo is cut off. If this produces a lot of deformation, you can opt for a vertically oriented gallery, for example: ![Facebook, a gallery example suited for vertical objects.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/07/image-3.jpg) Facebook, a gallery example suited for vertical objects. From landscape images, Facebook cuts off the sides in a gallery and focuses on the middle part. So compositions with the emphasis on one side lose a lot here. ### Twitter (X) On Twitter (X, you can recently use portrait and landscape images without ending up on the sore bed. Twitter will mainly be viewed on mobile, and so a **portrait image works well** then. When selecting a post with portrait images, a large part of the screen is filled. On Twitter, you can add up to four images to a single post. The images in the Twitter gallery with four images are displayed horizontally. So, portrait images are automatically cropped and this can lead to unwanted results. Especially with portrait images. ![Twitter image galeries examples](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/12/image-4.png) Twitter image galeries ### Threads Threads, a social media platform launched by Meta, offers a distinctive way to scroll through photos, accommodating both portrait and landscape orientations seamlessly. Users can view a series of photos in their original aspect ratios, avoiding the typical cropping issues seen on other platforms. This feature enhances the visual storytelling experience by preserving the integrity of the images, whether they are tall portraits or wide landscapes. Threads’ design optimally adjusts the display to ensure a smooth and visually appealing browsing experience across diverse photo formats. ![Threads with example of slider.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/07/Threads-orientation.jpeg) Threads with example of slider. ### TikTok TikTok is definitely mobile first, and this means all content is in portrait mode. So when you repurpose old video material, you probably end up zooming in or use the dual screen mode. ![TikTok screen examples](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/11/greenscreen.png.webp) TikTok screen examples ### Portrait is preferred In my personal experience and from what I read from others the portrait format is scoring best on social media. Especially when single photos are shown. Make sure to keep it in mind already with planning and making the photograph, rather then change format later. So enough image horror stories, quickly move on and leave the inn and bed of Procrustes behind. --- ### Tip you can use an app to add colour palettes to your pictures. This stands out but also switched portrait or landscape in a square picture. [The app is called Coolors ](https://rob-hoeijmakers.ghost.io/coolors-color-palette/) ![Compensate for odd aspect ratio ](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/10/IMG_6800.jpeg) Compensate for odd aspect ratio ### Graymail, what is it? URL: https://rob-hoeijmakers.ghost.io/graymail-what-is-it/ Last updated: 2026-04-25T13:16:03.000Z Graymail is about emails that form a gray area between wanted and unwanted messages. It's an interesting and handy concept for the reader, and certainly for the sender of emails. So what is graymail? Graymail is a term used to describe **emails that are not outright spam, but are not necessarily wanted or expected by the recipient**. Graymail can include newsletters, promotional emails, or updates from social media sites, and is often seen as a nuisance by users. ## Graymail is not spam The meaning of spam is probably known by everyone. Spam is not in a gray area. Unsolicited emails that bombard you with vague promises and offers. Most email software providers already know how to deal with it quite well, and this scourge seems to be under control. > Graymail is bulk email that does not fit the definition of spam because it is solicited, comes from a legitimate source, and has varying value to different recipients. ![Email, graymail and spam on a spectrum.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/11/DraggedImage-2.png) Email, graymail and spam on a spectrum. Graymail differs from spam in that it is typically sent by legitimate organizations and is not necessarily malicious in nature. However, it can still be **a source of frustration for users** who feel overwhelmed by the volume of emails they receive and may not have the time or interest to read all of them. ## Graymail is email that you open sometimes Graymail have in common with spam that they are sent in bulk. But these e-mails form a gray area between wanted and unwanted. You've signed up for one thing or another at some point, consciously and often unconsciously. You ask for a free e-book and with this you get the marketing emails or newsletter of a company or creator. But is also caused to receiving what feels like a never ending stream of messages. Think of LinkedIn updates, security messages, Instagram messages, notifications on delivery status, terms and conditions updates etc. etc. Nothing wrong with that per se, it's an exchange in the attention economy, and you might want to dip into the stream of information at some point in time again. But the engagement on those emails is very low, and they can clutter up your inbox. ## Google Gmail and Graymail To manage graymail, many email providers offer options for filtering or organizing emails into separate folders, so that users can more easily prioritize and manage their inboxes. Some users may choose to unsubscribe from newsletters or promotional emails that they no longer wish to receive, or to set up rules or filters to automatically sort these types of emails into a separate folder. Google's Gmail tries to address Graymail by applying filters for it. As a result, they don't show up in your inbox right away. ![Graymail filter in Gmail](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/11/Gmail-graymail.jpg) Graymail filter in Gmail ## Apple's Mail Privacy Protection and graymail Apple, the largest provider of email clients, has started to shield interactions on emails by what they called **Mail Privacy Protection (MPP)**. With this protection enabled, no longer information about the reader goes out when opening an email. About [50% of all emails are opened with an Apple client](https://www.litmus.com/blog/email-client-market-share-2021-q1/) and so this change has caused quite a stir among email marketers. The privacy measures have made it a lot harder to say for sure whether readers are opening or interacting with a received message. In that sense, the gray area for senders has just become a lot more gray. ## Tips to prevent being graymail Here are some tips to help prevent emails from being seen as graymail: 1. Make sure that you have **permission to send emails** to the recipient. If you are sending newsletters or promotional emails, make sure that the recipient has opted in to receive them. 2. Use **a clear and descriptive subject line** that accurately reflects the content of the email. This will help the recipient understand what the email is about and whether they are interested in reading it. 3. Include **a clear and concise message that is relevant to the recipient**. Avoid sending emails that are overly long or contain too much information, as this can be overwhelming and lead to the email being marked as graymail. 4. Offer **an easy way for the recipient to unsubscribe** from future emails. This will allow them to opt out of receiving emails from you if they no longer wish to receive them. 5. **Use segmentation** to send targeted emails to specific groups of recipients. This will help ensure that the emails you send are more relevant and engaging to the recipient. 6. **Monitor the response rate** to your emails and make adjustments as needed. If you notice that your emails are being marked as graymail or are not being opened, it may be a sign that your messaging is not resonating with your audience, and you may need to make changes to your email strategy. ### Mint your first Non-Fungible Token (NFT) URL: https://rob-hoeijmakers.ghost.io/my-first-nft-non-fungible-token/ Last updated: 2026-04-14T13:23:11.000Z NFT’s are pretty overwhelming if you are outside to the crypto crowd. I wrote a practical step by step record of how I minted my first NFT. A **Non-Fungible Token (NFT)** is a registered ownership of a digital artifact. On a blockchain, or decentralized records. So no one is really in charge of this administration in this way. You register a digital artifact there - a photograph, an image, an animation - and then register this work. You mint a certificate. That certificate (token) can then be traded. All rather abstract. When NFTs became known and the first items were sold for outlandish prices I had a rough idea what they were. I was still new to the world of crypto currencies and blockchain so I was very curious to purchase or offer an NFT. At first I failed and now I've given it another try. I'm taking you on this little journey. A practical, hands-on journey. After this short story, you'll know what it takes to create an NFT. ### Prerequisites What you need in short is a marketplace for of NFT services and a so-called Wallet. I searched briefly and chose the (non-curated) marketplace that calls itself the largest. What I was looking for was something with a low entry level and the highest possible ease of use. It became [opensea.io](https://opensea.io), the choice was still random and it is also not that I recommend this platform directly. For that I will do further research. ### Create the NFT The first thing I noticed is that I did not have to create an account. The Wallet served as my account. Like logging in with a bank account. Opensea has a list of Wallets you can choose from. So not every Wallet is suitable. I have a wallet at [Coinbase](https://coinbase.com) and linked it to OpenSea. Later I was able to create a profile though to make it look more personal. After linking Wallet to Opensea, an easy process I went through on the mobile phone, I was able to offer a digital artwork from the Wallet. I chose a picture of a mushroom I had taken, named it **Amanita**, and then it was on the NFT marketplace and a token was provided to my Wallet. ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/11/DraggedImage-1.png) ### Share your NFT Once you have a token you can promote or market it easily. I can share it through an address and send it to Social Media or through messages with a link: There is also a bidder but the honesty is that this is from an acquaintance and crypto enthusiast. I didn't accept the offer because there are the necessary fees for the transaction. ### Possibilities for NFT Creating the NFT was interesting to do and I learned from it. By doing, things become more tangible. There are numerous possibilities with NFTs. An interesting use case is that it can serve as a substitute or complement to current [copyright](https://www.nftlicense.org\). It appeals to the urge to collect. I myself have collected a lot in my life: soccer cards, coins, stamps, rare books etc. And I know how it can feel when a collection is not complete. Collecting is a purpose in its own. The NFT's also offers a great platform to the digital artist or craftsman. For example, you can set up a gallery to show and sell your work. [![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/11/_QwAZV4a.jpeg)](https://art.0xbanana.com/#) Art by 0xbanana.com ### The word non-fungible Oh yes, English is not my first language and I did not know the word "Fungible". Did you? The word "[fungible](https://www.etymonline.com/word/fungible#)" means something like "capable of being used in place of another; capable of being replaced". And so non-fungible is non-substitutable, unique. A definition of an NFT then is: An association (a kind of ownership) on a blockchain of an account to a unique (non-fungible) digital item via a smart contract. NFT is a way of linking ownership to digital objects. However, "fungible", exchangeable, are money, physical and digital. ## related links - [CryptoPunks](https://www.larvalabs.com/cryptopunks) - [CryptoPunks collection opensea.io](https://opensea.io/collection/cryptopunks) - [NFTs - all you need to know to sound smart on the internet](https://blog.mcgee.cat/nfts-all-you-need-to-know-to-sound-smart-on-the-internet) ### Google AMP is dead URL: https://rob-hoeijmakers.ghost.io/google-amp-is-dead/ Last updated: 2026-04-25T13:16:38.000Z No one likes to eat at the same table with someone who is known for stealing from your plate. And that was exactly what was happening with Google AMP. A short obituary. ## Google AMP Google AMP is an abbreviation for **Accelerated Mobile Pages**. The aim of this technology is to provide a slimmed down version of a webpage that shows up quickly on a mobile phone. A device that in the past had lower computing power and probably a slow connection. Google lured providers of digital content with possible higher ranking on AMP. A sort of preferential treatment. With the trade off that your highly valued content would like all the rest. ## Switch off Google AMP Today I read an [article on the blog of Plausible](https://plausible.io/blog/google-amp) that Google stopped the preferential treatment in their search results and that AMP is basically dead. This news made me feel happy; I always disliked the idea that a Tech Giant decides how my content should be presented. And if not, I will suffer at their hands. > AMP is no longer a Google requirement to create a fast-loading website. You can develop [a lightweight site](https://plausible.io/lightweight-web-analytics) with a great page experience and top Core Web Vitals scores even without using the restricted and Google-controlled way of building a website. So a small victory for design freedom and I actively switched off the AMP-feature. Less testing, fewer worries of how readers experience my stories. No more stealing from my plate. ### Monospaced fonts: From Typewriter to the Matrix URL: https://rob-hoeijmakers.ghost.io/single-spaced-fonts/ Last updated: 2026-05-10T11:23:12.000Z I was reading Stephen King’s book on writing, and in this fantastic book he writes about **single-spaced fonts** a couple of times. He talks about these fonts as a clear measurement for productivity, and I started wondering what those single space fonts are. > *Although no copies of that particular masterpiece survive (at least to my knowledge), I believe it was eight pages long, each page single-spaced and paragraph breaks kept to an absolute minimum (each stencil cost nineteen cents, remember).* Stephen King, On Writing: A Memoir of the Craft ## Monospaced fonts Turns out that he is from an era where the type machine was the norm when writing. These mechanical machines use the energy of the button press to stamp a letter or symbol. You push a button and an iron cast of the letter flies to the paper. Just before impact a lint with ink is pushed between the letter and the paper, TIK, the letter appears on the paper. ![Typewriter close up. Credit Bob Shand. ](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/10/6127766837_46596e4948_k.jpg) Typewriter close up. Credit[ Bob Shand](https://www.flickr.com/photos/bobthelomond/). All these letters and symbols use up the same width. And so does a space or for instance a small glyph like a dot or a comma. All the letters, word, lines line up on a matrix. So now you can count how much characters go on a sheet of paper. It is an analogue form of count. Not very precise per sheet, but it probably evens out pretty good when there’s a stack of paper. So if you are writing a novel or a script, you can estimate the number of pages you need to aim for. The fonts used for this are referred to as single space fonts or monospaced fonts. There are multiple types and sizes that go with the typewriter. Most used is the [Courier](https://en.wikipedia.org/wiki/Courier%5F%28typeface%29). ![Type writer font.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/10/unnamed-2.png) Type writer font. ## Early computer font When I was reading about the typewriter and its fonts, it suddenly occurred to me that in the early computing days the same sort of approach was taken. On the mono coloured screen, there was a so-called cursor blinking. An indication where text would appear when you typed. That cursor showed up as a little stripe or rectangular block. Always the same width. The font used was also monospaced, as with the typewriter. It filled a matrix. And so, for a long time, it also appeared on the paper when printed. ![Output Apple-1 display](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/10/0_gsJ0uPDpBBjz2Ove.gif) Output Apple-1 display Programmers still often write computer code with editors that use a monospaced font like Courier. I'm not sure why that is, maybe out of tradition. Or maybe because the number of letters or lines give an estimate of file sizes. These editing programs easily could display disproportional fonts like the ones you are reading this article with. > Both screenplays and stage play scripts frequently use monospaced fonts, to make it easier to judge the time a script will last for from the number of pages. The industry standard is 12 point Courier. A tradition holds that on this format one page of script will take one minute of screen or stage time. > Wikipedia, [Monospaced font](https://en.wikipedia.org/wiki/Monospaced%5Ffont) ## Matrix, the green letter tears The word matrix now is associated easily with the film, and then it struck me that the iconic green cyphers they use are monospaced and that is the reason they can droop down so nicely in green lines. The letters of the matrix are all on a matrix. A digital rain. The single-spaced fonts from Stephen King’s typewriter took me to the early computer fonts. These fonts are placed on an invisible matrix, and the film graphics uses this matrix to show the patterns. --- ### Further reading - [Typeface: to serif or not to serif](https://rob-hoeijmakers.ghost.io/typeface-to-serif-or-not-to-serif/) - [Typography insights](https://rob-hoeijmakers.ghost.io/typography-insights/) - [Typography, in motion](https://rob-hoeijmakers.ghost.io/typography-in-motion/) ### Site Search explained URL: https://rob-hoeijmakers.ghost.io/website-search/ Last updated: 2026-04-25T13:17:06.000Z You go into the woods to collect them, which can be full of relaxation and anticipation. You take what you find home and prepare for conservation. And in the third phase you get to eat what was in store. All three phases are important and can be professional fun or pain. Let me take you through the steps. ## Collecting the data You need to go out into the data space, which in this story is a website, to collect the data you want to index. This typically is done by a (ro)bot or crawler that starts somewhere on your indication and follows the links to discover the space and pick up the data. Another possibility is to use an API to collect all the information from the backend of your website. Picture a forest and you need to find the mushrooms. Sometimes it’s more like harvesting, and sometimes it is more like hard work. Ditches, slippery branches, wild animals, insects. It all depends on how well-structured the pages are and if they are properly connected. In this first phase, you typically give instructions and limits to the crawler. Start search here and here, don’t go beyond the forest or a certain path, don’t pick up blue mushrooms, watch the bears and blow the horn once done. ![Clean the data up like these mushrooms. Photo Rob Hoeijmakers.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/10/IMG_1184.jpeg) Clean the data up like these mushrooms. Photo Rob Hoeijmakers. ## Preparing and improving the data Once you have retrieved the data, you can enrich it. Let’s say you took the mushrooms home, you want to clean them up (some with water, other with a brush), sort them a bit (insects) and maybe get rid of some you might suspect to be poisonous. Preparing the data you have might involve a number of processes. One is that you can work with semantics. You can use synonyms and hypernyms to give more meaning to what you have. You can check for words that belong together and mark them as such. It might be necessary to split data based on the language it is in. And then there are tasks that might be very specific for the recipe you have in mind. Product names, people, categories might need to be marked. Once all this work on the data is done, it is put in an index for fast retrieval. A bit like the index on a book. ![Prepare for conservation or consumption. Photo Rob Hoeijmakers.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/12/IMG_2759.jpeg) Prepare for conservation or consumption. Photo Rob Hoeijmakers. ## Find results The third step is to offer a search box and page to do the actual searching. In the simplest variant, it is just entering search words and see a list of results. But you can go much further, and there are a number of possibilities to improve the experience of searching. You might want to have filters or facets, autosuggest, spelling corrections, highlight of results in context, recommendations or push certain results you deem more valuable. Also, here Natural Language Processing (NLP) might come in, and the person searching can type in daily language to search. Basically this third step is the goal and if the first two are done properly the visitor should be able to get what is necessary and maybe even be inspired to look further because of recommendations. It depends on how well the interactions are shown and working, if there is an enticing User Experience. ## The site search on this website In the past, I mostly worked with Google Search Application and a solution based on [Elasticsearch](https://www.elastic.co). For this website, I started with an API version that came with the template I use (Liebling for Ghost). That extracted data on articles and offered a simple search. This turned out not to be effective and sometimes plainly wrong. Plus, it was also not full text search. So I searched around and gave the free version of search engine [SiteSearch 360](https://www.sitesearch360.com) a spin. They crawled my site and all articles full text and with some iterations the crawler found them all and put the data in an index. Then I was able to use a dictionary and some other tricks to make the data even more useful. In order to serve up a search box and page, they provided me with an easy to implement solution. Now the search shows dynamic results and with a nice preview of the pages found. In my account, I can see some basic and valuable analytics of where visitors searched for and if they were successful. So for now I loved it! Really well done and very extensive. ![Enjoy the results. Photo Rob Hoeijmakers.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/10/IMG_2760-1.jpeg) Enjoy the results. Photo Rob Hoeijmakers. 💡 As of mid 2022 Ghost has [Site Search](https://ghost.org/changelog/search/). This functionality is easy to set up and fits nicely in the templates. On the downside, the capabilities are rather limited and it does not do full text search. ### Further reading on Site Search Algolia is a provider of site search capabilities and can be considered a market leader. Ivana Ivanovic, Senior Content Strategist of Algolia, wrote [this guide on site search](https://www.algolia.com/blog/product/the-ultimate-guide-to-site-search/?ref=hoeijmakers.net). ### Web analytics made simple by using Plausible URL: https://rob-hoeijmakers.ghost.io/cookie-free-analytics/ Last updated: 2026-05-10T08:54:22.000Z On this website, I started an experiment today. A change with some impact. I used to measure website traffic with [Google Analytics](https://analytics.google.com), but now switched to [Plausible](https://plausible.io). I want to explain the reasoning and show a bit of the direct outcome. ## Privacy Let's start with the heavy stuff first. A topic we all have options and feelings about but is harder to rationalize: privacy. Google Analytics earmarks your behaviour and browsing on a personal level by placing cookies. Most people authenticate themselves to Google for checking mail, viewing YouTube or other services. With these options available, Google can stitch all your data and devices together and follow your behaviour on an individual level. Next to that, they will connect these details to the IP-numbers (addresses) of your devices and store all this data to analyse and sell to advertisers. The EU law ([GDPR](https://rob-hoeijmakers.ghost.io/gdpr/)) basically tells a website publisher to inform visitors on these policies and give the visitor the option to opt-out from such tracking that involve cookies (earmarks in the browser) or the storing of your IP-numbers. Plausible does not use cookies, does not store your identifiable data and their servers are based in the EU. As such, they are completely GDPR-compliant. ## Ease of use Another aspect of moving away from Google Analytics is the complexity of the user interface. The whole environment feels like Bluebeard's Castle, and a whole slew of apps has sprung up to try and present data in more simple formats. Plausible is a very simple application that shows me as a publisher the basics I need. ## Speed In order to read all the data that Google needs, you have to install a heavy script that takes its toll in the loading of the actual page. The irony is that Google puts more and more emphasis on page loading speeds for high ranking in their search results. I made some screenshots before and after the removal of the Google Analytics script with the Google measurement tools: ![Google Page Speed Insights before and after removing Google Analytics](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/09/speedinsights.jpg) Google Page Speed Insights before and after removing Google Analytics ## Transparency Another nice option from Plausible is to be transparent and show what you measure and what you can view as a publisher. Check it out for yourself: One disadvantage of being private is that Plausible can't see it is me who is visiting, and so I cannot easily filter out my own visits. And I have to be honest, I tried to leave Google Analytics behind before, but came back to ask for forgiveness. Now I have made my motifs transparent as well, I try to be stronger: "My name is Rob, I am a recovering Google Analytics User, I am Plausible since 29th of September 🤪 ". ## Cookies on a website Check for cookies on websites: --- ### Further reading - [Cookieserve - Free online cookie checker for websites](https://www.cookieserve.com) - [Introduction to GDPR](https://rob-hoeijmakers.ghost.io/gdpr/) ### iPhone Live Text URL: https://rob-hoeijmakers.ghost.io/iphone-live-text/ Last updated: 2026-04-25T08:31:38.000Z As someone who loves words and technology, I just had to share this functionality. It more or less freaked me out. Since the latest update of iOS, a simple Spotlight Search goes through my photo’s and finds texts that are **in** the photos. But the precision and depth of this is just spectacular. Let me explain. If you are a bit like me, you probably have many photos on the iPhone. It is my hobby to shoot some quick snapshots everywhere I go, almost like an extended memory. I have more than 30,000 pictures in the photo app. So I got excited when I read about an update to **Spotlight**. That is a search that searches through your collection of content and apps. And after the update I played around with the Spotlight search, I searched for the word SOUL and to my surprise I see in the results a completely non-related photo of a second hand bookshop: ![iOS spotlight example on iPhone showing results from Visual Look up](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/09/IMG_F97AC0A7589D-1.jpeg "Results of the search for the word SOUL") Spotlight Search on iOS 15 on the word SOUL Out of curiosity I open the result and on the photo is an orange dot. I zoom in on the photo and the dot remains on a certain spot on the photo: ![Example screenshot of a search with Visual Look Up on iPhone.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/09/IMG_4AAF411292FF-1.jpeg "Photo of an Amsterdam second hand bookstore") The result of searching for SOUL I zoom in closer and closer, and at some point the text becomes highlighted and readable. On the back of the book, all the way in the corner of the store and the photo, the title says: the global SOUL. Literally, my jaw dropped. What a power, what a precision and speed. Scary and exciting at the same moment. Zooming on the result to find the word SOUL in the left corner. ## Live Text through the camera In the iOS 15 update, there are many more features around text. You can copy and paste the texts out of the stored photo’s, but you can do the same with Live Photo’s. That is also called **Live Text**. You point at a text, for instance a phone number, and dial it straight from here. Really handy. iPhone Live text, click on web link in a live picture ### Ghost for blogging URL: https://rob-hoeijmakers.ghost.io/ghost-blog/ Last updated: 2026-04-25T13:16:41.000Z A couple of months ago, I set up this blog with Ghost. I found the fully hosted version a bit expensive because I was just exploring. I was in the mood for more exploration and decided to fully [host Ghost myself](https://rob-hoeijmakers.ghost.io/self-host-ghost-blog/). Furthermore, I rented a space at Digital Ocean and harnessed my set-up with a CDN from Cloudflare. Most probably not the shortest and easiest route, so time for a quick summary. ## Ghost publishing platform How do I now feel about Ghost after using it for a couple of months? Am I still ravingly happy, or have there been some bumps and scratches along the way? Actually I am really content, Ghost is living up to the promise. I publish new content, shifted theme direction of the topics, renamed the domain, upgraded to version 4 and all with all it works as I hoped it would. My content is well indexed in Google, I have proper search on my site, and I am still very happy with the design. What I also really like is the modular and open set up of Ghost. I integrated Google Analytics and will add Disqus at some point as well. As mentioned, setting op [Site Search](https://rob-hoeijmakers.ghost.io/website-search/) was easy as well. Of course, this was all done in the English language, and I am not sure if Ghost has multilingual capabilities. ## Community Ghost is much more than a blog. Ghost is a publishing platform that lets you build a community of subscribers. I only added the newsletter functionality recently, and now I started building the community. I have to be honest, I still need to start sending out the first newsletter. Furthermore, I am not sure if I will use Ghost or [MailChimp](https://mailchimp.com) for sending emails. Once I have decided and done that, I will update here. It took me some time to get accustomed to Version 4 from Ghost. With this version they launched the membership part and I struggled a bit. Later it turned out that it requires email on the server to be configured, and that was not clear to me. Now that I have configured the email sending service [MailGun](https://www.mailgun.com) the community part started working as expected, and I can get creative and see how to fully use the community features. ## Self-hosting or hosted? The self-hosting is more cumbersome if you are time constrained or rather more focussed on creating than maintaining. Self-hosting is definitely affordable, I only pay Digital Ocean for hosting, but I put a number of hours into upgrades and connections like Mailgun. I would **not self-host**, pay for service and I would go with Ghost or with a service that offers hosting. You can for instance check out [Midnight](https://getmidnight.com), they offer such a service. ## Can you recommend using Ghost for publishing? I have been in digital publishing for a long time, and I am a Digital Producer myself, and yes, after some months of working with the publishing platform Ghost, **I am happy to recommend Ghost**. Fun to use, powerful and versatile. And choose a hosted version if you are serious with your publishing. I also [compared Ghost to WordPress](https://rob-hoeijmakers.ghost.io/wordpress-vs-ghost-for-bloggers/). ### Does my browser support WebP and/or AVIF? URL: https://rob-hoeijmakers.ghost.io/browser-support-avif-webp/ Last updated: 2026-04-14T13:24:52.000Z ### WebP support If your browser supports WebP the label on the image will indicate this and a green check will be shown. If the image label says WebP the browser supports the format, when the label says JPG the browser does not support and falls back on JPG. A red cross appears.‌ ![WebP browser check](/content/files/2022/03/Amsterdam-JPG.jpg) Here is the simplified HTML. I use it in this page. It requests a WebP image and has the JPG as fallback. ‌ ```html Amsterdam ``` HTML code to check for WebP ## ****AVIF support** If the browser supports AVIF the label on the image will indicate this and a green check will be shown. If the image label says AVIF the browser supports the format, when the label says JPG the browser does not support and falls back on JPG. A red cross appears.‌ ![AVIF browser check](/content/files/2022/03/Amsterdam-JPG.jpg) Here is the simplified HTML. I use it in this page. And again, it now requests an AVIF image and has the JPG as fallback. ‌ ```html Amsterdam ``` HTML code to check for AVIF ## ****Files** In order to get this working on this blog page, I uploaded the different formats as a file. I leave them here visible on the page, you might want to download and play around.‌ [ Amsterdam AVIF Amsterdam-AVIF.avif 128 KB download-circle ](https://rob-hoeijmakers.ghost.io/content/files/2022/03/Amsterdam-AVIF.avif "Download") [ Amsterdam WebP Amsterdam-WebP.webp 169 KB download-circle ](https://rob-hoeijmakers.ghost.io/content/files/2022/03/Amsterdam-WebP.webp "Download") [ Amsterdam JPG Amsterdam-JPG.jpg 352 KB download-circle ](https://rob-hoeijmakers.ghost.io/content/files/2022/03/Amsterdam-JPG.jpg "Download") ## ****Read more** I wrote two blog articles on image formats. One goes into a [bit of depth on WebP and AVIF](https://rob-hoeijmakers.ghost.io/webp-avif/), the other article is a bit older and [I recommend people to stick to JPG and PNG](https://rob-hoeijmakers.ghost.io/image-formats-for-the-web/). But the times are changing, not sure if that article aged well. ### Micro content. Small text, big impact. URL: https://rob-hoeijmakers.ghost.io/micro-content-big-impact/ Last updated: 2026-05-10T08:54:22.000Z As a long time digital producer, I was never always fully aware of the impact that micro content has. It was a bit like water for a fish. Vital but not noticed. So time for a closer look on what micro content is and where it is used. ## Three types of micro content Micro content, in the realm of digital information, refers **to succinct and bite-sized pieces of information, typically presented as concise text**. It serves a variety of purposes, primarily functioning as informative titles or guiding elements such as buttons or links for actionable engagement. Broadly, three fundamental applications of micro content come to light: 1. **Enhanced Clarity:** By leveraging micro content, one can offer succinct explanations or descriptions that illuminate the core topic or subject matter. 2. **Content Labelling:** Micro content excels as a tool for effectively labeling and categorizing various content pieces, making navigation seamless and user-friendly. 3. **Action Indicators:** Utilizing micro content to prompt users towards specific actions empowers them to engage proactively, driving desired outcomes. Micro content proves its worth by enabling quick comprehension, effortless content organization, and encouraging user interaction, making it an indispensable asset in the digital information landscape. Let's go in a bit more depth. ## Topic clarification Used this way you guide the reader through a text with for instance subtitles and subheaders. They are clearly visible and often marked with a bigger or bolder font. Micro content is also used to caption an image on a digital page, and most probably also when you use text to describe content in Facebook or Instagram. A usage of micro content that has impact is the page title or email subject line. Will a lead click on your page in Google, or will they open your marketing email? ![Micro content example screenshot. ](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/08/DraggedImage.png) In this example Eric from Leadpages squeezes all the content he can into a tiny space. ## Labels to content Digital content lives in a context, you also have micro content as [metadata](https://rob-hoeijmakers.ghost.io/metadata/). Small bits of text that label the information for context. That can be used for a grouping of content, sorting, a search engine. Often this content is not visible directly to the reader. Sometimes it is visible, for instance as hashtags. ## Pointers for action Often there is micro content required to label an action. For instance, on a button or a link. These are mostly vital in the guidance of the reader. What will happen when I click, why would I click, how to undo a wrong action when shopping. They often convey a call for action. ![Buttons that contain micro content and a tiny icon to support the message.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/08/micro-buttons.jpg) Buttons that contain micro content and a tiny icon to support the message. ## Managing micro content Often, micro content is made and stored separately from the main content. As such, they are differently managed as well. Mostly I make and add micro content on the fly. So for instance, I share an (existing) article on Social Media, I will add micro content like hashtags while creating the posts. Or write a title or subject specific to a purpose. The key is to align, orchestrate all these little bits of content and bring them in line with your story. ## Conclusion on micro content Don’t overlook micro content. Although small and short, micro content has a big impact on how your information is perceived and how it performs. --- ### Further reading - [How to Design Killer Micro-Content](https://designshack.net/articles/graphics/how-to-design-killer-micro-content) ### A Story Tellers tale URL: https://rob-hoeijmakers.ghost.io/story-teller/ Last updated: 2026-05-14T21:03:32.000Z I guess when you put my DNA under the microscope, you would read a story. From the side of my mother, I come from a family of story tellers. For long, I associated telling a story with entertainment and the realms of fiction. Only to read, at 51, the [book of Donald Miller called StoryBrand](https://rob-hoeijmakers.ghost.io/review-building-a-story-brand-donald-miller/). This crystal clear book explains how a story has deep roots in history and psychology. That a story is the perfect vehicle to entertain but also to inform, train and inspire the audience and prepare them for a change to come. Like our ancestors, must have used these stories to share and prepare a coming generation for the joys and dangers of the journey of life. And these stories, through all cultures, share elements and compositions. This message resonated with me strongly. From the roots of these cultural depths, Donald Miller grows a model to analyse facts and create a more compelling story around them. The model is called StoryBrand, and it works for a company but also for a campaign. The model is then put to form in a framework called the **SB7 framework**. With the framework, you get to see who is the hero of the story and how the hero or heroine fights a way through challenges and dangers with the help of an emphatic and clever guide. The guide gives the hero a plan, a map. Towards a brighter future that transforms the hero into a happier, better, healthier person that steers away from the abyss. The luring threats and dangers of being useless, stupid, ugly or plain lonely. A world filled with dangers, uncertainty and villains. ![The StoryBrand SB7 framework. A clear story.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/07/Screen-Shot-2018-01-21-at-10.54.04-AM.png) The StoryBrand SB7 framework. A clear story. I just love the book. The story of stories has struck a cord, and it inspired me to start practising the writing. I decided to pivot my career as a technical Digital Marketeer towards Story Teller. The thrill of a good, exciting and enervating story that helps business people connect to customers. I am excited to learn and practice, to build on experience I already have but now can use to bring liveliness, excitement and change to the marketing message. Thanks to [Donald Miller, his book StoryBrand ](https://rob-hoeijmakers.ghost.io/review-building-a-story-brand-donald-miller/)and the SB7 Framework I am back in a long line of Story Tellers. ### Writing my first StoryBrand URL: https://rob-hoeijmakers.ghost.io/storybrand/ Last updated: 2026-05-14T21:03:02.000Z Locked myself up in the Swiss mountains to focus on writing a [StoryBrand](https://rob-hoeijmakers.ghost.io/review-building-a-story-brand-donald-miller/) for a client. StoryBrand is such a simple and powerful method. Dry facts come to life with shaping a story. The audience and the message become clear and vivid. I am in the flow and will make sure to share results once done. That is, when I safely have got off this storm shaken mountain 😉 ![Writing a StoryBrand to clarify the message](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/07/1626515401038.jpeg) Writing a StoryBrand to clarify the message ## StoryBrand method in short StoryBrand is a marketing methodology developed by Donald Miller that helps businesses clarify their marketing message by focusing on the story of the customer's journey. The method emphasizes the importance of understanding the customer's needs and motivations, and presents the business as the hero who can help the customer solve their problems and achieve their goals. According to the StoryBrand framework, there are seven elements that make up a compelling story: 1. A character who is struggling with a problem 2. A guide who can help the character solve the problem 3. A plan that the character can follow to solve the problem 4. A goal that the character wants to achieve 5. A conflict that is keeping the character from achieving the goal 6. A transformation that occurs when the character overcomes the conflict 7. A moral of the story that resonates with the audience The StoryBrand method encourages businesses to tell a clear, compelling story about how they can help their customers solve their problems and achieve their goals, rather than simply listing the features and benefits of their products or services. It suggests that businesses should focus on building relationships with their customers by understanding their needs and motivations, and presenting themselves as a guide who can help the customer achieve success. ## Review of the StoryBrand book I loved reading the book and I learned a lot from it! It has become part of my day to day toolset, and [I reviewed the book very positively](https://rob-hoeijmakers.ghost.io/review-building-a-story-brand-donald-miller/). ### Semantic web and the knowledge graph URL: https://rob-hoeijmakers.ghost.io/semantic-web/ Last updated: 2026-04-25T13:17:10.000Z I wrote about [metadata](https://rob-hoeijmakers.ghost.io/metadata/) and that touches on the semantic web. With the semantic web, meaning is attached to information (often with metadata and structured data) that makes it readable for machines. Also, storage of this semantic information is done differently, it is stored as Knowledge Graph and retrieved by different types of query languages. I will make this abstract idea a bit more tangible. ![Standardized schema for data enables connections and relations](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/03/2.Standardization.png) Standardized schema for data enables connections and relations ## Describe something Key is to describe the world around us and standardize all the elements and relations. Such a description is called a schema and it uses schema markup to be published into a web page. The best way to explain is to demonstrate. Think of a person and all attributes of someone and their walk of life. Here is a schema for a person: And all the ways to contain this information in markup code. There are different types of code to wrap around the information. Microdata, RDFa and JSON-LD. ## Example of schema I will use Rembrandt van Rijn as an example and look him up in Wikidata: The page gives an extensive listing of the person and relations to other people and objects. The topic Rembrandt has a unique value called identifier: Q5598\. The beauty is that not only things have an identifier but also attributes as date of birth or the values of an attribute like male for gender. You can see the multiple threads of a semantic web. ## Knowledge graph All these things and relations together form a knowledge graph. In this context, graph means a network of connections. This is an abstract concept, but also made more concrete by using graphs as the basis to store and retrieve information. Where a database uses list and relations, the knowledge graph has taken meaning in its design and the knowledge is queried with a different sort of language which is much more adequate for the goal. ![Example of Knowledge Graph](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/03/1-rPYBoPWMQnIF0tvWlfNWAQ.png) Example of Knowledge Graph ## Make the concept useful, for people and machines Knowledge graphs are used by people and machines. One of the most tangible uses is on Google Search. Look up Rembrandt van Rijn and you will see the Knowledge Graph at work. It will, in the search results, display biographical information on the painter but also on people from the same sort of profession of works painted by him. ![Marked information is made with Knowledge Graph of Rembrandt](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/03/rembrandt.jpg) Marked information is made with Knowledge Graph of Rembrandt Next to a machine collecting and presenting information (and some meaning) on the world around us, the semantic web is used for machines to exchange information on a range of topics. Think of companies or a listing of cryptocurrencies and of course AI. ### Metadata explained URL: https://rob-hoeijmakers.ghost.io/metadata/ Last updated: 2026-04-25T13:17:10.000Z Metadata is essentially [data](https://rob-hoeijmakers.ghost.io/etymology-of-data/) about data. It provides information about various attributes of a piece of information, such as its origin, format, size, creation date, authorship, and more. Metadata helps in managing, categorising, and understanding the actual data it describes. I will make this concept clearer and more tangible by making comparisons to our daily world, and then highlight some digital examples. One of the places that shows the most metadata is probably the supermarket. All the cans, boxes, bags etc often hide its contents and as such carry metadata in the form of labels. The label typically describes what is in the packaging, the brand, the weight en and information or prescriptions on how to store. ![Information on the things contained in the box can be seen as metadata.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/03/IMG_0746.jpeg) Information on the things contained in the box can be seen as metadata. The same happens with books. The outside of a book carries in information about the actual contents. Some can be pretty obvious as a name and a title, but often you see the publisher, excerpts and images. And then on the back of the book, a repetition of some information to find it back when the book is put on a shelf between others. ## Digital images and video Also, digital images and video files are accompanied by metadata. Metadata starts with the file and a file extension that indicates the format. But also in the digital image file itself, there is metadata. This can be a wide range of information like camera brand, type, aperture, GPS information, copyright. To make this metadata exchangeable and practical, they are standardized. [EXIF](https://www.exifdata.com/) is the standard for images. Video has metadata bundled up in the container file for the video. ![EXIF metadata displayed for a photo](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/03/Exif-metadata.jpg) EXIF metadata displayed for a photo ## Webpages Pages on the web have a unique name to identify them, but also the file itself carries metadata. Pretty common are title, author, keywords, but the metadata on a page can be extended freely. Search engines use this metadata to query, sort and rank the content. ![HTML metadata script example](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/03/Schermafbeelding-2021-03-01-om-21.30.14.png) HTML metadata 💬 Check out the article called [The outside of a web page](https://rob-hoeijmakers.ghost.io/the-outside-of-a-web-page/) on this topic. There you find explained how to work with Open Graph and make sure your page is properly wrapped in metadata packaging. ## Social media Also, metadata is used for sharing webpages on social media like Twitter and Facebook. This information will give the author or producer more grip on what will be shown when shared on these platforms. And select an image with the proper [orientation and aspect ratio](https://rob-hoeijmakers.ghost.io/aspect-ratio/). ![Twitter card preview of this post on the function and importance of metadata.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/02/image-5.png) Twitter card preview of this post on the function and importance of metadata. ## Other applications using metadata Basically all digital objects have metadata. Sometimes the metadata has an official standard, sometimes they are more loosely defined and extensible. When you put something into a collection, you will often be asked to provide metadata. An app to an App Store, a photo to be shared. ![Data on a picture in Apple Photo (exif)](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/07/Schermafbeelding-2021-03-01-om-21.29.22.jpg) Data on a picture in Apple Photo (exif) ## Creating order and adding value with metadata Metadata is information about information. Metadata can be part of the digital object or attached to it. Most of the time, the metadata will be like an iceberg. Some information is displayed to the consumer of the digital object, most of will be under the water line and used to create order: sorting, storing, ranking, grouping. Adding metadata might seem boring work, it requires discipline and focus, but it provides real value having proper metadata. Trust me on my word, but in my experience **the metadata on content is equally important as its content**, for sure on the longer term, so challenge yourself, take an extra coffee and make sure to add metadata. ### Aspect ratio and orientation URL: https://rob-hoeijmakers.ghost.io/aspect-ratio/ Last updated: 2026-04-25T13:15:58.000Z **Aspect Ratio** – a term that often seems shrouded in technical mystique. I'll admit, I used to be puzzled by it too. But let's unravel the enigma together. Think about the word 'aspect.' It's like looking at different sides of an object, right? And 'ratio'? Well, that's all about numbers, reasons, and calculations. Now, when you put these two words together, 'Aspect Ratio,' the magic happens. At its core, **Aspect Ratio is the mathematical relationship between two sides of a rectangle**. That may sound a bit abstract, but stay with me – it's fascinating. Imagine every screen, image, and video in the digital world as a canvas, and the aspect ratio as its blueprint. We're about to dive into the world of screens, images, and videos to uncover the secrets of aspect ratios. Ready to explore? ## Aspect Ratio of a screen The screen is of course a physical object and as such has a fixed Aspect Ratio. In TV sets, these ratios are more or less standardized. They used to be 4:3 in the past (more square) and have gone more rectangular later on. To 16:9\. If you want to see an even wider screen, you can go to the cinema. For computer screens, there are different aspect ratio’s. In general, they will be 16:9 for a monitor and 16:10 for a laptop. ![Some common examples of screens and there aspect ratio.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/02/what-is-video-aspect-ratio.png) Some common examples of screens and there aspect ratio. The tablet and the mobile phone are special cases in the sense that they can be used as a tall screen or a wide screen. ![Some examples of Aspect Ratio's on mobile phone screens (@Bullfrag)](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/02/smartphone-buying-guide-aspect-ratio.jpg) Some examples of Aspect Ratio's on mobile phone screens (@[Bullfrag](https://www.bullfrag.com/mobile-screen-formats-what-are-they-for-types-and-evolution/)) ## Aspect Ratio of an image Next to the screens also images and video have an aspect ratio. Images can have an array of ratio’s. Even a single image can have be used for different ratio’s. Either by cropping or scaling. ## Orientation of the image Is the image orientation to be lying or standing? When the image is lying down and broad, wider than it is high, it is called **landscape**. When the image orientation is more high than broad, they are called **portrait**. ![Portrait and landscape example. They could have the same aspect ratio but have a different orientation.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/02/Portrait-landscape-orientation0-1.jpg) Portrait and landscape. They could have the same aspect ratio but have a different orientation. ### Dimension This is how many pixels an image will have on either sides of the rectangle. It can be described as size as well. ### Practical use When you work with templates for sites of application, an image will have a fixed aspect ratio, dimension and orientation. You can see this on the feeds and timelines of for instance Twitter and Facebook. There are guides that can help you find the image with the right orientation and aspect ratio. In digital productions, finding the right image and size is so often necessary, and I wrote a separate article on [landscape and portrait images (orientation)](https://rob-hoeijmakers.ghost.io/portrait-and-landscape-images-on-social-media/) for social media. ![Aspect Ratio of images and video's displayed on a mobile phone (@effectivespend)](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/02/all-together-now.jpg) Aspect Ratio of images and video's displayed on a mobile phone (@[effectivespend](https://www.effectivespend.com/blog/facebook-ad-specs-2019/)) ## Aspect ratio of a video Video mostly will match the aspect ratio of the screen it is intended to be played on. But on digital devices like phone, tablet and computer they are fixed by the platform you publish on. Vimeo, YouTube, Twitch, Twitter. I would say the vast majority of online video has an aspect ratio of 16 by 9. ![Image showing Classic TV (4:3) and Widescreen (16:9), the most common aspect ratio's for video](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/02/Classic-versus-Widescreen-TV-Aspect-Ratio.jpg) Most common aspect ratio's for video ### Resolutions for screens and images URL: https://rob-hoeijmakers.ghost.io/resolutions-for-screens-and-images/ Last updated: 2026-04-14T14:21:21.000Z The word resolution is to a large extent a technical term used by digital producers. To the consumer it mostly translates as sharpness, clearness of a display (and has impact on file size and bandwidth). We all turn into producers, and this article can help you navigate the topic of resolution. The information will give you good understanding of the different kinds of resolution and help you produce sharp looking images and video’s for your audience. So let's jump right in! The word **resolution** is being used to describe these three different concepts: 1. Screen resolution: physical pixels in a screen 2. Point resolution: points used to show something on a screen 3. Image resolution: pixels used to represent a digital image or video frame Here is an image to illustrate the different values these concepts will give: ![Overview of different resolutions (GBKSoft)](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/02/3.png) Overview of different resolutions (GBKSoft) Let's go into a bit more depth for all three different uses of the word resolution, and the differences will get much clearer. ## Screen resolution The resolution of a screen is pretty factual, and this is a **physical property**. It typically says something like: the screen resolution is 1080 by 780 pixels. So you could take a [magnifying glass](https://rob-hoeijmakers.ghost.io/three-colors-of-light-and-a-drop-of-water/) and examine the screen. You would actually see these pixels as a tiny hole radiating coloured light. ![@Ariel Waldman, Left: iPhone 6S, Right: iPhone X](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/11/twitter_DP70PtSV4AA-gFs.jpg) @Ariel Waldman, Left: iPhone 6S, Right: iPhone X Some typical [screen resolutions](https://en.wikipedia.org/wiki/Display%5Fresolution) with an aspect ratio of 16:9 are QHD (2560 x 1440) and 4K UHD (3840 x 2160). ## Point resolution The resolution of display refers to **how many points are used to show something** on the physical screen. This also is often called screen resolution. Let’s take an iPhone 12 Mini as an example. The screen has 1.080 physical pixels horizontally and 2.340 pixels vertically. The resolution for display of this device is 375 points by 812 points. ![Different point resolutions for iPhones (UX Collective)](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/02/1_W8y_AgAq2wNluK6dT46MJg.png) Different point resolutions for iPhones (UX Collective) For a long time these two kind of resolution matched 1 on 1\. One pixel in the image matched the physical pixel in the display. Now that the screens have become so good (the eye can’t see individual pixels any more at the right viewing distance) there are more physical pixels than virtual pixels. This site gives you information of the screen you are currently using: ## Image resolution Take a deep breath, there is one more meaning of resolution that is thrown around. The resolution of a digital image and video. Here, the term is used to explain **how many pixels were used to represent the image** or video frame. In traditional television this is a pretty well-defined set of standards, on digital devices it can take many measures or shapes. Let’s say you shoot a picture with a 12 mbps camera, the resulting photo will have a resolution that might have a height of 3024 pixels and a width of 4032 pixels. If you shoot a video with the same device, it probably will give you an option to pick a resolution. For instance: 720p, 1080p or 4K. So here the word resolution also get its notion of **sharpness of the image**. When it can be captured and displayed in high resolution (so with many pixels) our eyes will see a more clear picture. Because it involves more data to represent, the image file size and/or bandwidth for streaming will increase. ![Example of image resolution and zooming in. Left image has resolution of 4000 x 3000 pixels](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/07/Image-resolution.jpg) Example of image resolution and zooming in. Left image has resolution of 4000 x 3000 pixels. ## Summary In this article, I tried to resolve some of the confusion around the meaning of the word **resolution** in the context of screens and digital images. The term is sometimes used for different things that have overlapping meaning. How many lights points does a physical screen have (screen resolution), how many points will be the default display for interacting with it (point resolution) and how many discrete, coloured dots can a camera capture and store (image resolution). ### The miracle of your screen URL: https://rob-hoeijmakers.ghost.io/three-colors-of-light-and-a-drop-of-water/ Last updated: 2026-04-25T13:17:09.000Z I have to be honest, and I watch a digital screen all the time. I look at it more than I look outside through a window. Watch, phone, tablet, laptop or computer. And I never fully realized how these screens work. How they are able to present information or are able to show the beauty of a wide array of things. Of course, I had some cognition, I work in digital environments for more than two decades, but the full realization came by watching and understanding a simple experience: > a drop of water on a led screen will act as a magnifier and a prism And then you suddenly see the three colours, where all this digital marvel is based on. A red, a green and a blue light. ![LED, three colours of light seen through a magnifying drop of water](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/02/IMG_9502-1.jpeg) LED, three colours of light seen through a magnifying drop of water ## Three colours and LED The things emitting these lights are the LED’s, light emitting diodes. And there are three physical separate entities that produce its individual colour of light. A red LED, a green LED and a blue LED. All in a very tiny space, a space that gets more miniature with the year. The three colours now can work together, they act as palette and can generate a gamut of colours. ![Red, Green and Blue (RGB) Colour Wheel](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/02/2560px-RGB_color_wheel_10.svg.png) Red, Green and Blue (RGB) Colour Wheel One such a light mixing unit is called a pixel, and a screen consists of a raster of them with certain height and width. Mixing these three colours of light (RGB Color Model) is the basis for most modern screens. There is more about screens and resolutions in my article [Resolutions for screens and images](https://rob-hoeijmakers.ghost.io/resolutions-for-screens-and-images/). ## Colours are joy And a moral and emotional story that made me appreciate being able to see colour even more: > Watch this student see colors for the first time [pic.twitter.com/Ea9S2xVlTW](https://t.co/Ea9S2xVlTW) > > — Vala Afshar (@ValaAfshar) [February 13, 2021](https://twitter.com/ValaAfshar/status/1360680079400067076?ref%5Fsrc=twsrc%5Etfw) ### Record and edit a professional screencast URL: https://rob-hoeijmakers.ghost.io/screencast/ Last updated: 2026-04-25T13:16:42.000Z My personal aim was to have a Screencast set up where I can share actions on my screen, comment the actions, both with spoken instructions and visual pointers and distribute it on a modest scale. The secondary requirements were to have clean and pleasant audio and video with a technical quality that will last 1 to 3 years. Jump in if you are looking for the same. ## Computer Needless to say is that you need a decent computer that can process the information you feed it will producing. In my case I show events happening in the browser and I run [Camtasia](https://www.techsmith.com/video-editor.html) to record screen, the audio and video stream from the webcam. ## Software to record and edit With the screen recording and editing software from Techsmith Camtasia I am really happy. It is intuitive, versatile and affordable. They have good documentation and instruction video’s to get going and support when you hit a bump in the road. The software enables you to record the screen and then stream from your webcam at the same time. In post production you can then insert the webcam recording as a screen-in-screen addition to your production. With Camtasia you can also insert text, animations, transitions, music etc. You can check the features out at[ Camtasia from TechSmith](https://www.techsmith.com/video-editor.html) and run the free trial. ![Techsmith Camtasia for screencast example](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/02/Techsmith-Camtasia-Screencast.jpg) Techsmith Camtasia for screencast example ## Webcam I used a fairly new computer (Apple iMac 2017) but the webcam is poor. In terms of how it deals with light but also in terms of [resolution](https://rob-hoeijmakers.ghost.io/resolutions-for-screens-and-images/). It records video with a height of 720 pixels (Standard HD) and I personally find that to low of a resolution for a screencast. I upgraded to an external webcam ([Logitech C920](https://www.logitech.com/product/hd-pro-webcam-c920)) that has a resolution with a height of 1080 pixels (Full HD) and is fine for the purpose. ## Microphone The recording sound from the computer is good enough for Zoom calls but not for recording audio for instructive purposes. The build in microphone picked up too many sounds from mouse, keyboard and other ambient noise. I first tried to solve this with a podcast microphone (Blue Yeti) but that was aiming too high. That gives a rich, radio like voice but also picks up other noises (requires a quiet environment) and it is a bit of dominant device if you also record video. I bought a so called Lavalier microphone and made sure it was a wired one. This type of microphone you pin on your clothing and keep it close to the source of the sound (your mouth). The one I bought is digital (USB-C) and plug and play ([Boya M3 Lavalier](http://www.boya-mic.com/lavaliermicrophones/1429.html)). ![I prefered a lavalier microphone for screencast. Hands free, close to the source of the sound.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/02/Boya-M3-Digital-Lavalier-microphone.jpeg) I prefered a lavalier microphone for screencast. Hands free, close to the source of the sound. ## Lighting Once this was all in place I was able to properly record and edit screencasts. Had to work a bit on the lights in the room to make sure that the light sources gave soft, diffused light from multiple sides. I learned a lot from this video from Julie Schiro. Lighting hacks bij Julie Schiro. Check out her channel for other great tips to enhance your digital audio and video. With this setup I was able to produce my own high quality screencasts with clear, enjoyable audio, clean screen-in-screen video of the instructor and crisp display of the actions on the screen. ### High Dynamic Range (HDR) in photo's URL: https://rob-hoeijmakers.ghost.io/high-dynamic-range-hdr/ Last updated: 2026-04-25T13:16:40.000Z Everything in technology needs a label, ideally a three letter acronym. When you open this little box, some seemingly random words tumble out of it. **High Dynamic Range, HDR**, is a way to produce a photo that gives you brighter colours, even in darker circumstances. In this article I will explain how HDR works and how colour information is stored, coded and transferred to the viewing device. ## HDR technology and its history When you capture images with HDR you actually **capture multiple images**. These images have different exposure times and are then put together again, with algorithms, to form a well-balanced and lighted image. When I did research, I found out that this technology to pick out the best lighted parts out of multiple photos is not new. This technology was pioneered in the 1850s by an innovative photographer called **Gustave le Gray**. Le Gray went out to the beach and photographed the sea and the sky. He took multiple photographs with different exposures and then later manually added them together to a single photo again. Furthermore, he cleverly used the horizon to make the cut. ![Gustave le Gray, pioneering HDR. Von der Heydt museum, Wuppertal.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/11/6B50491B-04C5-4E3E-AA06-C6476833B30A.jpeg) Gustave le Gray, pioneering HDR. Von der Heydt museum, Wuppertal. ## How HDR Works for iPhone Now that we've touched on the concept of HDR and its historical roots, let's dive into how High Dynamic Range technology works in the [context of an iPhone](https://rob-hoeijmakers.ghost.io/taking-photos-with-the-iphone-12-mini-a-powerful-camera-in-a-compact-size/). When you activate HDR mode on your iPhone's camera, it's like giving your device a creative tool to capture more detail and vibrancy in your photos. Here's how the magic unfolds: 1. **Multiple Exposures:** Just like Gustave le Gray did with his pioneering work, your iPhone captures multiple images of the same scene but with different exposure settings. In simple terms, some shots are slightly underexposed to ensure the highlights are captured without being blown out, while others are slightly overexposed to bring out shadow details. 2. **Algorithmic Magic:** After capturing these exposures, your iPhone's smart software algorithms take over. These algorithms analyze the multiple images, identifying the best parts from each exposure. The aim is to create a final image that balances out the brightest and darkest elements, resulting in a photograph that captures a wider range of light and color. 3. **Merging for the Ideal Image:** The selected parts from each exposure are seamlessly merged together, creating a single image that's well-exposed throughout. This process helps overcome the limitations of traditional single-exposure photography, where you'd often have to compromise between well-lit subjects and well-lit backgrounds. 4. **Instantaneous Results:** Unlike Le Gray's manual combining of images, your iPhone's processing happens in an instant. Once you tap the shutter button, the device's hardware and software work together to perform the necessary captures and computations in a matter of seconds. 5. **In Your Hands:** After the merging is complete, what you see on your iPhone's screen is the final HDR image. It's a stunning representation of the scene that maintains more of the details in the brightest highlights and deepest shadows. By employing these steps, your iPhone harnesses HDR technology to create images that pop with vividness, even in challenging lighting situations. Whether it's capturing a vibrant sunset against a darkening skyline or a well-lit subject in a backlit setting, HDR helps your iPhone deliver a more true-to-life representation. 📷 Photos made with the iPhone are [stored as HEIC](https://rob-hoeijmakers.ghost.io/heic-a-blood-type-for-images/). This allows for the different exposures to be used and stored in a single file. When you export it as JPEG or WebP these layers will be lost. ### Audio compression and codecs URL: https://rob-hoeijmakers.ghost.io/audio-encoding/ Last updated: 2026-05-10T08:54:22.000Z This article is a primer on Digital Audio for streaming music to a device such as a mobile phone or wireless headphone. It is intended for the digital producer and the interested consumer. Uncompressed digital audio files are large. Compare it to a full balloon. To squeeze the size of the data for audio in smaller files, it needs compression while being transported (streaming). There is compression that is lossless (great for portability) and lossy (for standard playing and streaming). The lossless compression will not touch the recorded audio, while lossy will alter it slightly. So if you need to do further work on audio, save it uncompressed or compress it lossless. > This is a black & white photo. Colour grids were drawn on its surface and your brain is filling the colours in by itself. > My head is done in. [pic.twitter.com/9xo9EGx4ey](https://t.co/9xo9EGx4ey) > > — Daniel🎗🏴󠁧󠁢󠁷󠁬󠁳󠁿 ॐ (@DannyDutch) [September 30, 2022](https://twitter.com/DannyDutch/status/1575931031537991681?ref%5Fsrc=twsrc%5Etfw) ### Three categories audio files So here are three categories of digital audio files: 1. Uncompressed (WAV, AIFF, AU, PCM) 2. Lossless compression (FLAC, ALAC) 3. Lossy compression (MP3, AAC, WMA, OGG) The codec is software that compresses an audio file. The word derives from Coding and Decoding. There are different types of codecs. Some examples: AAC, MP3, WAV, WMA. If the process of encoding or decoding needs to scale up or speed up, there are dedicated hardware solutions. Here is an overview to give an impression on how much a stream can be compacted with different codec and bitrate: | Codec | Bitrate | Filesize | | ----- | --------- | -------- | | AAC | 160 kbps | 16.3 KB | | AAC | 256 kbps | 21.8 KB | | MP3 | 128 kbps | 13.4 KB | | MP3 | 256 kbps | 26.8 KB | | FLAC | 432 kbps | 53KB | | WAV | 1411 kbps | 155KB | | AIFF | 2 mbps | 250 KB | ### Wireless headphones and compression In order to go from source to wireless headphone, the audio also needs to be compressed. That is good to keep in mind if you have fine ears and stream uncompressed audio. Knowing which codecs are supported on your devices will help create an optimal wireless audio experience. At a minimum, make sure your Bluetooth audio devices support AAC as that will give you a solid experience across the biggest range of devices. Apple only supports SBC and AAC on its devices. Everything purchased or streamed via iTunes or Apple Music is encoded over AAC. The only exception is the Mac, which supports aptX. ![To illustrate the point of Digital to Analogue conversion](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/02/lightning_DAC.jpg) To illustrate the point of Digital to Analogue conversion By the way, headphones that use a Lightning cable or Bluetooth have a Digital to Analogue Converter (DAC) inside the headphones. These include Apple's AirPods and EarPods (the wired ones with the lightning plug). It means that audio comes out of the device as a digital stream and is converted to analogue waves in the headphones. ### Audio formats in applications | Application | Audio Codec | Sampling rate | Bit rate | | --------------- | ----------- | ------------- | ----------- | | Netflix | Multiple | | | | Whatsapp | OPUS | | | | CD | N/A | 44.1 kHz | | | FaceTime | AAC-LD | 16kHz | | | AM Radio | | 5 kHz | | | Phone call | | 8 kHz | | | Apple Music | AAC | | | | Spotify Free | AAC | 44.1 kHz | 128 kbits/s | | Spotify Premium | AAC | 44.1 kHz | 256 kbits/s | | Skype | SILK | Variable | | --- ### Further reading - [Digital audio explained with stacks of paper](https://rob-hoeijmakers.ghost.io/digital-audio/) ### Digital audio explained with stacks of paper URL: https://rob-hoeijmakers.ghost.io/digital-audio/ Last updated: 2026-05-10T08:54:22.000Z In this article, I explain storing and streaming Digital Audio and how to manage the data. I will make Digital Audio tangible and demonstrate and explain the file sizes with the visual metaphor of stacks of paper. ![the visual metaphor of stacks of paper](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/02/IMG_9225.jpg) the visual metaphor of stacks of paper ## Samples and bit depth A good start is **sample**. A sample is an atomic slice of the audio file, a data point of the sound wave. ![Analogue, natural sound wave.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/11/2857B78B-8735-47DB-8021-892FA05F709F.jpeg) Analogue, natural sound wave. ![Wave sliced up in data points called samples.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/11/727317D4-8232-42B5-8051-F264A6EAEE6B.jpeg) Wave sliced up in data points called samples. In our comparison with the paper it is like a single sheet of paper from the stack. ![Wave sliced up in data points called samples.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/02/IMG_9226--1-.jpeg) Audio sample compared to a sheet of paper With the single sample, **bit depth** comes in. This is how much information is contained on the sheet. So the white paper on the left will contain more information than the small yellow one. The white one represents a high bit depth and the yellow one a lower bit depth. For music, the bit depths you will come across are 16 bit and 24 bit. For voice, like on phone systems, you will see 8 bits. While the sample rate is concerned with capturing frequency accurately, bit depth is related to **dynamic range**. Dynamic range is the distance between the quietest and loudest sounds in a piece of music, and the quality of the resolution within this range. For many years, 16 bit was the standard, the depth used on CD's. While 16 bit is still very common, 24 bit is now becoming more widely used for Hi-Res (HD) audio. Consumers can now purchase music in lossless formats that support higher sample rates and bit depths. The below illustration shows this concept in detail: ![Sample rates](https://tomsrayaudiomastering.com/wp-content/uploads/2017/03/bit-depth-OPT.jpg) Sample rates ## Sample rate In order to produce sound, we now need to add time to the equation. The faster you move through the stack the more fidelity you get. Sample rate en bit rate How many samples of data are taken per second? This is normally measured in hertz, eg an audio file usually uses samples of 44.1 kHz (44,100 audio samples per second). A single voice signal on the phone occupies 8 kHz of bandwidth. Each sample is quantized into 8 bits, yielding a rate of 64 kbps, which is used universally on phone systems. ### Bitrate When audio streaming became the norm, bitrate was an important measurement. How much data is going to be needed to provide the right quality of sound. Are we moving slowly or quickly through the stack, and a how big are the sheets? We use bitrate to describe the fidelity of audio files. An MP3 file that was compressed at 320kbps, will have a much better dynamic range and sound quality as one compressed at 128kbps. Or, more information can be contained on a bigger sheet of paper than on the smaller ones. Bitrate is also the measure of the rate at which data is transferred from one point to another in time. Think of it as the volume of the pile of paper that needs to be downloaded. ![High resolution has a high bit depth and high sample rate. This means that the required bit rate to play the stream will be high as well.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/02/IMG_9224.jpeg) High resolution has a high bit depth and high sample rate. This means that the required bit rate to play the stream will be high as well. With higher bitrate, audio files with higher bit depth and sample rate can be streamed, thus increasing the quality of the audio. However, this means an increase in bandwidth used for transmission. ## Bitrate formula You can see that sample rate and bitrate are related, but not the same. Here is the formula for it: > Bitrate formula = Sample rate x Bit-depth x Number of Channels A typical, uncompressed high-quality audio file has a sample rate of 44,100 samples per second, a bit depth of 16 bits per sample and 2 channels of stereo audio. The bit rate for this file would be: > 44,100 samples per second × 16 bits per sample × 2 channels = 1,411,200 bits per second (or 1,411.2 kbps) ## Variable bitrate Bit depth is fixed for unencoded streams, but with lossy compression codecs (like MP3 and AAC) it is calculated during encoding and can vary from sample to sample. More on encoding in [this article](https://rob-hoeijmakers.ghost.io/audio-encoding/). ![Variable sample sizes.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/02/IMG_9229.jpeg) Variable sample sizes. --- ### Further reading - [Audio compression and codecs](https://rob-hoeijmakers.ghost.io/audio-encoding/) ### Brand Indicators for Message Identification (BIMI) URL: https://rob-hoeijmakers.ghost.io/brand-indicators-for-message-identification-bimi/ Last updated: 2026-05-10T08:54:23.000Z There is no clear, visual sign that the sender is doing everything to protect the integrity of their emails. A new initiative called **Brand Indicators for Message Identification (BIMI)** is about to improve this. A company is allowed to use a logo in the receiver's inbox when they have checked a number of security and privacy boxes (DMARC compliancy). In order to do have such a logo displayed, the company must be registered, and they need to [buy a digital certificate](https://www.digicert.com/verified-mark-certificates/) called Verified Marc Certificate (VMC). ![Brand indicators at work](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2020/12/vmc-phone.png) Brand indicators at work ## My personal experience with BIMI I wanted to set up a BIMI logo and played around. I first increased the level of strictness on DMARC (quarantine) and then generated a BIMI-record. There were some warnings and hoped it would still be enough to qualify. I inserted the information in a text record for DNS and started validating. ### Certificate (VMC) Then I find out I needed to request an actual certificate. And for such a certificate is a waiting list. I put myself on the list. If you are interested, you can maybe start with [the waiting list for the Verified Marc Certificate](https://www.digicert.com/verified-mark-certificates/). The costs are substantial, though. An annual certificate costs $1299. ## Conclusion Displaying a logo with an email can be a strong defender against fishing, it increases brand trust and visibility, but by design it is also hard to acquire and use the logo. There are many e-mail clients out there and so it is hard to force visibility rules you can communicate. 🍏 Update October 2022\. With the release of iOS16 and MacOS Ventura the Mail App from Apple also supports email sender branding. I can't help noticing that not a lot of companies use BIMI yet. --- ### Further reading - [BIMI Inspector - BIMI Group](https://bimigroup.org/bimi-generator/) ### Measure Page Speeds with Core Web Vitals URL: https://rob-hoeijmakers.ghost.io/measure-page-speeds-with-core-web-vitals/ Last updated: 2026-04-25T13:17:10.000Z How speedy a page loads looks like a simple question. You clock the time it takes in controlled situation and there you go. The reality is a bit more complex. You want to track and analyse the experience of real users. There are loads of different browsers, network speeds, devices, screen sizes and dependencies on third parties. A visitor is not happy when they can see the page but not really click on it. The speeds of your own servers or of third parties you rely on might fluctuate and the experience of speed is subjective. There are many metrics in use and ideally you would want to boil them down to the essence, a core. Something as many people agree upon as possible to make the metrics also have authority and to make them exchangeable. A bit like the civic time. Agreeing on strong metrics also provides a proper fundament for measuring optimization and improvements. ## Core Web Vitals Google has the power, the tools and the brains to set some standards. And even if you don’t agree you probably will suffer because the standards are more and more [part of how Google ranks your pages](https://moz.com/blog/core-web-vitals). The standards Google now have set for Page Speeds, and made part of their key tools, are called [Core Web Vitals](https://web.dev/vitals/). Google boiled the metrics soup down to three key figures: - Largest Contentful Paint (LCP) - First Input Delay (FID) - Cumulative Layout Shift (CLS) ![Core Web Vitals as shown by Addy Osmani](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2020/12/core-web-vitals-addy.png) Core Web Vitals as shown by [Addy Osmani](https://addyosmani.com) ## Largest Contentful Paint (LCP) Largest Contentful Paint (LCP) is all about loading. LCP measures loading performance. To provide a good user experience, LCP should occur within 2.5 seconds of when the page first starts loading. The Largest Contentful Paint (LCP) metric reports the **render time of the largest image or text block visible within the viewport**. ## First Input Delay (FID) First Input Delay (FID) measures interactivity. When can the user start interacting with the page. To provide a good user experience, pages should have a FID of less than 100 milliseconds. FID measures the time from when a user first interacts with a page (i.e. when they click a link, tap on a button, or use a custom, JavaScript-powered control) to the time when the browser is actually able to begin processing event handlers in response to that interaction. For the actual individual visitor this probably best translates to how the interaction feels. Does something happen instantly when I click/tap or do I sense delay. Note that this is something that is measured by recording real usage (see below). ## Cumulative Layout Shift (CLS) Cumulative Layout Shift (CLS) measures visual stability. Is the layout stable or are there shifts in how the content is presented. To provide a good user experience, pages should maintain a CLS of less than 0.1. These layouts shifts you would probably call flickering. You want to press a link and then suddenly something pops up or rearranges itself. Very annoying at times. > To minimize layout shifts, reserve enough space for UI elements that may be loaded in late. Fixed dimensions for banners, images etc. help. [pic.twitter.com/aaIjeyscqy](https://t.co/aaIjeyscqy) > > — Addy Osmani (@addyosmani) [December 2, 2020](https://twitter.com/addyosmani/status/1334029534123868160?ref%5Fsrc=twsrc%5Etfw) Addy Osmani on Cumulative Layout Shifts ## Measurement tools for page speeds The tools you can use to measure the above key values come in two types. The first is a lab set up and the second field work. ### Measure in the lab You provide the URL of the page and then this is tested for speed. You can use [Pagespeed Insights](https://developers.google.com/speed/pagespeed/insights/) or the Lighthouse module in the Chrome browser. These tools provide separate reports for mobile and desktop. There is also a [chrome browser extension](https://addyosmani.com/blog/web-vitals-extension/). I really like the excellent service that [Webpagetest](https://webpagetest.org) offers. It offers great visual and detailed insights and it uses Web Vitals. ![Webpagetest screenshot.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2020/12/Webpagetest.jpg "Webpagetest") Webpagetest screenshot. ### Measure in the field The second type is called Real User Measurement (RUM). With this type the actual cumulative speeds in the wild are used. Sort of field work. It is also the only way to find First Input Delay because this needs real users to click around. For this site I use the excellent reporting module in the [Cloudflare CDN](https://www.cloudflare.com). ![Cloudflare screenshot.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2020/12/First-Input-Delay.jpg "Real User Monitoring report on Web Vitals in Cloudflare") Cloudflare screenshot. ### Web vitals essential metrics [Web VitalsEssential metrics for a healthy site![](https://web.dev/images/apple-touch-icon.png)web.devPhilip Walton![](https://web-dev.imgix.net/image/admin/BHaoqqR73jDWe6FL2kfw.png?auto=format&fit=max&w=1200&fm=auto)](https://web.dev/vitals/) ### Self-hosting a Ghost blog URL: https://rob-hoeijmakers.ghost.io/self-host-ghost-blog/ Last updated: 2026-04-25T13:17:07.000Z I have done some exploring on static sites. I was looking for an optimal set-up to post articles and create pages. Here is a summary of how I arrived at the technology set up for this site: **Ghost, Digital Ocean and Cloudflare.** ![Ghost on Digital Ocean and fronted by Cloudflare](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2020/12/Ghost-on-Digital-Ocean-and-Cloudflare.jpg) Ghost on Digital Ocean and fronted by Cloudflare ## Using Ghost This blog is set up and managed with Ghost. This is an [open source publishing technology](https://ghost.org/), and it is a joy to use. The management interface is clear, speedy and stable. Setting it up and finding a template was really easy, and you can connect the template with GitHub. So if you make changes to the layout, you can easily stage it on your own computer and keep track of the changes. Laying out a blog post or page is easy. The features to use images, video, links, tweets etc. I found of very high quality. You sense that a lot of attention have gone into them without showing off. Ghost has a [CMS](https://rob-hoeijmakers.ghost.io/content-management-system-cms/) built into it and is very much optimized for Search Engines. It has all the proper fields for managing an optimal display on social media like Facebook and Twitter. ![My blog (back then)](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2020/12/Ghost-interface.jpg "Ghost Interface") My blog (back then) ## Hosting Ghost on Digital Ocean I started with the hosted trial service and that was a great way to get to learn how it works. For the purpose I use Ghost for, I found it too expensive, and so I went for hosting it myself. I explored **Digital Ocean** for this purpose. That service was new for me, but once I was familiar with the concept, it took me less time than expected. I created a so-called Droplet (virtual server with networking) and then installed and configured Ghost. I was impatient and did not wait for the domain to point to the proper address, and that caused a bit of extra work. But once that was corrected, I was able to request SSL and set up the working site. Moving over the content from the trial version was a bit of a puzzle. The texts can be migrated easily, the assets like images are much harder to do. So once you have invested time and effort in the hosted version it will not be easy to move everything over, so you need to make the decision on hosting as early as possible. ![Droplet](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2020/12/Digital-Ocean-Niksen.jpg "Digital Ocean Niksen Droplet") Droplet ## Costs for hosting Ghost On Digital Ocean you basically get two months for free and then there is a variable pricing model. At the time of writing I am not sure how much the hosting will cost but will add it once I see the costs. ## Cloudflare CDN To increase loading speed even further, I fronted Ghost and Digital Ocean with the CDN from **Cloudflare**. This offers distribution over multiple locations in the world, but also minification of files (compression of texts) and the option to use HTTP-3\. Next to this, it gives you insight on how well the site is performing in loading the pages. Cloudflare reports on the key values defined by Google called the [Core Web Vitals](https://rob-hoeijmakers.ghost.io/measure-page-speeds-with-core-web-vitals/). ![Speed on Cloudflare](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2020/12/Cloudflare.jpg "Cloudflare") Speed on Cloudflare ## Using Ghost To produce articles, I use Ulysses for writing in Markdown. The pages are rendered by a GitHub hosted template designed and coded by Eduardo Gómez. The template is called **Liebling**. Both the [**workflow**](https://rob-hoeijmakers.ghost.io/blog-writing-workflow/) and the **code quality** of the pages are amazing. 🛠️ Update: I have switched to the hosted version of Ghost and wrote an article about this option: [Publish your blog with Ghost](https://www.hoeijmakers.net/ghost-blog/) ### Using Squarespace for a year URL: https://rob-hoeijmakers.ghost.io/squarespace-the-experiences-after-a-year/ Last updated: 2026-04-25T14:52:08.000Z For a year now, we've been using [Squarespace](https://rob-hoeijmakers.ghost.io/squarespace/) to create and manage sites. Before that, we had already built up extensive experience with a whole range of programs and online applications. Wordpress, Dreamweaver, Joomla, Adobe Experience Manager, Medium, Blogger and [many more](https://rob-hoeijmakers.ghost.io/four-ways-to-build-and-maintain-website/). A relationship of a year is a great time to evaluate. Puppy love is over, is there real love? ## Squarespace enthusiasts First a confession. We really have become Squarespace enthusiasts. This article is therefore not entirely an objective consideration but rather a summary of the experiences, properties and a number of limitations. And argumentation why it works so well for us. But this article is not intended for advertising. ## Start with Squarespace A nice start with Squarespace is that you can try it out for two weeks. That gives you time to see if it meets your requirements and, of course, how it feels to work with it. Squarespace is an **Online Web Builder**, so everything you do basically goes through the browser. The first thing you do is create an account or connect an existing account from Google, for example. We always prefer independent accounts, and so we use those. To help you choose a design, there is a quick step-by-step plan (wizard) that helps you choose. ## Design and build website If you have a trial account, you can save the chosen templates and log in to Squarespace. Log in to the side where you produce, so produce layouts, design and texts. You will soon experience that construction and design are very close together. That is because you work with pre-formed templates that you can adjust after choice. You don't really have to worry about navigation, the elements of a page such as text and images, because they are present as building blocks. One of the things that distinguishes Squarespace is the beautifully, stylishly designed templates. They all have a professional and qualitative appearance. They are also well optimized for both the small and the large screens. The designs adapt well for use on mobile or computer, and you can also view the design when creating and filling in. ![Squarespace webbuilder main screen](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2023/08/squarespace-webbuilder.jpg) Squarespace webbuilder main screen ## Design Designing and managing pages is simple and intuitive. Everything is done via the browser or the mobile app. Completing the site is very easy and there is excellent online support through documentation, webinars and a forum. As soon as you understand the basic operation of adding and placing a component (building block), it is fairly self-explanatory. ## Bring site online In all probability, you already have a domain in use. Connecting to Squarespace is a bit of a hassle, and you may need to get some help from your hoster (hosting provider). Incidentally, there is also an option to connect GSuite from Google to Squarespace, and then you could do with only Domain hosting. You then use Gmail via GSuite. That is a bit too far for here, but the bottom line is that putting your site live is not always easy, but that is not so much due to Squarespace. This is a bit of a challenge in general. You can also host your domain there, but we have no experience with that yet. ## Maintenance and optimization Once our site was launched, many of the benefits of Squarespace became clearer to us. The code of the pages is very suitable for [search engine optimization (SEO)](https://rob-hoeijmakers.ghost.io/optimize-squarespace-for-search-engines/) and also for Social Media. There are extensive statistics and a link with Google Analytics is a small thing. You can use a handy app for maintaining the site as well as for statistics. Adding and changing texts is really easy, and so are images. 💡 ****Update 26th of October 2023** I am a still enthusiastic about Squarespace and I am a circle member by now. In most cases I prefer it above WordPress and in this extensive article [Squarespace versus Wordpress](https://rob-hoeijmakers.ghost.io/squarespace-vs-wordpress/) I list my reasons. For blogging, like on this site, I prefer to use [Ghost](https://rob-hoeijmakers.ghost.io/ghost-blog/). For some use cases, mostly related to e-commerce I might recommend [Wix](https://rob-hoeijmakers.ghost.io/why-choose-wix-to-manage-your-valuable-content/) or Shopify. ## Costs of Squarespace You don't buy software from Squarespace, you take a monthly or annual licence. There are currently four pricing plans. With the basic version (Squarespace Personal) you actually come a long way, but you cannot add your own code, for example. That is difficult if you want to use a Facebook Pixel and you will soon end up at Squarespace Business. Then there are also two packages for Ecommerce (Basic & Advanced), but we have no experience with that yet. ## Lesser points What are the drawbacks we found after a year? The greatest shortcoming is **version management**. You cannot prepare future versions of your pages, and you will soon be working with copies and so on. It is so very difficult to prepare or share larger changes. The other drawback is **multilingualism**. There are actually no possibilities for this, and that means that you should create and maintain multiple sites next to each other. The non-standard offering of the option to use **Facebook Pixel** is also a downside. Also, the options for **privacy settings** (GDPR) are fairly limited. A cookie warning is all there is by default and no opt-out options. ## Interim score after one year The great success of Online Webbuilders such as Squarespace, [Wix](https://rob-hoeijmakers.ghost.io/why-choose-wix-to-manage-your-valuable-content/) and Webflow is completely clear to us. Fast on the road, beautiful designs, good options, no code and no hassle about hosting. It is a logical choice for the commercial individual or smaller company who want to quickly create a professional site and also regularly want to work with it. It is more user-friendly and requires much less maintenance than, for example, [WordPress](https://rob-hoeijmakers.ghost.io/how-about-wordpress/). Furthermore, it seems to be a safe method and the hosting also meets the requirements. There are also useful modules, such as a blog that you can quickly prepare. The support via email and tickets has been very satisfactory so far. We were already very happy with the start, the design and the build, but after a year we are still fine with [the choice of Squarespace](https://rob-hoeijmakers.ghost.io/squarespace). 💡 ****Struggling to find the login page on Squarespace?** You're not alone. While Squarespace excels in offering a platform to create beautiful websites, finding the login section can be a bit elusive. This brief guide will simplify the process for you. ****Quick Steps to Login:** ****\- Visit**: Go directly to [account.squarespace.com](https://account.squarespace.com/). \- ****Choose Login Method**: Use your Google, Apple, Facebook, or Email credentials linked to your Squarespace account. \- ****Two-Factor Authentication**: If enabled, enter the verification code sent to your email or phone. \- ****Access Your Dashboard**: Once logged in, manage your website(s), view analytics, and explore design options. Remember, Squarespace may hide the login option to streamline the experience for new customers. But with these steps, you can quickly access your account and get back to managing your stunning Squarespace site. ### Optimize Squarespace for Search Engines URL: https://rob-hoeijmakers.ghost.io/optimize-squarespace-for-search-engines/ Last updated: 2026-05-10T08:54:23.000Z Making your Squarespace site optimal for Search Engines like Google offers you a number of benefits. It will result in **better ranking** and so probably more qualified visits. There is a big overlap between an optimized site and being **accessible** and **usable**. So better rendering and display on a larger range of devices. The third point is that Google rewards **speedy and secure** sites with higher ranking. And of course your visitors will be happy with a snappier site as well. What does it take to optimize your Squarespace site for Search Engines like Google? Let me first start with listing what Squarespace has built in already for Search Engine Optimization (SEO). In my opinion and experience, Squarespace have done a really great job and the understanding will make it easier to properly configure and optimize. ## Squarespace built-in SEO features What has Squarespace already built in that lays the foundation for an optimal site? I will address that question by looking at the site level and the page level. ### On the site level On the site level, a **good and clear structure** of the pages is key. Compare it to a neighbourhood that has a clear layout and proper street and house tags. You know where you are, and it is easy to understand where to go next. The most important aspect is clean, **consistent URL’s** and an **XML-sitemap**. Clean URLs are static and easy to read and understand by a search engine. Your normal visitors of course can also take clues from such URL’s. The Squarespace site comes with an XML-sitemap. By default, your pages will be on that sitemap, but there are page by page switches to take it off the list. ![Clear structure and meaningful names](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2020/11/goke-obasa-76BYtE6_RI0-unsplash.jpg) Clear structure and meaningful names Security and privacy is important for good ranking, and it is important to have an **SSL certificate** on your site. When you have a custom domain (you should) then you can request such a certificate through Squarespace. ### On the page level On a page by page level, you can use a number of options to optimize your content. For the page is a proper **SEO title and description,** a key feature. These are not in the page, but are an attribute of the page and can be configured through a special panel. For the content on the page, it is important to provide a **good structured text with headings**. Squarespace offers the means for this, and they have also put effort in delivering **clean markup**. For the images that are used, you can provide **worded tags** that give the search engine information on what the images are about. ![Squarespace SEO panel for the page](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2020/11/Squarespace-SEO-page-level.jpg) Squarespace SEO panel for the page ## Google Search Console A must-have tool for SEO is [Google Search Console](https://rob-hoeijmakers.ghost.io/google-search-console-gsc/). This tool you can access with your Google account, but can also be integrated with Squarespace. I prefer accessing it directly. In order for setting up the Google Search Console you need to have Google verify that you own the particular site and [the necessary steps for verification ](https://support.squarespace.com/hc/en-us/articles/205813918)are integrated in Squarespace. Google Search Console is a dashboard to your SEO for the Google Search engine. You can use it to view progress, manage indexed URL's and set settings. How are my site and pages doing? How often did they show up in a search and how many times did people click on my listed pages. But you can also see if there are issues with indexing, mobile pages or page speed. ![Google Search Engine interface](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2020/11/Google-search-console.png) Google Search Engine interface For quick checks on SEO analytics, you can also use the Squarespace website and app. There the key results from the Google Search Console are presented as well. ## Further reading After this high level overview, you are probably confident to dig in deeper. Squarespace in general offers good support content, and they have for sure done this around Search Engine Optimization. I can recommend these: --- ### Further reading - [Increasing your site’s visibility to search engines](https://support.squarespace.com/hc/en-us/articles/205814568) - [SEO checklist](https://support.squarespace.com/hc/en-us/articles/360002090267-SEO-checklist) ### Create your website within 60 minutes URL: https://rob-hoeijmakers.ghost.io/create-a-website-within-60-minutes/ Last updated: 2026-04-14T13:24:37.000Z 🛠️ Updated this article in october 2022 for the update on Adobe Express, formerly known as Adobe Spark. I also added the website builder from Canva. ### Single page site In this article, I will assume a single web page suitable for display on the small and large screen. The names for these types of simple sites with all elements on a single page vary: **Single page site, Single pager, One-page website,** [**Landing page**](https://rob-hoeijmakers.ghost.io/tools-for-creating-landing-pages-and-lead-magnets/)**.** Or for a single page with a temporary function**: holding page, lock screen, coming soon.** There are several online tools for creating and managing these types of sites. For all possibilities, the hosting is done by the service, but a domain can be added. That requires a bit more work and patience. The address of the site is of course unique, can contain a name and can be used as a 'coming soon' page or a springboard to another online module, for example a pop-up shop. ### Carrd With this service, you can build a “one page” website at lightning speed. This could be for your personal profile, a landing page for an email campaign or perhaps something else, more extensive. The service is free but has a more extensive option where you can also use statistics, forms, your own domain and more. Setting up a website is really quick and easy. You choose a template, adjust the captured areas for texts and photos, and you publish. I made this site in less than 15 minutes: ![Single page website in Carrd](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2020/11/Single-Page-Site.jpg) Single page website in Carrd. Everything within carrd.co is in the English language, both in the operation and in the sample sites, but in principle the language is easy to adjust. I have to wait and see whether the site can be found in a search engine, and you don't get statistics with the basic version. Adjusting and further maintenance is then again very easy ### Adobe Express Adobe also offers the possibility to quickly create a single page website for free. I tried the free version and made this in 15 minutes: Express is not specifically intended for websites, but also for a range of expressions on Social Media. With Express, you can make beautiful designs, and quickly share on Facebook, LinkedIn, Twitter, etc. Adobe Express got me up and running quickly, they offer a wide range of templates, but it is not really intended for publishing a permanent single page website. It is more like a very accessible version of Canva. Worth checking out, especially if you have already licensed Adobe: ### Linktree This is also a form of a Single Page, but with the aim of setting up a page quickly to point your audience to multiple places on the web. They also currently offer a free and a paid service. Setting it up couldn't be easier, and it took maybe five minutes to create an account. It is also a piece of cake, but I would not really call this a site. But this can be useful for ad hoc matters such as an event or news items. ![Mobile friendly list of links in Linktree](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2020/11/Linktree.jpg) Mobile friendly list of links in Linktree ### About me A site where you can quickly and efficiently create your own profile. With a photo, a short biography and links to the social media profiles. Works very intuitively, but the aim is to present a single person. So not very suitable for a product, company or event. Unfortunately, the free version has no statistics. ### Canva website builder Recently, Canva added a feature to design and publish a website or webpage. I gave it a quick try and here is my five minutes of work website: Also, Canva provides a rich set of templates and options within a simple flow. 📣 ****Update August 2024**: Canva has put some serious work into their webpage builder. I recently used it [to quickly create a holding page](https://rob-hoeijmakers.ghost.io/viral-photo-seo-digital-strategy/) and that was of great value. I connected my own domain to it but you can now also [get a domain from Canva directly](https://www.canva.com/domains/). ![](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2024/08/Canva---Launching-soon-2.jpg) Canva Website Builder. ### GitHub page For the more tech-savvy person, there is an option to create a page for free and quickly via GitHub Pages. You place your site (publicly) on GitHub and check it for publication. GitHub Pages also has a number of templates available. This option requires that you are already familiar with GitHub, and is therefore certainly not suitable for everyone. ![Your website with Github Pages](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2020/11/Github-pages-1.jpg) Your website with Github Pages ## Overview of services This overview is not completely exhaustive and if I missed a service [let me know](https://rob-hoeijmakers.ghost.io/contact/). All the services and ways that I have listed here serve their own purpose. When it comes to building a nice looking single page website quickly, carrd.co has the strongest cards in it hands. - Carrd: - Adobe Express: - Linktree: - About me: - Canva: - GitHub: If there are long-term ambitions with the site, I prefer GitHub or a more extensive online service like [Squarespace](https://rob-hoeijmakers.ghost.io/squarespace-the-experiences-after-a-year/) or [Wix](https://rob-hoeijmakers.ghost.io/why-choose-wix-to-manage-your-valuable-content/) for a website and [Ghost](https://rob-hoeijmakers.ghost.io/ghost-blog/) for blogging. Your hosting provider may also have other tools that can help you on your way. Earlier, I wrote about [four ways to build and maintain a website](https://rob-hoeijmakers.ghost.io/four-ways-to-build-and-maintain-website/). Have fun! ### For images on the web stick with JPG and PNG. For now. URL: https://rob-hoeijmakers.ghost.io/image-formats-for-the-web/ Last updated: 2026-04-14T14:21:31.000Z Just go with the flow and choose JPG or PNG. > “Football is a simple game. 22 Men run after a ball for 90 minutes and in the end the Germans win. ‘’ This is what came to my mind when considering which file format is best suited for online, digital images. There are literally dozens of options and professionals can talk for hours about this (just like in football) but in this article I’ll provide the information to just get started. I assume the most used and best supported formats for the web. And then there is a simple answer to the question of which formats are suitable for display in the browser: **JPG or PNG** And which of these two you should choose is actually easy to answer. Photos will be JPG, images (graphics, screenshots) will be PNG. ![A typical photo and a graphic. Photo by Robert Hoeijmakers.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2020/11/1-JBxCNwpa4WPqRjnBVpnKQw.jpeg) **A typical photo and a graphic. Photo by* [**Robert Hoeijmakers*](https://about.me/robhoeijmakers)**.* On the left a picture of a photo, use JPG. On the right a picture of an image, use PNG. Since the JPG and PNG formats are the lion’s share of what is currently used on the web, it is good to look at these formats in more detail so that you can optimize them. That is, they will meet these three criteria: 1. Picture looks good 2. Picture loads quickly 3. Image can be used for a search engine ## Quality and size Whether images look technically good depends on the source. So choose a good source. These could be the photos you take yourself, screenshots or [stock photos from Unsplash, for example](https://rob-hoeijmakers.ghost.io/three-ways-to-enrich-your-website-with-top-photos/). With JPG, you can make images smaller but not bigger. Not without loss of quality. So always work with a copy when editing. Just think of a haircut. You can get it shorter, not longer In addition to size, compression is important. Compression is the omission of excess information in the image. You may lose (some) quality with this, but you gain speed when loading. Another form of compression is shrinking the palette of colours, which works when the image contains large, well-defined colours. Compression is done on a scale from little too much compression. The easiest way to compress is to rely on the settings (presets) of a program and then check it visually. Experiment with the size and degree of compression. With Photoshop Express you can experiment with PNG or JPG and adjust quality with a slider: ![Photoshop Express and JPG quality slider. Photo is a screen capture of photoshop site.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2020/11/1-DMpjxKdMkgTK5_qkuF8MJg.gif) **Photoshop Express and JPG quality slider. Photo is a screen capture of* [**photoshop*](https://photoshop.adobe.com) *site.* The rule of thumb is that an image is no more than 300 KB in size. This overview shows what sort of file sizes to expect with different resolutions. ![Scaling a JPG and size reduction. Photo by Robert Hoeijmakers.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2020/11/1-Bj1bmBCZpZ9LHf3MMFWb9Q.jpeg) **Scaling a JPG and size reduction. Photo by* [**Robert Hoeijmakers*](https://about.me/robhoeijmakers)**.* Overview of file size for smaller JPG photo. The decrease in file size is greatest with the first reduction in scale. ## Findability In addition to beautiful and fast-loading images, findability in search engines is often important (SEO). Especially if the image has an independent purpose (and does not serve as decoration). Then give the file a meaningful and descriptive name. Provide captions and alternate text. Place the image with the relevant text so that the search engine can figure out what is on the image. ## Summary This article describes as simply as possible the two most common formats that images (photos and images) have on the web, namely JPG and PNG. And how you can format and place them in such a way that they are easy to find for search engines. 💥 Update, september 2022\. Noticed that website build platform [Wix](https://rob-hoeijmakers.ghost.io/why-choose-wix-to-manage-your-valuable-content/) sets WebP as the default image format for the websites you build on it. A bold move but with all the major browers now supporting WebP it definitely makes sense. ### WebP and AVIF, the challengers of JPG URL: https://rob-hoeijmakers.ghost.io/webp-avif/ Last updated: 2026-04-14T13:25:00.000Z The most common image formats are [JPG and PNG](https://rob-hoeijmakers.ghost.io/image-formats-for-the-web/). WebP is on the rise, a format strongly recommended by Google. The WebP (weppie) format is good at further compressing photos. AVIF, is much newer and actually a derivative of video formats. AVIF also produces much smaller files without noticeable loss of quality. In short, both WebP and AVIF are initiatives to replace the outdated JPG. However, it turns out that old and well-supported formats are not so easily supplanted. This is mainly due to the lack of support for these images by the main browsers. For example, Safari did not support WebP for a long time, but that is changing. There is currently very limited support for AVIF, but you can view with Chrome for Desktop. In addition to browsers, applications for building and maintaining websites will also have to support these formats, and that is certainly not the case yet. ## Check browser support for WebP and AVIF Here is a simple HTML page that's [checks to see if your browser supports WebP and /or AVIF](https://rob-hoeijmakers.ghost.io/browser-support-avif-webp/). ## WebP The first image on the test page should be WebP. If your browser does not support it, it will say JPG on the image. The image is 768 by 1008 pixels in size. In JPG the image weighs 333 KB and in WebP 171 KB. A saving of almost 50%. The HTML code used here is standard HTML5\. The code checks whether a certain format is supported and otherwise falls back to something familiar: ``` Amsterdam ``` HTML5 code to check for WebP support. When not available it will fall back to JPG. ## AVIF And the same test for the much newer [AVIF format](https://jakearchibald.com/2020/avif-has-landed/). The AVIF image is an additional 12% lighter than the WebP. ## JPG If your browser supports both WebP and AVIF (like the latest version of Chrome) you can see the JPG here to compare the quality: ![Photo by Robert Hoeijmakers](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2020/11/Amsterdam.jpg) **Photo by* [**Rob Hoeijmakers*](https://about.me/robhoeijmakers) ## Support for AVIF in applications The browsers are a good representation of how far the acceptance of a new file is. But also the applications that work with images. I did a test with [Squarespace](https://rob-hoeijmakers.ghost.io/squarespace), the online web builder, and with [Pinterest](https://www.pinterest.com) and both don’t even support WebP, let alone AVIF at the moment. ## Squoosh, online tool for conversion If you want to experiment with WebP or AVIF yourself, [Squoosh](https://rob-hoeijmakers.ghost.io/crisp-and-fast-loading-images-with-squoosh/) is a very handy site. ## ### An explanation and demo of App Clips URL: https://rob-hoeijmakers.ghost.io/an-explanation-and-demo-of-app-clips/ Last updated: 2026-04-25T13:16:02.000Z In September 2020 the latest update for the iPhone, iOS14, was made available. This includes a number of changes, but perhaps the largest was not yet noticed. That’s a feature called App Clip. Here, Apple offers the users the playing field and not the players. So the applications for this are bound to appear in the coming months. App Clips have the potential to shake up the digital and real world. I will explain App Clips further in this article and make it tangible with a short demonstration. ## What are App Clips? App Clips are fragments of the apps you normally install from the App Store. These fragments are not installed after calling, but are shown as a small window (Apple Clip Card). If you open them, then you will see little difference with the normal apps. Here’s an example of the steps to quickly order a cup of coffee: ![Example flow of steps in App Clip. Ordering a cup of coffee.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/11/1_5EUemAmJzY-_f1KOLEaYQw.png) Example flow of steps in App Clip. Ordering a cup of coffee. I made a short demo, that makes it visible and tangible. I scan a [QR tag](https://rob-hoeijmakers.ghost.io/qr-code-explained/) via camera and that starts an App Clip, in this case from Chibi. Furthermore, I also show the NFC tag. With an iPhone NFC reader/writer I have written the URL () into the NFC tag. The link of these App Clips with the real, tangible world makes it fascinating. Because in addition to browser, Apple Maps and iMessage (digital world), App Clips can be started in the real world via QR tag and NFC tag. Here the pieces fall together, and you see the power of the idea. In a way, it is very similar to the very popular Apple Pay. You touch a marked object with your phone and a functionality becomes available. No installation required and get started right away. This is suitable for many situations where you do not have the time or desire to install an App and that especially opens up possibilities for transactions. Even more so because you can log in and pay via Apple Pay. ## Sign up for Web Strategies ### Blog for the online professional. Subscribe Email sent! Check your inbox to complete your signup. No spam. Unsubscribe anytime. ## Try this App Clip example yourself You can try it yourself if you want. For example, via the QR tag below or via this link: ![QR Tag to start the Chibi App Clip.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2021/11/1_goZdwHhL38Q2RZ_sHZu70Q.jpeg) QR Tag to start the Chibi App Clip. Disadvantages or limitations? Sure! The Apple App Clips take place entirely in the Apple ecosystem. As a publisher, you must already have an existing app in the App Store and the App Clip must not be too large. *For deeper insight into the QR tag, check out this article about [the secrets behindthe QR Code](https://rob-hoeijmakers.ghost.io/qr-code-explained/)*. ### Share Kobo e-books with your family URL: https://rob-hoeijmakers.ghost.io/share-kobo-ebooks-with-your-partner/ Last updated: 2026-04-14T13:20:53.000Z Both me and my partner have a Kobo reader. You would think sharing a book within the family is easy and might even go over Wi-Fi. Zoofffff. Like we share paper books: "*I am done reading, the book is absolute worth reading, here you go. Please don’t make dog ears".* Of course, this is not how it works. The Digital Rights of these e-books are managed on a personal basis. Legally the e-books are even not fully yours, you own a licence. When you want to share an e-book, you create a copy, and that is not allowed within this system. But giving away your e-reader to share the e-book is probably not practical as well. > Legally the ebooks are even not fully yours, you own a license. Rather than beaming the book over, you will need to go through some time-consuming steps. So, how to move around the limitations and share e-books within the family? ## Step-by-step guide to share e-books To start, and try to make things simple, I made this overview of the flow of things: ![](https://miro.medium.com/max/1400/1*00ZRH5-zLc7kYF5e0RBNow.jpeg) Steps to share ebooks on Kobo Through your Kobo online store, [download the e-book](https://help.kobo.com/hc/en-us/articles/360019527954-Download-books-from-your-Kobo-account-to-export-to-another-device-or-app) you want to share. Once logged in, the site offers an option to download. Doing this will not download the actual e-book but a file (with the extension ACSM, and it contains sort of a token). That file you need to open with a program called Adobe Digital Editions. From there you can download the protected e-book. ## Remove the digital seal You then remove the seal by importing the e-book into Calibre. The file to import resides in a folder that Adobe created. Calibre removes the seal (called Digital Rights Management or DRM) if you installed a specific plugin. The installation of the plugin is described [on this DRM removal plugin page](https://www.epubor.com/calibre-drm-removal-plugins.html). Once the DRM is removed from the e-book, you can send the EPUB-file off to your partner. You can do this by locating the EPUB-file and for instance use email to send the file. The receiving party then also needs to use Calibre to get the file in their library and move it onto the Kobo reader. ## Legal aspects It might be that this procedure is not legal in the country you live. Paying for books is the right thing to do, this is not intended for commercial purposes. As it is worded on the DeDRM plugin site: > *Please only use these tools for gaining full access to your own ebooks for archiving/conversion/convenience. De-drmed ebooks should not be uploaded to open servers, torrents, or other methods of mass distribution. No help will be given to people doing such things. Authors, retailers and publishers all need to make a living, so that they can continue to produce books for us to read. Don’t be a parasite.* ## Summary of steps Short **summary** of steps: 1. Make sure to install and configure Adobe Digital Editions, Calibre and DeDRM plugin 2. Download the e-book ACSM link to computer 3. Download the e-book file to Adobe Digital Editions 4. Import into Calibre (DRM is removed) 5. Exchange e-book file I think setting it all up for the first time, installing software, creating an Adobe ID and getting familiar with Calibre will take **an hour at minimum**. It definitely helps if you are already using Calibre. But hang in there, it is worth the trouble. The second time you follow these steps, it will take less than **3 minutes**. I hope it was helpful, although I understand these are not easy and simple steps. I wrote more articles on Kobo, one on [first aid with problems with the reader](https://rob-hoeijmakers.ghost.io/kobo-clara-crashes-and-freezes/) and a popular one is [how to load your free ebooks through iPhone.](https://rob-hoeijmakers.ghost.io/load-free-ebooks-on-a-kobo-using-the-iphone/) Great for travel and holidays. If you liked the article consider [subscribing to free updates](https://rob-hoeijmakers.ghost.io/newsletter/), I write about digital life and content creation. ### Load free ebooks on a Kobo using the iPhone URL: https://rob-hoeijmakers.ghost.io/load-free-ebooks-on-a-kobo-using-the-iphone/ Last updated: 2026-05-10T08:54:23.000Z Do you have the need to load alternative or free e-books onto your Kobo when you are on the move? Or off the normal tracks? I explain how loading alternative e-books from iPhone to Kobo works and what you need to have and pack. Of course, you can turn your iPhone into a Wi-Fi hotspot and buy items through the official shops. But suppose you have e-books that come out of another online resource like [Gutenberg](http://www.gutenberg.org/) or [Z-Library](https://z-lib.org/). How can you make that work? I explain it extensively below in three steps: 1. The hardware prerequisites 2. Hooking up the devices 3. Doing the right steps for copying The video’s can be handy when you like to see it in action. I am not affiliated to Apple, Kobo or any of the sites or manufactories I mention. I make these stories and videos because I like to figure things out and share my experience. ## **The hardware prerequisites: the camera adapter and the standard cable** How can you then transfer a downloaded e-book to the Kobo? First I bought an adapter. But the really important thing is that the adapter needs to have an entry (called Lightning) for charging. Like you see here: ![The Camera Adapter for iPhone that I use for transfering ebooks to Kobo Clara](https://miro.medium.com/max/1400/1*MjdEWn1egBWdEBxpsxmFiw.jpeg) The camera adapter in close up So again, use the camera adapter with the charger entry, the iPhone notices the difference in hardware and will complain about not enough power to operate the attached device. For my setup, it did not matter if I plugged in the charger or not, but reported it did for his setup. So to be on the safe side, plug in the charger as well. An extra benefit is that you can charge Kobo and iPhone together. The adapter is referred to as *Lightning to USB 3.0 Camera Adapter* and you can buy it from Apple or another company. I bought mine for 20 euro at a web store. ## **Hooking up the iPhone and the Kobo** Once you have the adapter, you can use the regular USB cable (USB Mini) for the Kobo. You plug it into the adapter and then into iPhone and Kobo. Order does not matter. I show it quickly in this video with my iPhone 8 and a Kobo Clara: Connect Kobo to iPhone through camera adapter Once attached you will see a message on the Kobo which you need to confirm. It is the same message you see when you hook it up to a computer. After confirming (pushing the confirm button) the two devices are connected. On the iPhone, the Kobo will show up as a drive under files. ![An iPhone showing the connected Kobo e-reader as a location.](https://miro.medium.com/max/1400/1*bWF304nlb8q6RmLKOaokPg.jpeg) The Kobo Reader shows up as a location to store items And here is a screenshot from the iPhone that gives you an idea how this works. ![You can use USB On The Go (OTG) to turn the iPhone into a host for devices like Kobo.](https://storage.ghost.io/c/20/27/2027934d-fd7f-4725-bcbf-86fe1f55dd12/content/images/2022/03/IMG_663067EBA135-1.jpeg) Screenshot of the iPhone with Kobo connected ## **Copy downloaded e-book to Kobo** The next steps I found a bit cumbersome, but with the steps written out it is much easier. Months might go by between using this workflow. So here are the steps: 1. Download the e-book 2. Locate the EPUB file 3. Select it for copy 4. Go to Kobo drive 5. Paste the file It sometimes helps to see it at work, so I made this clip of me actually performing the above steps after downloading an e-book from a website: Steps on iPhone to move downloaded ebook to attached Kobo Clara And here is the same workflow, but then as a screen capture: Copy ebook from iPhone to connected Kobo Clara ## **Closing the connection** Now the downloaded e-book is on the Kobo. Unfortunately, I did not find an option to remove to safely remove the Kobo from the iPhone, and so I advise playing it safe and first power down the iPhone. And then disconnect the cable. > Make sure that you remember your PIN before you restart the iPhone. When you disconnect, the reader will extract the e-book and add it to the list. Ready for reading. The iPhone can be powered up again. Have fun with it! 💡 ****Should this also work on an Android device?** I am not 100% sure but I found this on [StackExchange](https://ebooks.stackexchange.com/a/9161): **Simply use a cable Micro USB to USB-C. You connect USB-C to your Android phone and Micro USB to your Kobo eReader and voilà! Using any file explorer application on your phone, you can simply copy epub files directly to the eReader mass storage.* --- ### Further reading - [Connecting a Kobo Reader to an iPad via USB-C: Easier Than Expected](https://rob-hoeijmakers.ghost.io/connect-kobo-reader-to-an-ipad-via-usb-c/)