Anthropic put out Fable 5 a couple of weeks ago. The best model they’ve ever made. Within days, it was gone, and not because anything went wrong with it. The US government told them to pull it.

If you’re not a US citizen, you don’t get to use it. Living in San Francisco doesn’t help. Working at Anthropic doesn’t help either, they had to cut off their own foreign staff.

I live in Germany, so that includes me. Probably many of you reading this too.

The cutoff

What happened, roughly. In mid-June the Commerce Department hit Anthropic with export controls and ordered it to block Fable 5 and its bigger sibling Mythos 5 for every foreign national, inside the country and out. Anthropic can’t really tell who’s a citizen from an API call, so it just switched both models off for everyone. Worldwide.

The official reason was national security, and that was about all they said out loud. The reporting filled in the rest. A group linked to China had apparently gotten into the new model, and officials were nervous that Fable 5 could be talked into helping with real cyberattacks, the kind that go after banks and their ancient backend systems. Anthropic argued that one narrow jailbreak shouldn’t kill an entire commercial product. The government didn’t agree.

The logic is what the US already does with Nvidia and AMD chips. Some countries don’t get the good silicon. Now they don’t get the good software either.

OpenAI Too

And it isn’t only Anthropic.

OpenAI’s newest, GPT-5.6, got the same thing in a softer wrapper. The administration asked OpenAI to hold back the wide release and keep it to around 20 partners the government had vetted.

Sam Altman reportedly told staff that access gets approved one customer at a time, by Washington. So it’s not a blanket foreigner ban like Anthropic’s. It’s a guest list. And you’re not on it.

Different method, same result. The two best models in the world right now both sit behind a gate in Washington.

This switch is scary

A few American companies build the best models on earth, and the American government decides who gets near them.

That’s a scary and powerful switch, and unfortunately the lunatic in DC has his hand on it.

We’ve been here before with other things. Musk cut Starlink coverage over a war zone when it suited him, because he could. Europe got nervous enough about American payment rails that it’s now building its own digital euro.

Now it’s AI’s turn.

Which is, in some sense, worse, because AI is becoming the thing everything else runs on top of. Everyone in the world.

We need our own

Europe needs its own. So do other parts of the word. But we’re freaking far away.

AI models are not a nice-to-have anymore. Real labs and real models, built in the EU and elsewhere, that no foreign government can reach over and switch off.

We’ve got the start of one. Mistral, in France.

Their whole thing is open weights, models you can download and run on your own machines, which does matter. You can’t recall a model that’s already sitting on someone’s servers in Frankfurt. The rest of Europe’s AI scene is there too, just small… very small.

Aleph Alpha in Germany gave up chasing the frontier a while back. Black Forest Labs does image models.

It’s a start. Not much more than that. We are very far from having great AI models for everyday users.

The money is one thing

Just look at the money.

OpenAI is pulling in something in the mid-twenty-billions a year. Anthropic is past a thirty-billion run rate. Mistral expects around one billion euros this year. Round those however you like, they’re not in the same stratosphere.

The models say the same thing. Mistral’s best is the weakest reasoner of the frontier bunch. It’s okay. It can do a lot. But it isn’t the best by any means, and it isn’t pretending to be.

Then it gets circular… unfortunately.

You can’t train a top model without a huge pile of chips, the best chips are American, and the US decides who gets them. The same rule that just locked us out of Fable also makes it harder for us to ever build our own.

We’re behind, and a big reason we stay behind is the country we’re trying to catch.

See the problem…

To be fair

Some of this is a real concern.

If a hostile group did get into the model, and it really could speed up an attack on a bank, I understand hitting the brakes. Not every national-security line is a cover story.

There’s also a contradiction the US hasn’t sorted out. Trump’s AI pitch is that the world should run on American models, American AI everywhere as the default. You don’t get there by slamming the door on your best ones.

Every time Washington yanks a model, Mistral’s boring open-weight approach will get more attractive. A model you host yourself can’t be switched off by anyone.

The Bottom Line

Winning the AI race, is not in the cards for us in the EU. Maybe ever.

Just having something nobody across an ocean can turn off on a Tuesday should be the goal, though.

We don’t have that yet. We’ve got a small start, a long road, and a chip supply issue.

I’d love to work with a European AI model. Can’t right now. Not really.

I’m somewhere in the middle.

A few hundred, maybe 1 or 2K dollars a month from writing. A newsletter. Then a handful of products that bring in money. Affiliate links. Partner. Things like that. Not nothing. It’s great. But not a true living either. At least in Germany, where I live.

For a few years that place felt like a fine place to be. Keep publishing, keep getting a little better, and the number creeps up.

That part is getting harder I feel like. The work didn’t get worse. But the middle itself is thinning out. Don’t you think?

The creator middle class gap

The creator economy as a whole is huge now.

Something north of $200 billion. More than 1.5 million Americans call themselves full-time creators, roughly seven times as many as in 2020.

So you’d think there’s more money to go around. There is. It just isn’t spread the way you’d hope.

One report said: the top 10% of creators took 62% of all payments in 2025, up from 53% two years earlier. The top 1% alone grabbed about a fifth of it.

And here’s the number that actually says it. Average creator earnings went up. The median went down, from around $3,500 to $3,000.

When the average rises while the middle falls, you know exactly where the money went.

It’s the same shape as the wage gap. A few people way up top. A lot of people at the bottom making close to nothing. And the part in between getting squeezed from both ends.

There’s almost no middle class anymore.

Brands skip the middle

For a lot of creators, the money was never the platform alone. It was brand deals. And brands changed how they spend.

When budgets are tight, they go in two directions. Up, to the big names with the reach. Or down, to the nano and micro creators who are cheap and come in bulk.

The middle gets skipped. Again. Too expensive to be a bargain. Too small to be a headline.

Then there’s AI

The other squeeze hits writers and bloggers, creative creators harder than anyone.

Search used to be a great equalizer. A small site could rank, get found, and pull in readers for years. That’s how a lot of the middle class survived.

Google’s AI Overviews are eating that. AI chats are eating that. They answer the question right there, so nobody needs to click into the article or content source. When that summary shows up, something like 1% of people click through to the source. 1%…

Traffic to news sites dropped about a quarter in the year after AI Overviews rolled out. Some small publishers lost 60%. A few lost more.

AI competes with us for attention. And it changes the reader too. I wrote a whole piece on how AI isn’t replacing writers, it’s replacing readers. People skim a summary and move on. The slow, careful stuff the middle class is good at gets compressed into three bullet points by a machine before a human ever sees it.

Platforms reward extremes

Substack tells a similar story. Gross writer revenue keeps climbing, past $450 million. Sounds great.

But the median writer there makes something like $4,000 a year. A few dozen people make over a million. Most make a few hundred bucks, if that.

Algorithms don’t help the middle.

They reward whatever gets a reaction. Outrage, drama, big claims, short bursts. The steady writer publishing twice a week with no theatrics is exactly the kind of work that gets buried…

Good used to be „good“ enough. Publish consistently, build trust, help people. Now the rules are about hooks and funnels and being loud.

The market rewards spectacle.

It wears you down

When someone at the top posts “I made $35K this week,” it’s meant to inspire (often). For people at the bottom, it may.

For the middle, it stings.

Because you’re closer to them than it feels, but the distance looks enormous. And the questions start. Is my niche wrong? Am I not hustling enough? Should I be louder, more extreme, less thoughtful?

I don’t think it’s jealousy. It’s the tension of working hard and watching the chart pretend you don’t exist.

The middle works. Top creators automate. The bottom dabbles. The middle edits the post, answers the comments, builds the actual community. And then gets told that wasn’t the right content format.

There is another side

The middle is dying in one place and holding in another. It’s dying in short-form, the cheap interchangeable stuff a brand can reassemble out of fifty micro creators. It’s holding in long-form, where a real following is something nobody can buy back.

That’s the thing the middle still owns. Trust. A reader who actually reads. Brands and readers both lean on the people who aren’t only chasing virality. They want credibility.

We can’t lean on one platform.

The deals dry up, the algorithm change, the search traffic goes down, and if all your eggs are in one basket you go down with it.

The Bottom Line

The creator middle class is getting pressed from every direction at once. Brands, algorithms, search, AI, and a culture that only points the camera at the extremes.

I’ve written about this creator middle class problem before, and I keep coming back to it, because it’s personal… I am in there.

The people in the middle are the ones who do great work. But can’t survive on it.

I wrote that there is no Apple Workspace. No console, no business email, no team chat. Still true.

But Apple is closer than it looks. Closer than Google was in 2006.

They already sell Apple One, a bundle of consumer services. They already run Apple Business Manager, a console for company devices. Two halves of a business suite, sitting in the same building, never introduced to each other.

So let me build it for them. Call it Apple One Business.

The console

Apple Business Manager already exists. It deploys Macs and iPhones, hands out apps, creates Managed Apple Accounts. It just stops at the hardware.

Let it keep going. Add a person, and they get an email address, a calendar, Drive space, iWork, and the chat. One click. Remove them, and all of it dies at once, the leaver locked out of every document and thread. Set a password rule, it covers everyone.

That is the one thing Apple is missing, and they are one product decision away from it.

Email

iCloud Mail already takes custom domains. Five of them, on iCloud+, today.

Make it real. you@company.com, with an admin who can create and kill mailboxes, group aliases someone can actually manage, retention rules, shared boxes for support@ and billing@. The inbox is already good. It just needs a manager behind it.

Files

iCloud Drive already syncs everything across every Apple device. Add team spaces.

A folder that belongs to the company, not to a person, with real permissions. The freelancer sees one project and nothing else. When they leave, the files stay. Google solved this years ago. Apple already has the storage and the sync running. It just ties them to one Apple Account at a time.

iWork

Pages, Numbers, and Keynote are good apps with weak web versions. Fix the web first.

Then make the .docx and .xlsx export clean enough that the rest of the business world stops wincing when a Pages file lands in their inbox. iWork already does live collaboration over iCloud. Two people, same document. Push it the last mile and it competes.

The work apps

Notes is already a fine wiki. Give a team a shared space, the internal links it already supports, and it would cover what a small shop needs. I went deep on running Apple Notes like Notion in a separate piece. Most of that is one admin layer away from being a team product.

Reminders is the bigger opportunity. Right now it reminds you to buy milk. But the bones are there: lists, tags, sections, smart lists.

Grow it into a real project tool. Project views. Boards a whole team shares. Who is doing what, and due when. A simple report at the end of the week. Microsoft ships Planner, To Do, and Project for exactly this. Apple ships one app and calls it personal.

A good task manager and a real project tool are two of the most important things in any business. It does not have to stay personal.

Chat and video

It gets a bit trickier here. There is no Apple Slack. There is no Apple Teams.

But Apple could build one on iMessage. Channels, threads tied to a project, an admin who can pull a leaver out of every conversation at once, and the freelancer on Android allowed in through a link.

FaceTime already runs in a browser. Add scheduling tied to the calendar, recordings stored where an admin can find them, host controls a company would sign off on.

The consumer versions are pretty good. The business versions just need the admin stuff.

The bundle

Apple One already bundles consumer services into one monthly price. Apple One Business does the same, per seat. Competing in price with Office 356 or Google Workspace, or if the really wanted to light up the market, make it much more affordable for small companies. Maybe not payment per seat but just a flat monthly fee per company up to X users.

For more than that: A few dollars a month per user, like Google/Microsoft.

Bundle: Email, Drive, iWork, Notes, Reminders, chat, video, one console to run all of it. On the best hardware in the room, which Apple also happens to sell.

They would get paid for the suite and the Mac it runs on. I suspect the margins would be fine.

Will they?

Probably not.

An enterprise suite means contracts, SLAs, compliance audits, a sales team, and someone answering the phone at 3am when email goes down.

For now, Apple sells the device and let Microsoft and Google fight over the software on top.

The Bottom Line

Apple has every piece for this Apple One Business Suite. The hardware, the apps, the storage, the console, the bundle, the brand, the money.

They could be the workspace, not just the laptop the workspace runs on.

They just have to want to. And they don’t. For now…

Some time ago, Wero showed up in my banking app. A private payment system, built by 16 European banks, to send money across the continent without touching an American card network. I just wrote about Wero.

The digital euro is another version of that idea. Money issued by the European Central Bank directly. Not a bank. Not a card. The central bank.

And after years of “studying it,” it’s now on a real timeline.

It’s also the most controversial money project Europe has ever attempted. For good reasons.

Timeline

The European co-legislators are expected to adopt the digital euro regulation during 2026. Legal green light, so to speak. Without it, nothing happens.

The ECB published a call for expression of interest in March 2026. Payment providers had until mid-May to apply to help build and test it. Selection wraps up by the end of June.

A 12-month pilot starts in the second half of 2027.

And a potential first issuance… 2029. Assuming the law passes in 2026.

What is it

A digital euro is central bank money in your phone. Digital cash. Backed by the ECB, not by a commercial bank.

That distinction is interesting. The money in your checking account isn’t really “the bank’s IOU to you.” If the bank fails, that’s a problem. Central bank money can’t fail the same way. Cash works like that today. The digital euro extends it to the phone.

It’s designed to sit next to cash, not replace it. For now…

There’s also an offline mode. Pay someone phone to phone, no internet, and only the two of you know about it. Cash privacy. In theory.

Sovereignty

This is why it exists.

Right now, most digital payments in Europe route through Visa and Mastercard. American companies. The ECB’s own people have publicly warned that Washington could, in theory, switch off Europe’s payment rails. Because nearly all of them run on US infrastructure.

Wero is one answer to that. The digital euro is another one. State-backed money that no foreign company and no foreign government can flip off.

Europe is trying to ditch Visa and Mastercard with Wero. The digital euro goes one layer deeper.

Wero moves your bank money around. The digital euro is the money itself.

Christine Lagarde has pushed this for years. The European Parliament voted in February to formally back it, framing it as essential for monetary sovereignty.

The catch… there always is one

Central bank money in digital form means the central bank can, technically, see a lot. Critics call it a trojan horse for financial surveillance. Every transaction visible to the state. Combined with AI, profiling… yeah.

The ECB’s answer: privacy by design.

A hard no to programmable money. A guarantee that cash stays. And the offline mode for anonymity.

But the offline mode, the truly private one, comes with caps. Small per-transaction limits. So your anonymous spending is fine for a coffee, but traceable for a couch.

There’s also a holding limit. Something around 3,000 euros is the number. You can’t park your life savings in digital euros, because if everyone did that during a panic, the banks would empty out overnight…

So the design is a tightrope. Private enough that people trust it. Limited enough that it doesn’t break the banking system. Useful enough that anyone bothers.

Really hard to get all 3 right…

Downsides… because there are always some

It’s slow. 2029 for a first issuance is still a long time (in tech).

Which raises the obvious question: do we even need it? Wero works and gets better. Apple Pay and Google Pay work. The case for a public digital currency as sovereignty is a harder sell to someone who just wants to split a dinner bill.

The banks aren’t thrilled either. A digital euro competes with their deposits. The holding limit exists partly to keep them calm.

And the privacy promise only holds if the final design respects it. Promises made in 2026 might not be the same in 2029. We’ll see.

The Bottom Line

Sure, Europe should own its money rails. That part isn’t really up for debate for me.

But money is the most personal infrastructure there is.

The digital euro could be a great thing. Or it could be the most monitored money in the free world.

Every big AI company lets you build your own little assistant.

OpenAI has CustomGPTs. Google has Gems. Anthropic has Skills. Microsoft and Mistral have their own versions of the same thing. Or do they? Are these the same thing or not?

They sound like five different products. Or they’re one product wearing five outfits.

Let me explain what these actually are. Because the marketing pages won’t.

The same trick

None of these are new AI models. They’re all the same model you already use, with a costume and a briefing.

A CustomGPT is still ChatGPT. A Gem is still Gemini. A Skill is still Claude. You didn’t build a smarter AI. You wrapped the existing one in (smarter) instructions.

The wrapper usually has three parts:

• Instructions. A standing prompt that tells the model who to be and how to act. “You’re my newsletter editor. Cut fluff. Keep my voice.”

• Knowledge. Files you upload so it can answer from your stuff, not just its training. Your style guide, your past work, a product manual.

• Tools. Sometimes. The ability to actually do something, like call an API or run code.

So: instructions tell it how to act. Knowledge tells it what to know. Tools let it act.

Most of these wrappers stop after the first two. The third one is where they split apart, and I’ll get to that.

CustomGPTs

OpenAI started this race in late 2023.

You give a GPT some instructions, upload a few files as knowledge, switch on capabilities like web browsing or image generation, and you can wire it to an external API with something called Actions. Then you share it with a link.

A simple example. Say you run a shop. You could build a GPT that knows your return policy, answers in your tone, and checks live order status through your store’s API. Set it up once, then everyone just chats with it.

OpenAI built a whole GPT Store around this. You’d build a clever GPT, people would use it, and you’d get paid. A little app store for prompts.

There are thousands of GPTs now and almost nobody finds them. I wrote about the GPT Store when it launched.

As a tool for yourself, a CustomGPT is fine. You can shape it to your liking and it remembers the setup.

Gems

Google’s version is the simplest of the three.

A Gem is a saved set of instructions for Gemini, with optional file uploads for context. Google ships some premade ones (a coding partner, a writing editor, a brainstorm buddy) and you can make your own.

Think of it as a chatbot with a job description pinned to it. A Gem that always edits to your style rules. A Gem that knows your coding stack so you stop re-explaining it. You set the briefing once, then talk to it.

That’s basically it. No store. No API actions for most people. It lives in the Gemini app and inside Google Workspace, so a Gem can lean on your Docs and Drive if you let it.

If a CustomGPT is a saved prompt with extras, a Gem is a saved prompt with fewer extras. Lighter. Less to configure. Also less it can do.

Skills

Anthropic’s Skills are the one that’s a bit different. They’re also the newest, from late 2025.

A Skill is a file, or a set of files, or a folder. Inside sits a set of instructions, and next to that you can put scripts, code, and resources the model can use.

Two things make them stand out.

First, Claude loads a Skill only when the task calls for it. You don’t pick it from a menu. Claude reads a short description of each Skill you have, and pulls in the full thing the moment the work matches. So you can have a stack of them sitting there and the right one shows up on its own.

Second, the tools part. A Skill can carry real, runnable code, and Claude can execute it. There are Skills that build an actual Word document, fill a PDF, or run a script against your data. Not describe the file. Make the file. So a Skill doesn’t just say how to do a thing. It can do the thing.

They work across Claude’s apps, the API, and Claude Code, the coding tool I use every day.

In December 2025 Anthropic published the Skill format as an open standard. Which means a Skill you build isn’t trapped inside Claude. Other tools that adopt the standard, like Cursor, can run it too. That’s the opposite of how a CustomGPT or a Gem works.

A CustomGPT and a Gem mostly talk. A Skill can act. And it can travel.

The rest

Everyone else has a flavor of this too.

Microsoft has Copilot agents and Copilot Studio, aimed at companies wiring AI into their internal systems. Mistral’s Le Chat has agents.

And then there are Projects in ChatGPT and Claude. Custom instructions buried in your settings. Those are versions of the same trick. A standing instruction the model reads every time, so you don’t repeat yourself. With the chat memories.

Talk or do

If you want one way to sort all of this, here it is.

Does the wrapper just talk, or can it run something.

Most of them just talk. A CustomGPT with knowledge files is a better-briefed chatbot. A Gem is a better-briefed chatbot. Useful, but you could get 90% of the way there by pasting a good prompt yourself.

The ones that can run code, like Skills, or a CustomGPT hooked to a real API through Actions, are a different category. They reach out and change something in the real world. That’s the part worth caring about.

The second question is whether you’re stuck. A Gem only lives in Gemini. A CustomGPT only lives in ChatGPT. A Skill, now that it’s an open standard, can move. If you’re going to invest hours building one of these, that matters more than it sounds.

Everything else is presets.

Downsides

Lock-in first, because for most of them it’s the big one.

Your CustomGPT only works in ChatGPT. Your Gem only works in Gemini. Build your workflow on one of those and you’ve quietly chosen that company for as long as you use it. Skills are the exception, since the format is open now, but they’re still the new kid and the ecosystem around them is young.

Second, a wrapper is only as good as the model under it. A clever costume on a weak model is still a weak model. You’re not buying intelligence. You’re buying a configuration.

Third, the knowledge goes stale. You upload your files once and forget them, and six months later the assistant is confidently working from an old version. Nobody sends you a reminder.

The Bottom Line

These are not five inventions. They’re one idea with five names.

Instructions, maybe some files, maybe the ability to actually do something. Pick a name based on which AI you already pay for, not because one of them is secretly a different species.

The only split that means anything is whether the thing can run code or only chat, and whether it can leave the building it was born in. That’s what I’d ask before building anything on top of one of them.

The rest is branding.

A writer is nothing without a reader. If you found this helpful, consider becoming my dear email friend. Nothing would make me happier.

I run Apple everything. The phone, the watch, the MacBook, the iCloud subscription… everything.

But when I want to set up a small team, I can’t reach for Apple. Not once. Because there is nothing to reach for.

Microsoft 365 and Google Workspace are not office suites. They are operating systems for organizations. Email, chat, video, docs, files, calendars, tasks, and one admin who controls all of it from one screen.

Apple has none of that. It has lovely individual apps for one person. Free mostly. That is a different thing.

The suite

Mail, Calendar, Notes, Reminder, iCloud, Pages, Numbers, and Keynote are good. They are free (mostly).

iWork as it was called once even does real-time collaboration over iCloud. Two people, same document, both typing. It works. Not like Google Docs, but still.

It falls apart, however, the second a team gets involved. The web versions are weak. Everyone needs an Apple account. And the rest of the business world lives in .docx and .xlsx, where Apple’s export is… basic. Microsoft Office is still a thing for a reason.

Google Docs solved sharing in 2006. Apple is still catching up to that, in 2026.

Email

iCloud Mail is consumer email. You can attach a custom domain through iCloud+ now, up to five of them, shared with your family plan.

That is not business email. There is no admin console, no mailbox management for staff, no retention rules, no security policies, no group aliases that someone can actually administer.

Google Workspace gives you all of that for a few dollars per user. Apple gives you a nicer inbox for yourself.

Chat

iMessage is a great consumer messenger. Blue bubbles, encryption, the works.

It is also Apple-only and person-to-person. There is no Apple version of Teams or Slack. No channels, no threads tied to a project, no way to add the freelancer on Android, no admin who can pull a leaver out of every conversation at once.

For a team, iMessage is a group chat with your friends. That is all it wants to be.

Video

FaceTime got links. It runs in a browser now, even on Windows and Android if someone sends you a link.

Still consumer. No scheduling tied to calendars across an org, no recordings stored where IT can find them, no waiting rooms, no host controls that a company would sign off on. It is a call with grandma, not a Monday standup with twelve people and a shared deck.

Tasks

Reminders is really nice for me. Shared lists, locations, the sections. I use it daily.

But it is personal task management. There is no project view, no tracking it across a team, no real board, no reporting. It is a grocery list with a kanban view, and I mean that with love.

Compare it to what sits inside Microsoft 365, which ships Planner, To Do, Lists, and Project, all talking to each other. Apple ships one app that reminds you to buy milk.

A good task manager and project management tool are two of the most important thigns in business settings.

The admin console

Microsoft and Google both give you one place to manage every human in the company. Add a person, they get email, chat, docs, and a calendar in one click. Remove a person, all of it dies at once. Set a password rule, it applies to everyone. See who has access to what.

Apple has Apple Business Manager. But that manages devices. It deploys iPhones and Macs, hands out apps, sets up Managed Apple IDs. It does not run your email, your meetings, your files, or your team’s work.

Apple manages the hardware. The actual work happens somewhere else, and that somewhere is always Microsoft or Google.

Does Apple even want to?

Apple does not need an Apple Workspace. Their whole model is selling you a device. Software exists to make the device worth buying, not to become a service business with support tickets and uptime promises for some company in Ohio.

An enterprise suite means contracts, SLAs, a sales team, compliance audits, and answering the phone at 3am when email goes down. Apple has never wanted any of that. They want to sell the best laptop in the room and let you run whatever suite you like on top of it.

Apple wins the hardware and lets Microsoft and Google fight over the software (in enterprise). Both of those suites run beautifully on a Mac. Apple gets paid either way.

So this is not really a flaw. It is a choice. I just wish the choice were different (for me). But it is how it is.

Europe is looking, and not at Apple

There is one more place this shows up. When European governments talk about getting off American software, they reach for Nextcloud, LibreOffice, Open-Xchange, Matrix. The whole sovereign tech conversation is happening right now.

Apple is never in it. Not because Apple is American, but because Apple has nothing to offer a public office that needs a managed suite for dozens of people. There is no product to even consider.

When the world goes shopping for a 365 replacement, Apple is not on the shelf.

The Bottom Line

I love Apple’s apps. As one person, I want for almost nothing.

The moment there are two of us, I open Google. Three of us, I think about Microsoft. Apple builds the best individual rooms I have ever worked in.

It just never built the office building.

A writer is nothing without a reader. If you found this helpful, consider becoming my dear email friend. Nothing would make me happier.

I wrote about the best productivity suite for a small team recently. Google, Microsoft, Apple, build-your-own. And one European option I gave about three sentences.

That option was kSuite. Three sentences is not enough.

Most people have never heard of it. But it’s a pretty cool option. So, let’s look into kSuite.

What is kSuite

kSuite is a productivity suite made by Infomaniak*, a Swiss company. Think Google Workspace, but the whole thing lives in Switzerland.

You get the parts you’d expect. kMail for email. kDrive for files. Calendar and Contacts. kMeet for video calls. kChat for team messaging, which is the Slack and Teams slot. And OnlyOffice built in, so you can open and edit Word, Excel and PowerPoint files (docx, xlsx, pptx) right in the browser, two people at once, the way you do in Docs.

So it’s not a single app. It’s the full office stack a small team runs on. One login, one bill, one admin panel.

There’s also an AI assistant baked in now, called Euria. It runs on open models hosted by Infomaniak in Europe, not piped off to OpenAI. I haven’t leaned on it much, so I’ll leave it there.

Also, kSuite has more tools, like SwissTransfer which is WeTransfer but safe, Chk which is a URL shortener, kPaste which transfers notes safely, and more.

In short, kSuite has a lot. For cheap. Safe and secure.

Who makes it

Infomaniak has been around since 1994. Geneva. They run their own data centers, they’re not just renting space on AWS and putting a Swiss flag on it. They host websites, email and cloud for a lot of Europe already.

And in May 2026 the founder handed the majority of the company to a Swiss public-interest foundation. Which sounds like a footnote, but it means the company can’t be bought by Google or Microsoft or some private equity fund that wants to gut it. It’s locked. That’s rare in tech now.

The green angle is real too. They cool the servers with outside air, no air conditioning, since 2013. Their newest data center recycles all of its energy as heat and warms around 6,000 households nearby. Whether that matters to you is your call. But it’s there, and it’s not greenwashing.

The storage

This is pretty dop.e

kSuite Pro is about a few euros per months and gives every user 1 TB of storage. Google Workspace starts around 7.99 euros and gives you 30 GB.

That’s a lot more space for a couple of euros. I had to read it twice.

For a writer, a designer, anyone sitting on photo libraries or video, that is a lot of cloud space. 30 GB fills up quickly. 1 TB does not. And email isn’t even included here, because you get unlimited space for emails on top.

There’s a free tier* too, called my kSuite. 20 GB for mail, 15 GB for kDrive, the apps included. The catch is no email aliases on the free plan.

Where the data sleeps

Google and Microsoft are American companies. Under the CLOUD Act, US authorities can ask them for data even when it sits on European servers.

For most people that’s an abstract worry. For a tax office, a doctor, a law firm, a public agency, it’s not. And this fact is getting more relevant by the day with the EU trying to be more independent from US tech and life.

kSuite is Swiss. The data is processed in Switzerland, under Swiss and EU privacy law, outside the reach of the CLOUD Act. Infomaniak doesn’t scan your mail to sell ads, because ads aren’t their business. You’re the customer, not the product. And Swiss data protection is even more strict than the EU’s… which is already damn strict.

This used to be a niche concern. It isn’t anymore. Germany and Denmark are moving public offices off Microsoft, and procurement departments across Europe are asking the same question now.

Downsides

It’s not Google, and you’ll feel it in some areas.

OnlyOffice is good, but on a very heavy spreadsheet or a huge document it’s a step slower than Google Docs or Office 365. If you live in complex Excel macros, some VBA stuff doesn’t carry over cleanly.

kChat does the job, but it lacks some of the automations and integrations Slack people are used to. It’s a chat tool, not an app platform.

And the big one… nobody knows it.

Your clients have heard of Gmail and Outlook. They have not heard of kSuite. So when you send a kMeet link or a kDrive share, expect a few “is this safe?” replies at first. The polish is also a half-step behind Google in small ways. Not dramatic. Just… a bit.

None of that is a dealbreaker. You trade a bit of slickness and brand recognition for your data and a mountain of storage.

The Bottom Line

If you want a good productivity suite for you or if you run a small team and you’ve been uneasy about renting your entire office from an ad company, kSuite is a great option now.

It won’t impress anyone with brand names. It’ll just give you mail, files, docs, calls and chat that work, for less money, with a hundred times the room, and Switzerland between your files.

I think that’s worth a free account* to find out. The worst case is you go back to Google and you’ve lost nothing.

this is an affiliate or SparkLoop partner link. I may earn a commission.

What if you ran a small team of people? A small team.

Five people. Maybe ten. A web agency, a tax office, a dental practice, a tiny startup.

What’s the best tool package for small teams? The best productivity suite?

I don’t know. But I can tell you the options.

Let’s go through them.

Google Workspace

For most small teams, this should be the default. Although in many places around the world, Microsoft is still the default.

But the default should be the default for a reason. And many reasons point to Google.

Gmail is the email everyone knows. Docs, Sheets and Slides open in a browser, sync instantly, and let two people type in the same file without anyone emailing “final_v3_REALfinal.docx” around. Drive holds everything. Meet does the calls. It starts around 7 dollars per user per month.

It works fast, reliable and user-friendly. It also scales pretty well.

The collaboration is the whole point. Real-time editing was Google’s idea long before Microsoft caught up, and Google still does it best.

If your team is non-technical and just needs to work together without thinking about it, you stop here. You really do.

The catch is that you are renting your whole office from an ad company. Google reads less of your business data than people assume, but they still do and the soft lock-in is there. Once 40 people live in Gmail and Drive, leaving is a project, not a decision.

Microsoft 365

Let’s be real, most bigger companies and most older companies (at least in Germany where I live) run on Microsoft entirely. Still. Always have, probably always will.

If your team lives in Office files, it’s pretty hard to convince them otherwise. Even though it’s all technically possible now.

But teams with MS background usually don’t want a browser version of Word. They want the actual Word, with every track-change, mail-merge and pivot-table feature.

Microsoft 365 Business Standard runs around 12 dollars per user per month and bundles a lot. Teams for chat and calls. Outlook. SharePoint and OneDrive for files. Planner for tasks. On paper, it does everything Google does and more.

On paper.

In practice, it’s heavier. Teams is a famously bloated piece of software that tries to be Slack, Zoom, a file server and a project tool at once, and does none of them as cleanly as the apps it copied. The admin center is a maze. Things break in ways that need an IT person.

But if you are a Windows shop, or you send complex spreadsheets to clients who also use Excel, the compatibility is worth it and the last straw Microsoft has, honestly. Office is still a thing for this exact reason, and I wrote about exactly that a while back.

I just don’t enjoy using it. (You might. Many people do.)

What I miss most about the MS suite is a robust task manager, because MS planner is a joke. But to be fair, Google isn’t better here. Google Tasks is even less capable.

Apple

Apple does not make a productivity suite for teams. I wonder why because they have all the pieces in place and could easily build a business suite.

I like Apple products a lot. I write on a Mac. I have an iPhone, AirPods, HomePods, Apple TVs… you get the point.

But iCloud Drive, Pages, Numbers and Keynote are built for one person, or maybe a family. There is no real shared admin, no proper team permissions, no shared inbox.

Apple Business Manager exists, but it’s a device and app management tool. It deploys Macs and iPhones to staff. It is not where your team writes documents and runs email.

You can’t run a five-person office on iCloud. Unfortunately. Because it would be so cool. Notes added to that and Reminders (maybe a more capable business version with real task management for multiple people), then passwords for security, and more.

Apple has it all, they could be the business suite for all teams with Apple-devices only. And they could be cheaper than all the other productivity suite too. I’d go that route any day over Microsoft or Google.

Notion, Slack, Zoom

Then there’s the build-your-own-stack idea. Notion or Anytype for docs and wikis. Slack for chat. Zoom for calls. Maybe Gmail still in the background for email, or Proton Mail or Fastmail or whatever you want.

Each of these tools is the best at its one job. Slack is much nicer than Teams. Zoom still beats Meet on the hard calls. Notion is a great brain for a company if you can get people to use it.

But adding it up, you’re paying for four or five subscriptions instead of one. Notion, Slack and Zoom each land somewhere around 8 to 15 dollars per user per month once you’re on a paid plan. Stack them and a single person costs more than a full Microsoft license that already includes all of it.

And nothing talks to each other cleanly out of the box. Your files are in three places. Your team forgets which tool a decision lives in. Onboarding a new hire means five logins, not one.

It’s the best experience and the worst maintenance. For a small team without an ops person, I’d skip it.

The European option

In the EU, things got interesting recently.

If data sovereignty matters to you, and in Europe it’s starting to matter to procurement departments, not just activists, there are now suites that don’t route your business through American servers.

Infomaniak’s kSuite is a close thing to a Google Workspace replacement that stays inside Switzerland. Mail, drive, docs, calendar, video calls. It has a free tier and paid plans, and it’s a real company with real data centers, not just AWS rentals.

Nextcloud is the other route. It’s open source, you host it yourself or pay a provider to host it, and you own the whole thing. Files, calendar, contacts, office documents through the built-in editor. Germany and Denmark are moving public offices onto exactly this kind of setup.

The catch is obvious. Nextcloud needs someone who can run a server, or a provider you trust. kSuite is smaller, so the apps are a step behind Google on polish. Your clients might never have heard of either.

But the option is there now, and it wasn’t really there two years ago. That’s a good start.

Migration

Every one of these has the same problem… migration.

Picking a suite is easy. Leaving one is the hard part. Move 20 people off Google after three years, and you’re untangling shared drives, calendar invites, group emails and a hundred small permissions nobody documented.

So the question isn’t “which is best.” It’s “which one will I least regret being stuck inside in five years.”

For a Windows-heavy team that sends Excel files to clients, the answer is Microsoft, even if Teams annoys you daily. Unfortunately.

For everyone else who just wants to work together, it’s Google, even with the ad company in the background.

For a team that cares where its data sleeps, it’s worth the extra friction of going European.

And if they ever do it, for Apple fans a true Apple business suite would be ideal.

The Bottom Line

Most small teams should probably use Google Workspace.

The rest will still use Microsoft 365. Don’t try to run a team on Apple yet, no matter how much you love your Mac. Skip the Notion-Slack-Zoom stack unless you want to pay a lot more.

And if you’re in Europe and the question of who can read your files keeps is key, look at kSuite or Nextcloud before you decide.

Ten years ago, building a website was a real project.

You either learned to code it yourself, which took months, fiddled with web builders yourself, which were not really that great back then, or you paid someone a few thousand euros to do it for you.

None of it was fun.

That world is gone. In 2026, you can put up a real website for free, in an afternoon, with no code. You can build a full web app for the price of a couple of coffees a month. If you are running anything online, and you still don’t have a website, that is now a choice, not a limitation. And sure, that choice is yours.

But having a website is not optional the way it used to be. It is the one piece of the internet you actually own. Instagram can change the rules tomorrow. Your domain can’t.

So here is how I’d do it now, depending on what you need.

The free way

If you just want something, start with a link-in-bio site.

I use Carrd* for this. But you can also you Pop Site. Or other free small web builders that make it really easy to start.

It is free to start, and free actually means free, not a 7-day trial that bills you later. You pick a layout, drop in your links, your socials, a contact form, maybe a little about section, and you are live.

You can have something working in fifteen minutes.

It does not have to be a boring list of links either. People use it for portfolios, small business pages, a single product. Then you point your Instagram, your LinkedIn, your X bio at it.

The paid tiers are mainly for a custom domain and more pages. For most people who just need a presence, this is where I’d start. There is no reason not to.

Framer

If you want something that looks designed, Framer is nice.

It is a visual builder. If you have ever touched Figma or Canva, you will get it. You drag, you drop, you publish. Except the output is a real website with hosting, a CMS, and SEO built in. It’s not as easy. But it easier than anything 10 years ago.

What Framer and its large template section do well is aesthetic. The templates on their marketplace look like studio work, the kind a small agency would charge you five figures for. You grab one, swap in your content, and you are done in a day.

For a business, this is a good pick. A restaurant, a studio, a freelancer who wants to look sharp. Ten dollars a month for a basic site, thirty for the pro tier. If you run a business, thirty a month to have something that just works and never breaks is nothing.

You don’t think about hosting, a CMS, or where the search traffic goes. It is all in one place. That is most of the value.

WordPress

People love to hate WordPress. I do too. Bloated, clunky, too many plugins.

I still recommend it, and the main reason is wordpress.org which you can set up for free on your own domain and web host.

I’ve built client sites on it for years. The reason it sticks around is the reason it works. Whatever you want to do, someone has already done it and written it up. A plugin exists. A guide exists. When you hit a wall, the answer is one search away. No other platform has that.

And it scales with you. A simple blog or a full e-commerce shop. Same tool.

You pay for hosting, and most hosts have a one-click WordPress install so you never touch a server or a database. Pair it with an AI agent now and the maintenance is easier, too.

AI agents

Coding agents like Claude Code and ChatGPT’s Codex have gotten so good at websites. You describe what you want in plain language and they build it in minutes. And the designs are getting better too.

I do have coding knowledge, so I’m not the typical “I can’t code at all” case. But I lean on these agents constantly now, and the gap between what a non-coder can ship and what a developer can ship has shrunk a lot.

You point Claude Code at a folder, ask for a clean personal site, answer a couple of questions, and it hands you something pretty decent. Half the buttons won’t link anywhere yet. But the structure is there, and you edit from a working draft instead of a blank page.

Then it goes further than a website. I’ve built apps that handle payments, user accounts, password resets, the whole boring backend that used to need a small team. The agent wrote most of it. It works.

The limit now is your imagination and your marketing. Building got easy. Marketing is the hard part. Still.

The catch

Nothing here is perfect.

Free tiers want you to upgrade. That is the deal. Carrd, Framer, all of them dangle the paid plan, and the costs add up if you stack three tools instead of picking one.

Framer locks you in. Your site lives in their builder. Leave the platform and you are rebuilding from scratch.

WordPress needs care. Updates, the odd broken plugin, a backup habit. Less now with AI helping, but it is not zero.

And the AI agents will lie to you with total confidence. They ship code that looks right but doesn’t work like you want it to. You still have to test what comes out. If you can’t read it at all, it’s harder.

But it’s all doable. For various skill sets and pockets.

The Bottom Line

The tools stopped being the reason.

For most of what you’d want online, there is now a free or cheap path that a normal person can finish in a weekend. Carrd or Pop Site for the simple landing page. Framer if you care how it looks. WordPress if you need low price and scalability. An AI agent if you want something custom.

I started when this was hard and expensive. It really isn’t anymore.

*this is an affiliate or SparkLoop* partner link. I may earn a commission.

On June 9th, eight European tech companies are launching Euro Office. A sovereign, open-source replacement for Microsoft 365.

Built in Europe, governed in Europe, stored in Europe.

What is Euro Office

Simple: Web-based editors for documents, spreadsheets, and presentations. Real-time collaboration. Full support for Microsoft file formats: DOCX, PPTX, XLSX. And Open Document Format.

Built on a fork of OnlyOffice, an existing open-source office platform. The interface is deliberately familiar. If you know where the bold button is in Word, you know where it is here.

The coalition behind it: Ionos, Nextcloud, XWiki, OpenProject, and several others covering hosting, file management, project management, and knowledge systems.

It goes live on GitHub for free on June 9th. It will also be integrated directly into Nextcloud Hub 26.

They announced the project in March and hit a stable 1.0 release in under three months. It’s urgent.

This moment in time

Germany spent €481 million on Microsoft licenses in 2025. Up 76% in just two years. One country, one vendor, one year.

And Germany is full of Microsoft software in every corner of every office.

Across all European governments, the estimated annual spend on American productivity software: over €2 billion.

But the money is almost secondary now.

In June 2025, a Microsoft executive testified under oath in a French Senate hearing that the company could not guarantee data sovereignty if American authorities demanded access under the Cloud Act. France responded by announcing 2.5 million civil servants would migrate off Teams, Zoom, and WebEx by 2027. Then France went further and announced it was planning to ditch Windows entirely for Linux.

Same in Schleswig-Holstein, where I live. I wrote about this when they started. 40,000 civil servants, are migrating off Microsoft’s entire stack: Office, Teams, Outlook, Exchange.

Transition cost: €9 million. Annual savings: €15 million.

Payback period: less than a year. In 2026, roughly 80% of workstations should be completed switched. The rest are on schedule.

Can small businesses use this too

So, Euro Office for the public sector. That’s all well and fine. But what about the millions of small companies running entirely on Microsoft. Can they benefit from Euro Office?

Probably, yes. With caveats.

Euro Office is free and open source. If you have a technical team, you can self-host it. If you don’t, providers like Ionos will presumably offer hosted versions. No GitHub skills required.

For a small team paying €12–25 per user per month for Microsoft 365, the math is fairly simple. Especially if what you mainly do is write documents, build spreadsheets, and share files.

The file format compatibility matters, of course. DOCX, XLSX, PPTX support means you can still exchange files with clients who use Microsoft products.

What you’d lose in v1.0: deep Active Directory integration, the enterprise admin console, advanced macro support. For a five-person agency or a small consultancy, those things probably don’t come up (yet).

What comes up is: can I open this Word file? Can I share this spreadsheet? Can we edit it at the same time?

Those, Euro Office does.

Worth trying before dismissing. Especially given what the alternative costs each month.

It doesn’t need to win

Euro Office doesn’t need to replace Microsoft 365 for most organizations.

It just needs to exist. For free. And without much friction. That will over time help people make the switch, especially the younger generations.

I wrote about this same logic with Wero, the European payment network. Wero doesn’t need to kill Visa and Mastercard. It just needs to exist so European payments can’t be switched off from Washington. Euro Office is the same idea applied to productivity software. It’s also cheaper.

For Euro Office, the demand signals are there, too. Nextcloud saw customer inquiries triple in 2025. In the Netherlands alone, up 8x. Austria’s Federal Ministry of Economy adopted Nextcloud. The city of Stuttgart migrated. Sweden launched a sovereign Nextcloud workspace for the Nordics. Nextcloud has committed €250 million in sovereign infrastructure investment through 2030.

An IDC Europe survey found 45% of European organizations said their interest in digital sovereignty solutions had grown specifically because of geopolitical instability in 2025.

Denmark’s digitalization minister put it plainly: “Europe must never make itself so dependent on so few that it can no longer act freely.”

The downsides of Euro Office

This is v1.0.

Microsoft has three decades of integration, enterprise tooling, and muscle memory on its side. The admin tools. Active Directory. The Excel macros that half of European finance runs on in the background. Euro Office does not have those yet.

Migration is work. Moving off Microsoft means changing habits, retraining staff, and rebuilding workflows that have been running on autopilot for years.

That being said, autopilot here means a lot of issues and problems with Microsoft tools. And those are not getting better.

Microsoft will probably respond by accelerating its “sovereign cloud” offerings in Europe. But the Cloud Act doesn’t go away because you put a European flag on a data center. A US court order still reaches the server room.

Lastly, Euro Office alone isn’t the complete answer.

Most of the infrastructure it runs on still sits on Amazon, Microsoft, and Google cloud. The layer underneath hasn’t been solved yet.

The Bottom Line

Euro Office won’t fix everything. It’s 1.0. It will be rough in places. The enterprise features will take time.

But it exists. June 9th. Free on GitHub. That’s a good start.

For small businesses: it’s worth a look before dismissing it as a government thing. The costs are zero, the file compatibility is there, and the alternative keeps getting more expensive.

A security researcher published six Windows exploits between April and May 2026. Several were being used by real attackers within days. Three still have no complete fix.

If you use Windows, this affects you. More than a billion people still do.

What means zero-day

Zero-day means zero days to fix it before it’s public. The patch doesn’t exist yet. Attackers already have the exploit.

The specific flaws here are LPE vulnerabilities. Local Privilege Escalation. If someone already has access to your machine, they can use these to take full control. Admin rights. System-level access. Everything.

Windows Defender and the Windows kernel are both affected.

The six exploits

The researcher goes by Nightmare Eclipse (also Chaotic Eclipse, Dead Eclipse). Starting April 2, 2026, they published six tools in roughly six weeks.

Bluehammer exploits a timing flaw in Windows Defender and grants full system control. This one was patched in the April 2026 Patch Tuesday update. It’s also on CISA’s Known Exploited Vulnerabilities list — confirmed used in real attacks.

Red Sun abuses a Defender file-recovery feature to inject code at the highest system level. No patch.

Undefend silently freezes Defender’s signature update pipeline. No security warnings appear. Your protection just stops working in the background. Also unpatched.

The remaining three — Yellow Key, Green Plasma, and Mini Plasma — have no official fixes either. Microsoft has flagged Yellow Key as “more likely” to be exploited in practice, since a working proof-of-concept is already public.

Hunter Labs confirmed active exploitation of Bluehammer, Red Sun, and Undefend by April 10. The attack chain: compromised Fortigate VPN credentials for initial access, then these exploits to escalate privileges once inside.

How we got here

Security research has a standard practice. You find a flaw, you report it privately, the vendor gets time to fix it, then everything goes public — usually around 90 days, ideally alongside a patch.

That’s not what happened here.

Nightmare Eclipse claims Microsoft’s Security Response Center ignored their reports. Their MSRC account was suspended. Reportedly, MSRC asked for a video demonstration of the exploit before they’d even look at it — an unusual requirement that drew criticism from the broader security community.

On March 26, the researcher posted a public warning: the exploit would be released if MSRC didn’t respond. According to their account, MSRC didn’t. On April 2, Bluehammer was public.

Five more followed.

Microsoft’s version is different. The MSRC blog states the vulnerabilities were not shared with Microsoft before publication, and that the disclosures put customers at unnecessary risk. Uncoordinated, unjustifiable — that’s the MSRC framing.

Both things can be partly true. The researcher’s approach was reckless. Microsoft’s process may have failed before that point. I don’t know what actually happened between them. I doubt many do.

Microsoft’s response

Swift. And not subtle.

Microsoft deleted Nightmare Eclipse’s GitHub account. GitHub belongs to Microsoft, so that’s within their power. The GitLab account followed shortly after.

The MSRC statement went further: Microsoft said it would pursue publishers of uncoordinated exploit disclosures, including through cooperation with law enforcement worldwide.

That’s a threat of prosecution.

The problem with threatening researchers

Legal action against a security researcher signals just one thing… stupidity.

A legal guide from the Harvard Law School Cyberlaw Clinic and the Electronic Frontier Foundation puts it plainly: organizations that threaten researchers with lawsuits see fewer responsible disclosures in the future. Researchers stop reporting. Or they report to someone else. Or they sell to someone you really don’t want to have the exploit.

Organizations with mature security programs generally don’t go the legal route. Because they understand what that costs them long-term.

MSRC was historically one of those mature programs. A serious institution, decades in the making. Which makes this response harder to explain.

What you should do right now

If you’re on Windows, a few things are worth checking.

The April 2026 Patch Tuesday update covers Bluehammer. If you haven’t installed it, do that now. Also make sure Windows Defender’s platform is current, not just the definitions.

For Yellow Key, there’s no full patch yet. Setting a BitLocker startup PIN and a BIOS/UEFI password limits the attack surface in the meantime.

For Red Sun and Undefend, there’s no complete fix. Apply updates as fast as possible when they arrive, and watch the next Patch Tuesday. That’s about it.

In enterprise environments, it’s worth checking whether any Fortigate VPN credentials were compromised. That was the confirmed entry point in the known attacks.

And mark July 14, 2026. Nightmare Eclipse has announced another disclosure that day, patched or not.

The Bottom Line

Publishing unpatched exploits harms real people. There’s no way around that, and Microsoft is right about that part.

But a company that responds to a researcher with account deletion and legal threats, rather than communication, isn’t making anyone safer. It also doesn’t fix any of the issues.

If MSRC’s process is broken, fixing it matters more than winning this particular fight. Because there will be a next researcher. And they’ll be watching how this one ended.

I personally stopped using Windows. This situation wasn’t the main reason. But I sure as hell am glad I did.

I already wrote about why I won’t touch Windows again.

Mostly because of macOS, and a bit Linux. But there’s a third option. And that is going to be big.

Android.

Google recently announced the Google Book. Android powered, AI first, built with Dell, Lenovo, Asus, Acer and HP, the usual Windows partners (and Chromebook partners).

The system inside is called Aluminium OS, based on Android 17, and it leaked in a 16-minute walkthrough before the official reveal.

On its own that would be a niche story. ChromeOS has been niche for years.

But layer it on top of what macOS and Linux already cover, and how Microsoft is screwing everything up with Windows 11, and the question becomes: What is Windows actually for in 2026 and beyond?

The entry tier

The cheap laptop tier belonged to Windows. Schools, students, low-budget office work. Some used Chromebooks, but that’s a very small percentage. The rest. Windows. Macs were always too pricey in that equation.

But now Google is going after that tier with Android on a laptop. Hard.

Real Android. Billions of users already know it. The app ecosystem exists. The Google Book runs Aluminium OS with a Windows 11 style taskbar, a desktop, real document folders, a task manager. It has a Link to iOS app to mirror iMessages.

And the AI is in from the start. Not bolted on. Not shitty like CoPilot. You shake your cursor and Gemini activates. You highlight something on screen and ask a question. You type “share photos from the Japan trip” and it pulls from Google Photos locally and drops them in your chat.

If that works as shown, it’s the first time someone made an AI feature on a laptop that I really works and helps.

Microsoft tried this with Copilot. Copilot is in Edge, in Notepad, in the taskbar, in Outlook, asking to summarize my Word docs and recall what I did yesterday. I didn’t ask. And much of it is not helpful at all.

Google is doing what Microsoft tried to do. But Gemini is just integrated better.

The power tier

If you need real work, performance, battery life, and long-lasting machines, you go macOS. That’s been true for years. I just wrote about my M1 Pro MacBook Pro still being great five years in.

Apple silicon. Battery life. A fanless laptop that compiles faster than a desktop did three years ago. Software that mostly respects the user.

You can argue about the Apple tax. You can argue about repairability. But for people who need their laptop to be a tool, the choice is made. And it’s similar for the Mac desktops like Mac Mini or Mac Studio. Even the iMac.

Microsoft has Surface. Surfaces are fine. But “fine” doesn’t cut it anymore. And they’re also expensive.

The tech-savvy tier

If you want control, if you want to tinker, if you don’t want any of the rest, Linux is right there.

Mint. Fedora. Arch. Zorin OS if you want something that looks like Windows but doesn’t behave like it.

The “Linux is hard” argument will be less and less of an issue for the more and more tech-savvy generations. Install times beat Windows. Updates don’t restart your machine mid-presentation. No ads in the start menu.

Used to be the nerd tier. It still is, kind of. But the nerds who used to dual boot now just install Linux and move on.

So what is Windows for

The cheap laptop user gets a Google Book with AI that works. I actually believe that GoogleBooks with Android baked in will be a huge thing. For kids, for schools, for the average, non-techy user that needs browsing and email, and many more.

The professional gets a Mac. Creatives, modern offices, companies that don’t want the IT hassle of Windows anymore.

The power user gets Linux. Full control. No cost.

That leaves Windows with two real groups. Enterprise IT departments locked into Active Directory. And gamers, who will start looking at a Steam Deck or a console sooner than later, anyway.

So the real last group: Habit.

People who learned Windows in 1998 and never had a reason to leave. That’s a real group. It’s also a shrinking one. Heavily shrinking.

I’m getting ahead

Aluminium OS is not out yet. The leaked demo ran slow in a virtual machine on a MacBook M4. Boot times were long.

The Google Book partners are Dell, HP, Lenovo. We know what they do to operating systems. Bloatware, trial versions, vendor apps you can’t remove. The Android phone market is full of it, too. The laptop market won’t be cleaner.

And professional software is the real wall. Photoshop and Lightroom on Android exist, but they’re cut down. Final Cut, Logic, DaVinci Resolve, the full Adobe suite, Excel with real macros.

None of that runs on Android (yet or well).

A Google Book in 2026 is for people who live in a browser and a few apps. Which is most people. Just not all people.

The Bottom Line

It’s really not about the GoogleBook in the end. The interesting fact is that every tier of the laptop (and even desktop) market now has or will have a good non-Microsoft answer.

Cheap and AI-first. Premium and powerful. Open and free.

Microsoft is the only one trying to be all three at once. But not getting one right. Windows 11 sucks because it doesn’t do anything well. And it has become an ad system. Ads in the start menu because apparently the trillion-dollar company needs the revenue. Copilot everywhere because they need an AI story. Ads for Xbox because… well, Xbox sucks.

I have used ChromeOS, macOS, and Linux. For work and play. I don’t use a Windows machine. Because the experience was just bad, and all the other 3 were (and are) better.

I think nobody under 30 has a reason to choose Windows anymore. And it’ll show. Soon.

The TÜV is the German authority that inspects every car on the road. Every other year. Three years in, if it’s brand new.

The 2026 report just dropped, covering 9.5 million data points.

The Tesla Model Y came in 110th. Out of 110.

Dead last. I couldn’t stop laughing when I heard.

But it’s also not funny. Because our German Tesla Ys are built in Brandenburg. And that’s not good for Brandenburg. Never mind Tesla.

Numbers

The Model Y’s defect rate hit 17.3%. Almost one in five flunked their first mandatory inspection.

No car has done this poorly in 10 years.

For comparison, the cars at the top of the same age bracket:

• Mini Cooper SE: 3.5%

• Audi Q4 e-tron: 4.0%

• Fiat 500e: 4.2%

• BMW i3: 4.6%

European builds dominate the top of the rankings, by the way. By a lot.

The Model 3 came in 108th, with 13.1%. Better than the Y. Still bad. Terrible.

Three Failure Points

The TÜV report names them:

• Axle suspensions buckling under the heavy battery weight.

• Brake discs deteriorating from chronic disuse (regenerative braking does most of the work).

• Lighting system defects.

But this has nothing to do with EVs. It’s a Tesla problem.

Every EV maker deals with the same physics. The Audi Q4 e-tron is heavier than a Model Y and fails four times less. A 5-year-old VW ID.3 is more reliable than a brand-new Tesla.

The Model Y at 2 to 3 years old has a failure rate of 17.3%. Mercedes just won a long-term quality award because their 10-year-old cars averaged 18.5%.

A brand-new Tesla is basically a decade-old Mercedes. LOL.

The Sales Side

The quality story lands during a sales collapse. Tesla’s European registrations dropped 27.8% in 2025. I wonder why… German sales crashed 48% in one year.

And the European EV market isn’t shrinking at all. It’s booming. Strongly.

• Overall EV sales: up 37% in January 2025.

• Tesla: down 45% in the same month.

• Continental market share: 0.8% in early 2026.

The other EV makers are doing fine. Tesla isn’t.

I bought an EV. And I love it. Obviously, it’s not a Tesla.

The Decoupling Pattern

We’ve seen this before.

Defense, where Germany is shifting most of its spend off US suppliers. Payment systems, where Wero is replacing Visa and Mastercard in the eurozone. Cloud infrastructure. Office software.

Now cars.

The Made in Europe Act is pushing domestic EV production hard. Tesla had a 5-year head start. European automakers are now driving past in cars that actually pass safety checks.

Not to mention the boom of Chinese EVs everywhere.

In Fairness

The TÜV only catches the 2-to-3-year-old window in this category. Older Tesla data could shift the picture. Maybe.

Tesla owners report different experiences too, of course. Some are happy. Some had endless service visits. The TÜV doesn’t measure satisfaction. It measures whether the car passes a safety check.

The Bottom Line

Europe spent decades letting US companies handle the critical stuff. Defense. Software. Payments. Cars.

Then Trump came (again), we noticed, and we’re changing.

The Made in Europe push is going strong. And it isn’t slowing done.

Wero is Europe’s payment answer.

A year ago, Wero showed up in my banking app. A new button for sending money instantly across Europe.

But it’s much more than that.

13 countries on paper. 130 million users in theory. The whole continent building its own payment rails.

Numbers

48.5 million registered users in the first 12 months. Germany, France, Belgium. Over €7.5 billion processed.

For a platform that only went live in mid-2025. Pretty solid.

For comparison: iDEAL, the Dutch system that’s getting fully migrated into Wero next year, took about five years to hit equivalent penetration in a single country. Wero did it across three.

Then this year, Revolut joined. That adds roughly 40 million European customers across every EU country where Revolut operates.

E-commerce is live in Germany and France. The Netherlands joins in 2026 with the full iDEAL migration. Luxembourg is next in the rollout queue. Point-of-sale payments (the moment you tap your phone at a shop terminal instead of routing through Visa or Mastercard) start rolling out from 2027.

The Gap

Mastercard alone has over 900 million cards in circulation across the EU.

Wero has 48.5 million users.

German adoption, despite Germany being the very first launch market, sits at roughly 5% of total transaction volume.

Visa and Mastercard together still process about 61% of all Eurozone card transactions. Globally, they move around $24 trillion a year between them.

13 EU member states have no domestic payment alternative at all. Every card swipe routes through American-owned infrastructure.

Wero is not closing that gap in two years. Probably not in five.

Leverage

Wero doesn’t need 50% market share to change the dynamics. It needs to reach the point where it exists as a credible operational alternative. A point where European regulators, banks, and governments can look at Visa and Mastercard and think: we have another option that works. With European infrastructure.

Not total volume. Capability across every payment type.

• Cross-border peer-to-peer between European countries is live this year.

• E-commerce online is live in Germany and France. Others will follow.

• Point-of-sale at physical terminals will come in 2027.

Interoperability with Spain’s Bizum, Italy’s Bancomat, Portugal’s MB Way, the Nordic MobilePay is coming in the first half of 2026 or is being built to do.

Once all that’s running, Visa and Mastercard can no longer set the price by default.

That’s the point.

Fees

Visa and Mastercard charge merchants between 1% and 3% per transaction. Sometimes considerably more on cross-border.

Wero’s fee structure sits around 0.7%.

For a European retailer processing millions of euros in card payments a year, that’s tens of thousands flowing to the bottom line instead of out of the continent.

Luca Rosignan at Capgemini’s research institute put it well to Cybernews: even partial success at capturing domestic European transactions could effectively cap what global card networks can charge here going forward.

Wero needs to exist at sufficient scale to permanently cap what Visa and Mastercard can charge European merchants.

And that seems to be working.

Backing

The European Parliament voted in February to formally back the digital euro, explicitly framing it as essential for monetary sovereignty. ECB president Christine Lagarde has been pushing for European payment infrastructure for a while now. The ECB’s own committee chair publicly warned that Washington can switch off Europe’s entire payment system at any moment, because nearly all of it routes through American networks.

Thirteen countries signed the Europa Alliance. 16 major European banks back Wero directly with capital and customer bases.

This is part of a broader pattern I wrote about earlier this year. Europe is decoupling across multiple tech layers at once. Payments are just the slowest layer to move.

So Wero is a state-backed, ECB-endorsed infrastructure project with legislative support and a geopolitical motivation that didn’t really exist three years ago.

A huge step.

Next steps

The central interoperability hub launches in the first half of 2026. That single technical connection creates a unified network across 130 million users in 13 countries, who can pay each other and pay merchants without a single transaction touching Visa or Mastercard.

E-commerce expands beyond Germany and France into Belgium and the Netherlands. Point-of-sale pilots begin in late 2026. Continental rollout through 2027.

iDEAL, which currently handles the vast majority of Dutch online payments, migrates onto Wero next year. That brings an entire country’s payment habit into the European sovereignty system basically overnight.

With that, Wero might hit 60 million users by the end of 2026. Probably over €15 billion in volume. Roughly double the first year.

Downsides

Adoption is still slow where it should be fastest. 5% of transaction volume in Germany after a full year is not great. Habit is hard. Apple Pay and Google Pay have the ease of use on their side.

The user experience for Wero is fine, but not exciting. There’s no Wero feature that makes you want to switch. It’s a payment system. It works. But it’s not “sexy”.

Cross-border parts are still being wired up. Right now, paying a friend in Italy with Wero from Germany is technically possible but still feels like a beta product. (For me anyway.)

And the political backing is conditional. A shift in the ECB, a new German government, and the urgency could fade.

The Bottom Line

The goal for Wero was never to capture the whole market.

The goal is to make sure that the next time Washington reminds Europe it can flip a switch on the payment rails, Europe can look at the switch and say: go ahead. We’ll have our own.

We’re still a long way from that being fully true. But at least it’s in the making.

I just wrote about why I won’t use Windows again.

Now comes the love letter to the machine that made that decision super easy.

My MacBook Pro is from late 2021. 16-inch, M1 Pro, 16GB RAM. I bought it for solo freelance work back then. Today it runs my new job too. Without complaints.

I haven’t thought about upgrading. Not once.

My new job runs on this thing

Whatever lands on my desk lands on this laptop. Claude Code with multiple agents running, browser automation, deploys, video calls, presentations, the occasional Photoshop. No company machine. No second setup. This one does it.

I’m half done while colleagues are still fixing their Windows updates. I’ve said that before.

16GB and AI workflows

16GB are the low end in 2026.

But on Mac, it’s fine.

I run Claude Code with parallel agents. I have Chrome, Safari, Mail, Notes, a couple of dev servers, Spotify, Wispr Flow, and the AI stuff all open at the same time. macOS swaps when it needs to. I don’t really notice.

I don’t run 70B models locally, of course. But for most AI workflows, it’s more than enough.

A full day, no charger

I work a full day from the kitchen table or an office or my backyard, and I don’t plug in until I go to bed. The battery just goes.

Battery health is above 90% after 5 years. That’s pretty nice. And even with less, it will still run longer than most other laptops on the market.

The screen replaces the monitor

16 inches is big. Bright. Sharp.

I don’t have an external monitor at home anymore. I used to. Don’t have space anymore. And I don’t miss it much.

(For me. People who live in 4-window grids on 32-inch monitors will disagree.)

Work happens wherever I am

I carry one thing.

I worked a full week from my parents’ place last month. The laptop went where I went, did what I needed it to do, and didn’t ask for anything.

No docking station. No power brick I have to carry constantly. No “wait, where did I save that file” because the file is just there. On this machine. Where I left it. (Or in the cloud anyway).

Downsides

It’s a bit heavier than other laptops, sure. 2.1 kg. Backpack-only, basically.

The fans do come on during really long and heavy AI sessions. Audible. Not loud. Just present.

Battery health will drop some more. That’s how it goes.

macOS 27 will support it. The one after that, I’m less sure. So there’s a clock somewhere, ticking with OS updates. But, I mean, any Windows user can talk about the „super smooth” Windows 10 to Windows 11 upgrade.

The Bottom Line

My M1 MacBook Pro has done 5 years of freelance work, all my apps, all my writing, all my videos and music, 3 operating system upgrades, career changes, and AI from the beginning

It hasn’t broken a sweat.

It’s not cheap. Not at all. MacBook Pros are expensive. But you get what you paid for. And that’s a lot. And very rare nowadays.

I’ll keep using it until it stops. And then I’ll feel bad about replacing it.

So for anyone who needs a new computer… just get a Mac if you can afford it. And be happy.

I started a new job that runs on Windows.

I hadn’t touched a Windows machine seriously in years. I figured… how bad can it be. I’ll just adapt.

I adapted. To the fact that I never want to use it again.

The UI

Windows 11 looks like it was designed by three different teams who never met. The Start menu is one style. The Settings app is another. The Control Panel … yeah. Right-click menus collapse what used to be one click into two.

Things load. Then they re-load. Then a context menu opens half a second after you clicked. You start to expect lag as part of the interaction.

The loading

The Search bar in the Start menu thinks for a moment before showing results. Sometimes it shows web results from Bing instead of the file I’m looking for.

Opening Outlook (the new one) takes longer than opening Apple Mail, Notion, and Safari combined on my Mac. And it shows ads in the inbox. More on that in a sec.

The AI everywhere

Copilot is in the taskbar. Copilot is in Edge. Copilot is in Notepad now. Copilot wants to summarize my Word docs. Copilot wants to rewrite my email. Copilot wants to recall what I did yesterday.

I didn’t ask. And much of it is not helpful at all. It’s slow, it suggests things that don’t apply to what I’m doing, and it interrupts the workflow it claims to support. Every productivity app on my Mac integrates AI more quietly and more usefully than Microsoft has managed in their own operating system.

And I work in AI a lot.

The ads

This is the worst part.

• Ads in the Start menu suggesting apps from the Microsoft Store.

• Ads during the setup process pushing OneDrive and Microsoft 365.

• Ads inside the new Outlook, between actual emails, styled to look like emails.

• Ads on the lock screen (“Spotlight,” sure).

• Notifications nudging me to switch my default browser back to Edge.

I paid for this machine. Well, the company did. The company paid for the Windows license. And I’m being marketed to while I work.

I’ve written about how embarrassing this has become before. Sitting in front of it again, on a clean install, reminded me I wasn’t being overdramatic.

Windows just sucks. I don’t get how people can live with that.

Sure, no system is perfect. Not macOS. Not Linux. But man, far from the worst. Both of them. Real far…

Options

I’m not telling anyone what to do. I want to, though.

But I get it. People have reasons. Software they need. Games. Workflows. Habits going back decades. I get it.

For me, the options were already there:

• macOS, if you want it to just work and don’t mind the Apple tax.

• Linux, if you want control and don’t mind tinkering.

Zorin OS is what I’d hand to anyone leaving Windows who doesn’t want to relearn everything from scratch, and doesn’t want Mac. It looks familiar. It runs on old hardware.

Even ChromeOS, for basic browser-based work, is calmer than Windows now. And Android on laptops will be too.

The Bottom Line

There’s really little Windows has going for it, other than the habits of millions of people using it. For decades. That’s it. Nothing else is a pro for Windows. There are only cons. In my eyes, anyway.

I told my new job I’ll be working on my own Mac. Smooth as a breeze. I am half done while the others are still fixing Windows issues.

Three days on Windows reminded me that I’m not picky about operating systems. I’m picky about being respected by the thing I look at for ten hours a day. And paid for.

A line is going around: “Women over 40 just surpassed teen girls in birth rates for the first time in U.S. history.”

Dramatic. Made for engagement. And true-ish.

But like most viral stats, the second look is quite interesting.

Numbers

The National Center for Health Statistics, part of the CDC, put it in a report from spring 2025. The data is from 2023.

So, it’s not actually that new. But the clickbaity headlines are.

Women over 40 had 4.1% of US births that year. Teens had 4.0%.

That’s it. Hair-thin.

Modern US birth records go back to 1933. Teens sat above the 40+ group the whole time. Often by a lot.

The crossing already happened back in 2022. The gap has widened a bit since. The 2025 CDC report made it more visible.

Teen births

Down 78% since 1990. From 59.9 per 1,000 to 13.1. I think, all in all, that it’s not really a bad thing. Most would agree here, I assume.

It’s not “by accident”, obviously.

• It’s decades of public health work.

• Long-acting birth control like IUDs and implants went from rare to routine.

• Sex ed got better. Teens started later. The share of sexually active teen girls dropped from 51% in 1988 to 42% by 2017.

“It tells me a range of strategies is paying off,” said Dr. Claire Brindis at UCSF. Sounds about right.

Older moms

Births to women over 40 jumped 193% since 1990. Births to women 35 to 39 jumped 90%. Those are incredible increases. But not even the striking. That is: Births to women 45 and over went up 450%.

But it also involves some nuance, because that number wouldn’t and doesn’t exist, the 450% increase we’re looking at, without IVF.

Other seasons stack up, too, though. Careers. Money. Education. Reproductive medicine that didn’t exist a generation ago.

Missing pieces

Older births don’t replace the missing younger ones. At least in statistics on paper.

US total fertility sits around 1.6. Replacement is 2.1.

Total US births fell 14% between 1990 and 2023. The “older mom” share is growing inside a shrinking pie.

Two stories

Teen pregnancy prevention worked. People are having kids later.

And the third story is that there’s falling fertility overall.

Also, health officials still warn about the medical part. Pregnancies after 35 carry higher risk. Chromosomal issues, complications. That hasn’t changed. Yet.

Teen pregnancy is down, and that’s good.

People are having kids later. That’s life now.

Well, the picture would also look quite different if official (and believable) data existed from before 1933. Way back. That would tell a totally different story, I imagine.

Apple makes great hardware. The software is… mostly fine.

I write a lot in Apple Notes. I read in Books. I take photos with Photos. I get my mail in Mail. Most of my day runs through their apps.

But there are a few apps I keep waiting for. And a few I use every day that need work. Today, ideally.

I wrote about hardware I wish Apple made a couple of weeks back. This is the software version.

Software I wish Apple made

#1 A real password manager. Apple has Passwords now. It covers the basics. But it’s still only half a 1Password. It can’t share with anyone outside your family. Software licenses, secure notes, attachments, identity fields — barely there. Anyone who’s tried to migrate off a serious password manager knows.

#2 A serious task app. Reminders has gotten better. It’s still not Things. It’s not Todoist. It’s not OmniFocus. Apple has the entire surface area (Mac, iPhone, iPad, Watch) to ship the best task manager on the planet. They keep building… fine.

#3 A real window manager. Stage Manager is half a window manager. Rectangle, Magnet, Raycast windowing — those exist because Stage Manager isn’t enough. Just buy Rectangle. Or copy it. Ship it built-in.

#4 A serious Calendar. Apple Calendar is the calendar equivalent of TextEdit. It opens, it shows you things, it doesn’t do much else.

Apple has the talent, the platform reach, and the install base to ship the best apps on every single category. They keep settling for “good enough plus a Sherlock.”

Apps that need fixing ASAP

#1 Mail. Apple Mail is minimal and clean. I like it. It’s also got sync issues every other week, search misses obvious emails, there’s no built-in snooze, no schedule send, no quick filters, no templates worth the name. For a free, built-in client it’s fine. For 2026 Apple could be much better.

#2 Notes. Much of my writing starts here. But some things are still missing. The formatting is stuck in 2018. No real markdown. Tables on iPad and Mac don’t behave the same way. The search is ok. The rest needs love.

#3 Music. I don’t really know what’s missing here, but I can’t shake the feeling that Apple Music or the Music app in general could be better.

#4 Files. Files on iOS is supposed to be the Finder bridge. It isn’t really, though. It’s not bad, but it could be enhanced. Also, sometimes the buttons and layouts are a bit weird.

A few that work

Not everything is broken.

Photos has gotten pretty good. Safari is fast and respects privacy (mostly). Books is nice. iCloud Photos sync is solid. iMessage is cool. FaceTime. I even like Pages, Numbers, and Keynotes.

And Mail, when it works, is the cleanest email client. That’s the frustrating part. The bones are good. The maintenance isn’t.

The bottom line

Apple ships the best phones, the best laptops, and some of the best chips on the planet (in my opinion). The software around them should match. It sometimes doesn’t.

Fix Mail first. Build a real password manager. Ship a window manager. Then talk to us about Vision Pro 3 for thousands of dollars. Baby steps.

I wrote about picking a social platform by format back in 2023. The principle held up. The map did not.

Most of the platforms I named are different now. Some are dead. Some swapped their identity. One didn’t exist yet and now eats much of my attention. So this is a redraw.

Platforms

Pick the platform that fits your format. Not the one that’s trending. Not all of them.

If you write, go where writing is the point. If you talk, go where audio lives. If you film, go where video gets watched. Spreading yourself thin across six platforms is the most reliable way to be invisible on all of them.

That hasn’t aged.

If you write

In 2023 I said Twitter.

Twitter is X now, and X is a different product. The text-first culture is still there. The reach is not, for most people I know. The algorithm rewards a narrow set of voices, and small accounts feel smaller than they used to.

Substack Notes is the (new) place text-first writers actually live. It works a lot like old Twitter. It feels light(er). It also pushes you toward your newsletter, which is the point.

Bluesky exists. It’s an option.

Threads is still around, technically. I never warmed to it. But some have success there.

So: Notes first. X if you have history there. Bluesky or Threads as a backup.

If you film

TikTok survived its forced-sale year in the US. Still the format leader. Still the place a short video can go from nothing to a million views overnight. The platform US lawmakers spent three years trying to kill is bigger than ever.

Instagram Reels is a good bet for many creators, still.

YouTube Shorts is even better than Instagram, I think. It pays. It ranks in search. It feeds the long-form channel if you have or want one. If you’re going to do short video anyway, Shorts is probably the highest-ROI version of it.

If you talk

I made fun of Clubhouse in 2023. Remember that one? Audio-only social media. It’s dead.

Audio didn’t die. It went back where it was working: podcasts, audiobooks. The “social audio” wrapper was the part nobody needed or wanted.

If voice is your thing, start a podcast. Don’t wait for the next Clubhouse. It isn’t coming.

If you make images

Instagram, still. Less obvious than it was, though.

Pinterest second. SEO-driven, long-tail, kinder to small accounts. It rewards a body of work over a single viral post. That fits a lot of creators better than they realize.

If you make photographs as art, neither is great anymore. The format that platform was built for got eaten by short video.

Trying all of them

I tried many. It didn’t work.

Six platforms means six content engines, six tones, six audiences who don’t overlap as much as you’d think. The math just doesn’t work for one person.

One platform you like, plus your owned channel (blog or newsletter), beats a thin presence everywhere. By a lot.

Downsides

Picking one platform means missing the next one. Probably.

If TikTok actually gets banned somewhere it matters, that’s a problem for anyone who put everything there. If X keeps drifting, the writers who stayed lose more than they realize.

Owning your audience somewhere a platform can’t take from you is a great idea. Email is still the best version of that.

The Bottom Line

The map in 2026 is quite different from the one in 2023. Clubhouse is gone. Twitter is X. Notes is the writer hangout. Threads are somewhat flatlining. TikTok is strong. Pinterest is underrated.

The principle holds up to this day, I believe: Pick what fits. Make the good stuff there. Send people from there to a list you own.

That’s been the play the whole time.

I publish on Substack. Almost daily. It’s a fantastic platform. It’s got a lot going for it. And mostly for free. But I like Medium more, and the gap is getting bigger.

Substack Notes ate the app

Notes has gotten huge. Good for Substack. Less good for me.

Every update pushes the app a little further toward social media and a little further from writing, I feel like. Many people enjoy this. I don’t really.

New feed tweaks. New video stuff. New ways to “engage.” Like another Twitter, just with longer captions.

I open the app to write. I close it because I scrolled.

That’s not Substack’s fault, really. They’re chasing growth, and Notes is where the growth is. But it changes what the platform feels like. And what it feels like to me right now is a social app with a newsletter and blog bolted on.

The email part isn’t enough for me

Substack’s email tools are fine for a free newsletter. They’re not enough for the way I work.

Most of my email workflows live in Gumroad. Product delivery, follow-ups, segmentation by what someone bought. That’s where list segmentation does the real work. Substack handles the public newsletter. But it’s basically just a sendout of a blog post. Said it before. Substack is more a blog with email attached than a “newsletter” tool.

So Substack’s email side has become the lighter half of my setup. The newsletter still goes out. But it’s not the engine. For anything.

My main driver of Substack growth is SEO. And that’s doing pretty well right now. It’s the reason I still have Substack.

Medium is my main thing

Medium pays me. Quite well most months.

Medium also ranks on Google. Medium has a custom domain that I own at stories.byburk.net. (Substack has that too, though). The reading experience is calm. There is no Notes feed pulling at me. Sure, it’s algorithmic. What isn’t? But it’s not social media, in my eyes. Or less than Substack, at least.

I do most of my writing for Medium first. Substack gets the same article, later. That’s the order, and it tells you what I treat as primary.

Substack has become a second-class citizen for me. And… I treat it like one.

The one thing that keeps me there

I said it. SEO.

The SEO bump on letters.byburk.net over the last months has been really nice. Search traffic is up. Old posts are getting found. The custom domain is doing its job.

I wrote a whole guide about how I do this in Substack SEO. And it’s starting to pay off.

So I’m not pulling the plug. The archive earns its keep, even if I’m not in love with the app (right now).

Will They?

Substack could pull me back in. Probably not by adding more social features. More like the opposite.

Better long-form tools. Better editor. Better post management. Less feed in my face when I open the app to draft.

Could happen. I doubt it.

This is me, not Substack

None of this means Substack isn’t great. It is.

For a beginner writer, Substack is probably the best place to start right now. One platform, free, easy, with a built-in audience graph through Notes and recommendations. For an established writer who wants a newsletter, a blog, and a social feed in one app, it’s basically perfect.

I’d encourage anyone to try Substack. And Medium. Both, ideally.

But if it’s not for you, there’s no shame in not doing it. Platforms are tools. Some fit, some don’t. Mine just fits a little less than it used to.

The Bottom Line

I’m not quitting Substack today. I’m just thinking about where it sits in my stack.

Medium first. Gumroad for email that does work. Substack for the SEO archive and the newsletter that goes out anyway.

Notes is the headline feature for Substack right now. It’s not mine.