Obviously this is a complete failure of governance. The very first thing they should have considered was whether or not these features made sense in the ways that they were being added. There should not be any necessary work to "rollback" features that do not make sense, because they should have not built them in the first place.

Even if we accept at face value that AI has made generation of code significantly cheaper, that doesn't justify the existence of worthless code. Taste comes from knowing what not to build.

Right now Windows is an unstable mess, filled with things that shouldn't have been built. The question Microsoft should ask themselves is why they built them in the first place, and how they will prevent this from happening again.

And then every single decision for every single product, at least outside of dev tools, is tainted by this?

Like I understand you always want to have a vision and direction and that needs to be set out by management. But it feels like they just dug down every other vision like "We want users, especially power users, to like using our product"? How does an organization let that happen? Who says "we should make people sign into ms paint" without getting laughed out of the room? Microsoft has great engineers on all levels of the organization. What's going on?

I have since been reminded why this was always misplaced hope. I will never update to Windows 11 or purchase any of their software again.

I'm similarly not updating Mac to their first ai-hype'd OS version. I've only heard poor reviews, zero interest in their glass and hyper-rounded corners

If MS rehires their QA teams and listen to the people on the ground, I'd imagine the very same devs who put AI in Notepad would be very happy to give us features we actually want.

Do you still remember Microsoft Home? My first Windows gaming experience, other than Minesweeper and Solitaire ofc, was the Fury3 demo, which contains only one level from the licensed game, on a Windows 95 Home Entertainment CD. There were also Encarta and other Home products that never took off. The only problem is frequent BSOD, which was solved by the NT kernel, so I guess XP was the pinnacle of personal computing OS, although it does fail some DOS and earlier Windows games.

Do we still have a compatible OS nowadays? Linux is mostly for server and business, and while it has gained some popularity as a desktop OS, it definitely still have a long way to reach the intimacy that early Windows offered, and I doubt it will ever regress back to a "Home computer OS". But maybe we can build on top of that. I mean we can build software on top of Linux that provides the friendly vibe.

Although I have to admit: The combination of AI and required new hardware has been a nice boost for switching to Linux.

Winslop is an open source tool that removes Windows 11 bloatware, disables AI features like Copilot, and restores useful settings: https://github.com/builtbybel/Winslop

What's concerning is that they lack judgment and proper insight into why pushing it in the first place was a bad idea.

If your OS truly is a product, users should not be beta-testers. This isn't an indie kickstarter game.

Forms of it are very powerful and have a lot of uses for sure. But there seems to be an enormous amount of top-down “figure out how to fit AI into our product/processes” for both producers and consumers.

It’s not an AI problem but rather a ram stuff down users throat even when they clearly don’t want it problem.

See broken start menu that does a web search instead showing your apps. See forced online install. See one drive everywhere.

Toning down the AI a bit won’t be enough

What would happen to the morale? What would people proudly say “this is what my company built”?

I’m trying to understand how they screwed up the AI thing so much.

There’s so much innovation, creativity possible using AI, especially as an integration in the OS.

Place your bets.

"And I want people to know that we made them dance, and I think that'll be a great day"

Hmmm, the "moonwalk" is a dance too...

* Kiosk Mode via the shell launcher delays logon times from <5s to 30-180s - just by turning it on, even if it doesn’t actually enable kiosk mode!

* Local changes via registry keys don’t “stick” consistently, even when the machine is entirely offline

* Offline activations using hardware keys fail across vendors without anyone knowing why (other than Microsoft, for the cost of a support call)

* Existing Windows 10 powercfg scripts and config files do not work with Win11. Our workaround was manually calling the exact same command twice, back-to-back, to force-apply a change.

* Installing language packs via the command line by any means available (Powershell’s add-windowscapability, DISM’s package installer, lpksetup, etc) do not actually populate the GUI with those packs as an option until we reinstall them from the GUI again.

* Adverts are everywhere, even on IoT LTSC Enterprise

Honestly, Microsoft completely lost the plot as to what an operating system is supposed to do in favor of turning it into an advertising and user surveillance machine masquerading as a useful OS.

I hate it.

(sarcasm)

Is not only AI bullshit in my notepad.

The excess adds, intrusive online stuff, terriblee performance for basic tasks like the File Explorer or even opening a menu.

Making everything a damn web page...

One clear example is outlook. Talk wathever you want, outlook, is the indistry standart for e-mail. And while not perfect it was very usefull. Then they keep pushing the new interface on everybody throats. The new interface takes like 1 gb of RAM when in use, agaisnt 200mb of the traditional one, while offering less options. Why would anyone who really cares about e-mail use that shit? People who just casually use e-mails dont use Outlook, they use the webmail.

They choose to ignore the users, and push top down changes into them. But the market dont really works this way for most people. Not every tech company needs to be like Apple.

Microsoft have been de-investing in its own companies to put more money into AI. Yes, they have made cuts on highly profitable business to burn money on AI. I hope that they reverse before they fire everyone that was able to build useful software.

Obviously the last part was disappointing. In hindsight, the push was far too aggressive, and too mediocre. We have a lot to learn from this try, about ourselves and about others.

Integration 2.0 will be or needs to be a bit more skewed towards low hanging fruits. I want a copilot for the help menu, and for the 1000 commands not in the toolbar, and a copilot enhancement to the search bar (which came from google!). And a natural language interface to bring up templates, open files, and look for images. Apple could have done this AI enhancement with settings, and added a start button to open apps.

As for "privacy", that is the first scream that comes out from people who have never touched the product. Assurety should have been made that all data would be local. This runs contrary to the "data is the new oil" direction, but it would have been a good way to entice the skeptics and eventually swallow them whole to give up their privacy (insert evil laugh here).

The taskbar, start menu, file explorer, desktop and system settings are areas that should be boring, consistent and just work. That hasn't been close to the case since Windows 7.. Windows 8 was the start of the decline and that was over 13 years ago. While I like some of the things that have come to Windows since then, there's probably twice as much to dislike. Hire/promote/raise someone to act as a UX czar for Windows all around... design a "final" UI/UX target that covers existing features (starting with a checklist of everything that currently can be configured everywhere)... and finish that design.

Stop shifting gears and enshitifying the overall experience to wring every cent of "value" out of every user out there. Get a full release cycle or two with no new features... just solid bug fixes and a more consistent UI/UX. I know there's kernel features for new GPUs, improved storage and CPU interactions and scheduling... but the UI/UX on top of the kernel is a mess and it shows too well.

Get your staff aligned on a common UX theme... I don't care if it's back to Fluent or something closer to where Win11 is, but at least get it consistent above all else. It's a mess. Fire half the managers... stop giving out bonuses based on new features in windows, and give bonuses on sections of config/settings that are marked as complete without falling back to legacy UI from the late 90's. Not that the 90's setup was that bad compared to the hybrid bastard we have today.

They are doing dumb shit for about 5 years now, and killing MS Office, a brand thats market leader for more than 30 years prooves that anybody who had conservative opinion on how software should be built have already abandoned the ship or was kicked out of it.

Now is being run by "visionary" marketing people, and the only way left is down.

I’m not sure why I need to know the history of screenshots that is Recall. Maybe this was simply the best they could do?

That said, Windows 11 is such an AI-fueled privacy dumpster fire that it’s getting replaced by Linux on my gaming PC this month. Then I’m only stuck on Windows for work, and even then I can still write code on Mac or Linux.

Anyway, I re encrypted my Windows install a few days ago. Linux is still an adventure the moment you need to do anything difficult.

The problem is that the engineers and coders that ran microsoft are no longer doing so. Their leadership is too disconnected.

It's the same recipe of enshittification you see all over, marketing, sales take over along with consultants and MBA's. Except, I think in Microsoft's case there are still enough people who know about the technology to be dangerous, they're not entirely clueless, they just know enough (like about AI in this case) to be dangerous.

Regardless of the sentiment on HN and elsewhere though, even excluding enterprise usage, I doubt windows' marketshare will suffer more than 10% YoY.

NOw that the hype is naturally going down, and people are encountering limitations in what modern AI can do, and having more realistic expectations, Microsoft hype train stoped in middle of the hill climb.

They pretty much exchanged gradual increases in revenue for a quick but ultimally short term profit.

Now I'm curious: will the executives, paid millions because they are visionaries well ahead mere mortals like us, be fired for this pathetically stupid strategic push?

Which makes me believe that their "walk back" is just to change the packaging of the same old "slop" being shoved down their customers throats.

2. Realize that Co-pilot is bad and needs to improve up to Microsoft's highest gold standards of trustworthiness.

3. Ditch Co-pilot branding inside the OS.

4. Make AI features private and offline by default unless the local hardware cannot run the specialized tiny model for that task, at which point it goes online for it. It might be slower, but if it does the thing, it's ok.

5. Allow companies and power users to provide their own local models that hook into these tasks, so they can host AI servers within the company and these AI tasks never reach outside of the company.

6. Make AI features more specific, targeted and useful instead of simply integrating it into the various functions and throwing it at users like "here, you figure out what to do with this thing, we don't know."

7. Don't expect people to want to chat with it in every app, just find a task that you know it succeeds at and expose that task rather than letting users figure out what it sucks at.

8. Don't make the AI integration APIs a case of increased surface area privacy and security risk that 3rd party system apps can hook into, to mass extract information out of every app on your system easily. Put limitations on it.

9. Add features to specify where AI can go and cannot go, just like the microphone. Folders, apps, online services. Even if it does use Co-pilot online, let users sculpt it.

10. Make it explicit and obvious when AI features are operating offline or online. If users have decades of understanding that Notepad is a private offline app, preserve that expectation as much as possible. Just because Outlook and OneNote are very online-oriented apps, it doesn't mean they want their local experience to be online in every way. If you force AI to go over all my cloud files, notes and e-mail without my permission, that is sociopathic behavior and I will ditch you, Microsoft.

Some day Co-pilot will probably be good. That isn't today. It's probably not this year or next year, but eventually. Until then, it needs to stay in a lane with freshly painted lines surrounded with sand barrels in case it wrecks.

It's not that I'm entirely opposed to some Microsoft AI feature existing in Windows, but manufacturing a user assumption that it is everywhere all the time is bad not just for Windows, but for society as a whole.

We've already seen how political and activist the public sphere became over the last decade, which reduces trust in the people who make software and services too. What do we do when Microsoft gets ideologically taken over and abuses its information access to people for political ends?

Show you can be trusted. When I put a little food bowl down for you, don't scratch me and we'll go from there.