I personally see two major factors :

• it's yet another way to collect data for monetization (selling to governments and advertisers).
• it's a symptom of market culture : it exists not to serve the user, but to be a product to sell to the user. In that light, it behaves and evolves exactly as any other thingy which tries to reinvent itself year after year, to keep customers interested, regardless of its intrinsic value as an operating system.

but that's just me, I can't claim to understand the inner workings of Microsoft

AI is obviously the next big thing. If your product isn't full AI right now, you'll miss out.

So basically I'm convinced the reason everyone's chasing AI is all just greed and fomo. People at the top of the ladder seem generally more detached from reality.

Three big things, from most to least obvious:

1. They own openai, which is a ponzi scheme.
2. They are spying on you. Obviously. This is a better excuse to do more of that more openly.
3. This creates a corporate controlled abstraction layer between you and everything you do with any networked device, puts more decision control and discernment in the corporation's hands. When Amazon tried to move to voice based ordering, it was on this same logic; that you would say 'alexa, order toilet paper' and they would choose the toilet paper to order based on what's best for them-sponsored, most profitable, etc, like you can only buy things from a selfish hyper sociopathic genie. LLM based ux doesn't just put the corporation between you and products, but between you and all information, between you and truth.

If your program/OS has to constantly phone home to operate, you can't easily distinguish the traffic related to it's operation from the traffic snooping all your data. Extra points if you willingly give the AI access or information.

Asskissing.

Decisionmakers at MS are divorced from reality. The cognitive dissonance is incredible. They make shit up as they go along and the rest eat it all up and re-spew some new godawful incarnation of the terrible ideas.

None of the decisions are based in reality or quality market research (what users want). Their market research boils down to AI is god and we need to make it even more ungodly.

Which isn't that surprising. When you have a single company dominating a market niche (say, desktop OS-es on PCs), the focus isn't increasing, retaining or in any way satisfying customers. It's making shit up as you go along, trying to get your department's lines a bit higher than the other one so you as a high-to-mid level manager get that sweet, sweet bonus.

It's not limited to MS and OS-es. The same thing applies to Google and the search and browser markets, as well as Nintendo and their ecosystem, or HP and printers.

You don't even need a monopoly for this shit to happen. You just need to be "too big to fail". Luckily, such unsustainable behavior won't be "too big to fail" forever, but it's impressive just for how long a company with such rotten echochamber decisionmaking behaviour can keep chugging along just fine, all the while hurting the customers and the economy in general through knock-on - or shall I say trickle down - effects.