As a shareholder (which I'm not), it's absolutely amazing.
As a human being though... it's simple to look at the history of the company, from its inception based on nepotism and locking-down was hitherto the common good, to going from one place of monopoly (OS, app, cloud) to another (extending to whatever is trendy at the moment e.g XR with HoloLens, AI with OpenAI, etc).
It's IMHO one of the very worst thing that could have happened to humanity in terms of cognitive empowerment. Apple is not far behind but in terms of locking up an entire ecosystem but Microsoft, sadly, is doing it better.
To clarify what I mean is that Microsoft is the business embodiment of learned helplessness. Most people would shrug at the quality of software they provide, the price, etc ONLY because they are convinced, wrongfully so, that they are is no legitimate alternative. If users were actually able to chose, not being coerced into but properly chose, by experiencing alternatives, the World would be totally different. Instead of having computer users who feel an adversarial relationship to their devices, we would have a much stronger relation of "this is MY device" the same way a lot (not all) of people have a repair toolbox at home. They know they can try to fix something in THEIR home, even improve it. Most people understand it won't be easy, they might mess it up, but it's possible to try. Not in software, and that's entirely Microsoft "success". Maybe in an alternative reality others, like Apple, would have made that happen to, but in our reality I blame Microsoft, Bill Gates upbringing from his legal mindset father and well connected mother.
We could have a world were users own their devices, have a challenging yet empowering relationship to technology, starting with software, and instead we have exploitative learning helplnessness. So yes, Microsoft is that bad.
My documented process https://fabien.benetou.fr/Content/SelfHostingArtificialIntelligence but honestly I just tinker with this. Most of that isn't useful IMHO except some pieces, e.g STT/TTS, from time to time. The LLM aspect itself is too unreliable, and I do like 2 relatively recent papers on the topic, namely :
which are respectively saying that the long-tail makes it practically impossible to train AI to be correct in rare cases and that "hallucinations" are a misnomer for marketing purposes to be replaced instead by "bullshit" used to convinced people without caring for veracity.
Still, despite all this criticism it is a very popular topic, hyped up to be the "future" of computing. Consequently I did want to both try and help others to do so rather than imagine that it was restricted to a kind of "elite". I try to keep the page up to date but so far, to be honest, I do it mostly defensively, to be able to genuinely criticize because I did take the time to try, not reject in block.
PS: I do try also state of the art, both close and open-source, via APIs e.g OpenAI or Mistral but only for evaluation purposes, not as tools part of my daily usage.