

I suspect it’s more likely that Meta won’t exist in any recognizable way in 2035 than anyone makes a huge profit on A.I. in the next decade. There will be advancements, to be sure, and compute will (hopefully) get cheaper and more efficient but 2035 seems like an Elon-Musk-level optimistic timeline.
I’m sort of agnostic on A.I. I don’t like it for much now but certainly see its potential. But look how long it took for The Internet to be universally adopted. And that’s assuming researchers even can solve the really hard last 20% of problems.
But someone smarter than me once put it better:
I would recommend installing a fairly vanilla Gnome distro (like Fedora or something) and then a KDE version (most major distros have a KDE spin) in a virtual machine. Gnome Boxes is a really easy way to do that. And then just customize the shit out of both of them and see what you like best.
Gnome is more of a macOS-like experience so to me, it feels more trackpad driven (though keyboard shortcuts are plentiful). Install some extensions if you don’t like something. Someone else probably also didn’t like something.
KDE is more like Windows. I’m less familiar with it but it’s on my Steam Deck so I use it a decent amount. It’s more mouse and keyboard driven, as far as I can tell. So, that’s why I think it would be fine to evaluate in a VM.
They’re both high quality, though, so it’s really about what you prefer. I like Gnome, obviously, but I prefer to code on a smallish laptop (for portability/travel reasons) and a dock whereas a lot of people want an elaborate multi-monitor situation and a different interface. Everyone has their own workflow. Both work equally well so it’s just a matter of taste and preference. (Most Linux decisions are like that and people get weirdly angry about it but that’s part of the fun. Choose your own adventure.)