You are complaining about counterfactuals. Nobody is bullying anyone; the people active on meta aren't "chasing away" users (almost no users actually come to meta in the first place, the big names don't do a lot of personal interaction and when they do they don't leave comments unless they can be exceedingly polite about it); all that's happening is that with LLMs people finally have the thing they wanted all along[0] so they no longer have to keep coming to SO demanding it to be that thing when it was very explicitly designed the entire time
not to be that thing,
specifically so that experts wouldn't have to
waste their time when they try to make the world better on a volunteer basis[1].
But most importantly: the site does not even remotely in any imaginable way empower those users. It actively hinders them, the staff have berated them over the years over alleged "unfriendliness"[2] and they constantly dump their ideas on the meta site, then ignore all feedback and push their ideas through anyway.
The main site is noticeably slower now. As far as I can tell, this is because they've taken code paths that the beta uses (especially whatever it is they use to load the code boxes and put a little JS widget at the top of them) and applied them back to the main site that worked perfectly well before.
[0] Basically: a magic robot that can listen to them and try to suggest an answer one on one, without caring about whether literally anyone else on the planet could benefit from that exchange, because the robot can tirelessly do that for the next person.
[1] I.e., so that humans could say something once and actually have it be relevant to many people.
[2] I.e.: the company makes money from views and a lot of people don't want to view a site where they have to actually meet any kind of standard whatsoever to participate; so it's the fault of people who have an actual vision for the site being useful.