Major AI companies form group to research, keep control of AI

I read this as more of a "both, and" situation. Lots of other industries have ISACs (Information Sharing and Analysis Centers), mostly focused on information security, and they can work quite well. The FS-ISAC, for example, is quite active in the financial services sector and (IMHO) reasonably effective.

It's always been a little baffling not to see FAANGs on this list (the IT ISAC is broader than just the tech industry): https://www.nationalisacs.org/member-isacs-3

So, I think this is a good step for the AI industry. Of course regulation is still a necessary component, but an industry consortium focused on sharing lessons learned and good practices is a good step, in my view.

I don't trust these groups to act in my best interest. If AI is as powerful as they say it is, it needs to be controlled by groups whose stakeholders are not capitalist corporations.

This is a terrible idea for the future of AI.

tbh I struggle to see these incredible promises of the future for AI to evolve the way everyone seems to think. I do see the tool value and the revolution that can have but the feverish talks about AGI and the coming AI apocalypse sound very self-aggrandizing by the people with their careers staked on it to me. Even if there is widespread and horrible societal impact it sounds to me like they are not putting the right people at the helm to work on that.