You're not wrong in saying AI is not helping matters, but that's not the root cause; merely an amplifier.

What pisses me off about this new round of doom preaching is the assumption that our extinction will be mandated by a hypothetical emergence of a Skynet level AGI, with the implication that just regulating AI (what does that even mean, BTW?) will solve all our problems. Firstly, we were well on our way to an extinction-level event before AI even showed up. Secondly, AGI is a made-up problem, akin to worrying about a meteor strike while our house is on fire and the tenants are gouging each other's eyes out on the rooftop. More and more I hear discussions around me about this bullshit non-issue, while we (as a society) continue to ignore the issues that will actually pull us past the point of no return.

And the kicker? I have this nagging suspicion that the whole "slow us down before we kill us all" clown show is triggered not by a sudden onset of morals in those involved, but (and I speculate here) by the inevitable collapse of their new frontier models in the lab. There's a limit to how large an LLM can grow, and I think they reached it. Saying it out loud though would collapse the whole Ponzi scheme, so they're begging for a regulated slowdown. As if the law would suddenly mean something for the people that pirated a metric fuckload of copyrighted works to train their stochastic parrots.