The first is that these people have often been indistinguishable from the ambitious and power hungry people I was decrying. Sam Altman was able to blend in for a long time by copying the rhetoric of AI safety types. When there is this much money to be made and power to be amassed, it's not hard to pretend to care.
The second is that I have often been disappointed by what the AI safety folks are concerned about. There has been a huge amount of talk about existential risk and not nearly as much about, say, the impact on children if education is outsourced to AI. The obsession with science fiction led to some very out there scenarios that may or may not still happen, but have nothing to do with the very real impacts of AI on people's lives right now. I believe that even the well intentioned have been too detached from humanity as a whole.
If you don't live in the US and you are taking advantage of the US security umbrella, sure, you can deny AI and enjoy a curated lifestyle.
But living in the US means we must deal with this.
I also think your whole unipolar vs. multipolar framework is ahistorical. It's always been more secure living within the cosmopolitan center of an empire, but there's never just been one empire.
I just think your view of history is too simplistic to be accurate or interesting. If/when the US declined as a global power, the results would be entirely unpredictable. They wouldn't adhere to the kind of formula you're describing.
The Social Network is looking more and more prophetic. When your only ambition is success for its own sake, driven by insecurities you can never name, you're going to make things that don't actually serve other people, or only do as an unfortunate side effect of you making money and gaining power.