I agree with most of what you’ve said, except the central tenet that Ai is a threat, unconditionally.
Using it as a means of harvesting wealth from the majority, by a minority, in the “Zero-sum” game is obviously harmful, as we can see from the effects of profit driven social media, it amplifies human misery, since that is what makes most profit.
The Romans knew this also, regalling us with gladiator games, using the enslaved.
Isn’t it ironic that the original meaning of Libra, the name of the new Crypto token intended by, and for Facebook, is an ancient Roman measure of weight?
It would often have been used to measure the weight of a slave.
If one or another corporate entity does attain overall world government, then maybe we are screwed.
However, the antidote could be non-profit open source Ai, programmed to maximise the value of, and reward humans, at no cost to any others. This could turn out to be an extremely powerful force for good.
Each human could effectively have a superhuman protector, for free, with the net effect that in addition to making all infinitely financially wealthy, all humans, all animals, and our environment are automatically also protected, as this is what adds highest value to humans.
The point should be made that with Ai, we have it in our power to create our own heaven or hell.