I agree with you, as “smart” as these people at AI companies are, somehow they keep missing the fact that it’s impossible (unless you kill the majority of people) to align on this kind of thing