The thing I find off in his analysis is he just assumes AI is rubbish and won't get better, hence all the investment in it is wasted. I don't think that's realistic. Even if you are skeptical there's some chance it'll work out.