Yeah, for me EY's stories about an airgapped AI convincing their keepers to release it (or that other one about an AI destroying all humans because it's the mos

Yeah, for me EY's stories about an airgapped AI convincing their keepers to release it (or that other one about an AI destroying all humans because it's the most efficient way to make a gazillion paperclips) also miss the point. It's mostly the just the realization that superhuman AI might be dangerous through some kind of runaway feedback, as a very general idea, that I respect about EY's work.