Comment by jordanb

The problem is the "AI Safety" people seem entirely focused on a sci-fi "the computer is a vengeful god" plot and not at all on the AI talking people into suicide or ruining children's educations. This makes them seem unserious and out of touch.