> "A.I. Safety" research is the worst of it, as it suggests this technology is so powerful that it's actually dangerous

Well that's an interesting way to misrepresent an entire important field of research based on what a few idiots said. There are serious people in that field who aren't addicted to posting on LessWrong.