You seem to be banging on a stereotype. Yudkowsky doesn't speak for everyone (not even everyone on lesswrong) and AI safety/paranoia is an incredibly polarizing issue.

There's a loosely structured community of people who self-identify as "rationalist". There are some general themes of belief but no coherent platform, per-se. Some of them are nutjobs. Maybe at an elevated rate from general society, but not dramatically so.