“Peter Steinberger has observed a phenomenon he calls "AI psychosis," where the public is overly gullible or trusting of claims made by AI agents.”