"These AIs are usually right about things I don't know anything about" sounds like the textbook example of risky thinking though.