Humans hate to think and make decisions, they like being told what to do. So having an AI doing the dangerous part of thinking, leaves humans to do what they d

Humans hate to think and make decisions, they like being told what to do. So having an AI doing the dangerous part of thinking, leaves humans to do what they do best: follow orders. Even better AI will take on the responsibility when anything fails: just get the AI to fix it, after all AI coded the mistake.