RE: Letting an AI break the rules in the middle, as long as the answer is right at the end — Steemit