Radford Neal comments on AI #132 Part 2: Actively Making It Worse