I fully agree about provocative content. Well except that some people enjoy feeling angry by believing something that isn't true. That's a kind of pleasure that's worth something, but it seems emotionally harmful in the long run. God in these olden-days moral judgements usually means emotional health of the individual or vaguely defined long-term success of their society.

My point is that it's not AI-vs-human that matters. AI makes provocative content easier to write but it also makes helpful technical documentation easier to write.