Experiments with AI based moderation on Discourse Meta

How often you get false positives or otherwise misses? This is relatively peaceful environment, though.

1 Like