Loading…
27
An old-school web dev made me rethink AI content moderation
Met a guy at a local meetup in Pittsburgh who runs a tiny forum moderation team, and he was explaining how they use AI to flag posts but then manually review everything before banning. He said 1 in 5 automated flags are false positives, but they still keep it because it catches stuff they'd miss. It made me wonder if we're too harsh on AI moderation, or if we just need better human checks in the loop. Has anyone else seen a mixed system like that actually work well, or does it just add more work for people?
1 comments
Log in to join the discussion
Log In1 Comment
troyreed27d ago
That 1 in 5 miss rate is like spam filters, you still check your junk folder.
1