Jimbabwe@lemmy.worldtoTechnology@lemmy.world•Child safety org launches AI model trained on real child sex abuse imagesEnglish
359·
2 days agoIf AI was reliable, maybe. MAYBE. But guess what? It turns out that “advanced autocomplete” does a shitty job of most things, and I bet false positives will be numerous.
Somebody was on reddit today