3 ms·
A very good outcome for AI safety would be if when improved models get released, malicious actors use them to break society in very visible ways. Looks like we'
by AntiDyatlov 6mo ago
A very good outcome for AI safety would be if when improved models get released, malicious actors use them to break society in very visible ways. Looks like we're getting close to that world.
- sourcecodeplz 6mo agoGives me Fight Club vibes.
- pants2 6mo agoIt would certainly be good news for cybersecurity employment!