x
“Endgame safety” for AGI — AI Alignment Forum