The Hacker Learns to Trust — AI Alignment Forum