x
Debiasing as Non-Self-Destruction — AI Alignment Forum