x
Understandability principle — AI Alignment Forum