x
Bayesian Phronesis: A Regularization Strategy for Agentic Virtue — LessWrong