Path: Multiple angles on Bayes's Rule — LessWrong