x
Reconciling Shannon and Bayes. — LessWrong