Scott Alexander comments on The Good Bayesian