Scott Alexander comments on Little Johny Bayesian