Scott Alexander comments on Techniques for probability estimates