Peter S. Park comments on The limited upside of interpretability