Esben Kran comments on The limited upside of interpretability