Molnar has written the book "Interpretable Machine Learning: A Guide for Making Black Box Models Explainable", in which he elaborates on the issue and examines methods for achieving explainability.