Marius Hobbhahn comments on Deep Learning Systems Are Not Less Interpretable Than Logic/​Probability/​Etc