Ilya Shpitser comments on Deep learning—deeper flaws?