DanielFilan comments on Deep learning—deeper flaws?