Daniel Paleka comments on The shallow reality of ‘deep learning theory’