Gordon Seidoh Worley comments on Can few-shot learning teach AI right from wrong?