John_Maxwell comments on A dilemma for prosaic AI alignment