North And comments on Analyzing the moral value of unaligned AIs