Matthew_Barnett comments on Analyzing the moral value of unaligned AIs