Shmi comments on Is friendly AI “trivial” if the AI cannot rewire human values?