Steven Byrnes comments on Can we simulate human evolution to create a somewhat aligned AGI?