Shmi comments on Understanding differences between humans and intelligence-in-general to build safe AGI