William_Saunders comments on Why I prioritize moral circle expansion over artificial intelligence alignment