RussellThor comments on The case for stopping AI safety research