DragonGod comments on Is AI risk assessment too anthropocentric?