Vladimir_Nesov comments on An illustrative model of backfire risks from pausing AI research