ESRogs comments on Call for cognitive science in AI safety