Gavin comments on The academic contribution to AI safety seems large