David T comments on Sensitive assumptions in longtermist modeling