Sebastian Schmidt comments on Towards more cooperative AI safety strategies