Vaniver comments on The Value Learning Problem