NickLaing comments on Bentham’s Bulldog is wrong about AI risk