James_Miller comments on Intelligence Amplification and Friendly AI