John_Maxwell comments on Some alternatives to “Friendly AI”