David_Gerard comments on Some alternatives to “Friendly AI”