lukeprog comments on So You Want to Save the World

[lukeprog]

Updated: Added a new problem category: “How can we be sure a Friendly AI development team will be altruistic?”