Vladimir_Nesov comments on London meetup: “The Friendly AI Problem”