" He acknowledged that "there is always some threat that in actually wanting to advance (friendly) AI we may possibly build the thing we've been worried about"; but Even so, that the top defense was "to empower as Lots of individuals as is possible to possess AI. If Everybody