" He acknowledged that "there is always some threat that in essentially wanting to progress (pleasant) AI we might build the matter we've been worried about"; but Even so, that the most beneficial protection was "to empower as Lots of people as you can to have AI. If All peop