" He acknowledged that "there is often some threat that in essentially endeavoring to progress (helpful) AI we may create the detail we have been concerned about"; but nonetheless, that the most effective defense was "to empower as Lots of individuals as is possible to acquire AI