"J. R. Molloy" wrote:
>
> Eliezer's definition is when AI exceeds intelligence of smartest humans.
No, hardware-improved humans also count, as long as it's genuine hardware
improvement and not just using Google or forming corporations or whatever
the latest Fad Pseudo-Singularity of the Month.
> I ask because a list member has expressed fear that a system
> which identifies incorrect thinking might do so with extropians. Wouldn't that
> actually be a friendly thing to do? I mean, if extropians think incorrectly, a
> friendly AI would be doing all sentient beings a big favor by removing that
> incorrect thinking, right?
By *identifying* the incorrect thinking, to those people who would want an
AI to help them identify incorrect thinking.
-- -- -- -- --
Eliezer S. Yudkowsky http://singinst.org/
Research Fellow, Singularity Institute for Artificial Intelligence
This archive was generated by hypermail 2b30 : Fri Oct 12 2001 - 14:39:59 MDT