Re: P-zombies in the NY Times

From: Dan Guernsey (dan-guernsey@utulsa.edu)
Date: Wed Aug 22 2007 - 19:01:37 MDT

Robin Lee Powell wrote: 
On Tue, Aug 21, 2007 at 03:27:00PM -0400, Eric Rauch wrote:
  
Also if this is a simulation doesn't that make us a boxed AI?
Didn't you guys decide that it would be tough to keep an AI in a
box?
    

The point was that a super-smart AI in a box that we don't talk to
is useless; we can't ask it to cure cancer if we don't talk to it.

A super-smart AI we talk to can get out.

If we're in a simulation, the guys outside the box aren't exactly
advertising the fact.

-Robin

We don't necessarily have to "talk" to it in order to get it to solve a problem. For instance, what if we are a simulation and the world around us is constructed to coerce us into solving a certain problem. You use a cure for cancer as an example: perhaps cancer introduced into this simulation because the guys outside the box face this same problem and expect us to find a solution.

On another note, it's unsettling to ponder what might happen to us once we solve the problem we were designed to solve (if we were a simulation, that is).

Dan

This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:58 MDT