From: Tomaz Kristan (me2icq@icqmail.com)
Date: Mon Jul 08 2002 - 10:44:24 MDT
On Mon, 08 July 2002, "Mike & Donna Deering" wrote:
> Would it be possible to change
> the design slightly to
> avoid volition, ego, self consciousness,
> while still maintaining the
> capabilities of complex problem solving,
> self improvement, super intelligence?
But of course - if you ask me. Just as I've said:
AI (=full context simulator) could make a sorted list of all possible actions for a situation, and it's up to you, which action you shall approve or take. The context may be any world description. Physics above all. This way, AI is no smarter than a calculator, and you are responsible for your decisions.
Consider the Heimlich Maneuver. It could came out from a (rough) human body simulator. It has, in a sense. From Dr. Heimlich's head, as far as I know. The (enormous) space of all situations + actions should be examined with a good software simulator and a list of top N or so - printed out.
Top N by the criteria of "still alive after the action, but not without it".
That (stupid) AI could do miracles for the First Aid, although it wouldn't be trivial to make such a simulator. But far from impossible.
My point is, that it is doable for virtually every field.
Any idea why not?
- Thomas
- Thomas
-------------------------------------------------------------
Sign up for ICQmail at http://www.icq.com/icqmail/signup.html
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:40 MDT