From: John K Clark (email@example.com)
Date: Wed Jun 11 2008 - 11:28:07 MDT
>> It can be proved that a finite set of axioms
>> cannot derive all that is true, so I see no
>> reasons why a finite set of goals can derive
>> all actions that can be performed.
"Stathis Papaioannou" <firstname.lastname@example.org>
> If that's true then a machine with a finite set of
> goals won't be able to perform all actions that
> can be performed.
Not only that but if we move from GŲdel to Turing then there are some
actions a goal oriented mind can never perform but we or it can never
know it canít be performed. The AI can never solve the problem nor can
it ever prove it is imposable to solve, it will just go on and on
getting nowhere. A mind that operated with a ridged goal structure would
soon be caught up in one of those infinite loops. A flexible mind needs
to say ďIím not getting anywhere, to hell with that super-goal; itís
more productive if I think about something elseĒ.
> Variability of goals has nothing to do
> with level of intelligence.
It has everything to do with the level of intelligence. As it learns
more it will discover some of its original goals were contradictory,
redundant, or just plain stupid.
400 years ago Johannes Kepler thought the top goal of astronomy was to
discover why there were 5 and only 5 planets, Mercury Venus Mars Jupiter
and Saturn. He thought there must be something special about the number
5 that caused this and it was the top goal to spell out exactly what
that was. Kepler was a very smart man and at the time that seemed like a
very reasonable goal, but as it turned out nobody ever achieved it and
today nobody is even trying to because we now know that goal was stupid.
>> What are the axioms of human behavior?
>> What is the top super-goal?
> If there is one, it could easily change, but it
> is not necessary to program a computer this way.
It is if you what the computer to be intelligent. At least thatís what I
think and evolution agrees with me.
John K Clark
-- John K Clark email@example.com -- http://www.fastmail.fm - Email service worth paying for. Try it for free
This archive was generated by hypermail 2.1.5 : Mon May 20 2013 - 04:01:21 MDT