Re: [FT]SML question

On Tue, 19 Jun 2001 14:04:21 -0700 (PDT), David Griffin
<carbon_dragon@yahoo.com> wrote:

>I'm certainly not a fighter pilot, but in the games
>I've played where you get the chance to choose what
>you do next -- dive, climb, turn sharp, fire a
>snap shot, etc. There seem to be a LOT of different
>choices. I don't think it would be quite so easy
>to distill all that in a program. 

Oh, it won't be easy! That's for sure. But it's a matter of choosing the
right
thing to do, or a reasonable thing to do, quickly. I understand that the
next
generation of avionics systems (maybe even the state-of-the-art today)
will
be/are capable of not just recovering the plane in the event the pilot
blacks
out, but attempt to get the plane into a superior combat position. 

>In Chess at first, we saw computers play badly 
>because they weren't capable of the sort of neural
>short circuiting we do in jumping to a conclusion
>without working through all the possibilities. But
>when they got faster and were able to work through
>all the possibilities, they didn't need that
>ability.

Well, that's not actually how they work. They have a database of moves
that
they search through. They don't do every possible permutation. They try
to
achieve a position stored in the database that can't be assailed. In
other
words, they have completed games stored in the database, and follow the
moves
in those games. If this sounds like "cheating", chess masters for
decades have
studied games in order to memorize winning moves.

It's also why no computer opponent has been able to beat a grand master
at Go.
Go is simpler, in moves, to chess, but the permutations are far more
complex.
It devolves into many games at once, each part of the board being a
different
battle. 

>In regards to the inertial dampers, I can see what
>you're saying. I guess what would matter is whether
>the fighter pilot could bring something to the
>equation that would be worth the loss in performance.

That's very true.

Allan Goodall		       awg@sympatico.ca
Goodall's Grotto:  http://www.vex.net/~agoodall

"Now, see, if you combine different colours of light,
 you get white! Try that with Play-Doh and you get
 brown! How come?" - Alan Moore & Kevin Nolan,