Prev: Re: AI's in full thrust Next: Re: I lost a...

Re: AI in FT ... again,...

From: Darren Douglas <ddouglas@v...>
Date: Tue, 15 Jul 1997 18:49:02 -0400
Subject: Re: AI in FT ... again,...

Chris McCurry wrote:
> 
> >One great reason, is the training time of pilots. A fighter can be
> >manufactured in
> >a lot less time than it takes to train a pilot to fly it.
> 
> >This allows for high use/loses. There are a pile of other good
reasons,
> >human
> >pilots for example have problems like ill health, fatigue, they need
> >sleep etc.
> 
> >Darren
> 
> Thought i still believe that from an over all stand point that the
living
> pilots do has a sort of advantage..
> the irrational thought..
> 
> but with AI fighters there could resonaly be a 10 AI fighter to 1
Human
> fighter..  due to training and so on..
> 
> unless...:  you have a AI that is smart enough to learn..  there is
sucj
> thing as self modifing code..
> and you have to admit that at a very small and uncomplex look that is
a
> limited AI.  (first steps at least).
> 
> CMC

I have just remembered, in the book SkunkWorks, about Lockheeds special
development
unit (blackbird, u2 spy plane, Stealth Fighter). They mention the
compterised
flight combat system on the stealth.

Apprantly, it could be programed with the mission profile, flight to the
target,
(using the best route for stealth capabilites and keep the aircraft at
the
best flight angle to avoid detection) drop the muntions if the pilot did
not
want to, or had not overidden due to avoiding defences etc. And then fly
home.

True it could not handle a dog flight for the take off and landing and
had
not dicussion making capabilities, (we still need human pilots) but add
another twentry or so years and see what we get. This system was
designed and
developed in the late seventies early eighties.

Darren

PS. Do read this book "SkunkWorks" its a real eye opener.

Prev: Re: AI's in full thrust Next: Re: I lost a...