..............
i'm gonna assume you have already studied Troman's source for Aivolution.
next i would recommend a review of games that have made claim to unusual
a.i. techniques HERE:
http://www.gameai.com/games.html
there are a couple premier middleware a.i. packages that make some use
of "biologic" technique and have been around for like 8 years and therefore
quite mature. i'm thinking for example of:
AIimplant: http://www.presagis.com/products/simula ... aiimplant/
to date the practical consensus in the game world over the last 10 years is that
the application of neural nets to game a.i. is over-rated and over-hyped.
(i mentioned utilizing NN to Alex Lee who wrote the original WZ a.i. back in '99 and
his evaluation was that it was much more trouble than it was worth and that he could
achieve the same goals with WZ script. which is in essence what Troman has
achieved over the years... and also by upgrading the WZ scripting lang itself..)
i would strongly recommend you get your hands on a copy of:
"Artificial Intelligence for Games" (w.CD) by Ian Millington &/or the
series of 4 books called:
"AI Game Programming Wisdom" (w/CDs)
on a final note...
have you considered WZs pathfinding scheme ? if you haven't you need to
because you can NOT isolate your proposition from the A* PF. i'm no expert
but from what i know WZ's PF is far, far, from ideal when your talking a NN ai
implementation. WZ PF is grid-based and NN is based on ? i'm thinking you
can see where i'm getting at...
...
