Very cool, I would definitely be interested in such a contribution. I'm very interested in AI in general.

At the moment I'm sort of trying to figure out how to make an AI player play my new "some sort of chess game" with more strategy than just near random moves. Turns out to be quite a challenge, but very cool stuff still,

Cheers

@ Phemox:
A-Star would work... he checks all possiblities and give each turn points so he can decide what is the best turn

I'm really looking forward to this paper. Larry told me it used the clever FEAR system of A* decision making, and while the papers he pointed me to helped me understand the concept I still need to know the best, most efficient way to code it. The GDC paper was very conceptual, rather than looking at how to programming issues, and while informative just wasn't enough by itself. I'm thinking this paper will help clear the gaps.

@Phemox:
Have you heard of deep blue? It works out almost every possibility that will happen in the game and takes the move that gives the best chances. By almost I mean the next few million moves that it and you can possibly make. However, that would be kind of hard, especially since that was a dedicated team and they were using a supercomputer.

Well, I fear that I would be in way over my head, but I will read it from top to bottom. Probably several times. I respect your work and would be grateful for the opportunity to share your knowledge, Larry. (Too bad there will probably be math involved... )

Mark

Actually, that's the idea.. The paper is intended to someone without any computer knowledge at all.(apart from the very basics of course...) but anytime i mention a finite state machine, or A*, or even the term 'game engine', I'll then have to devote at least 1 paragraph explaining it

So no math involved or anything... On the other hand, you'll need to make further research in order to find the actual algorithms that will help you implement what you'll read. I keep saying that, so please don't get let down... On the other other hand, after you're done with it, you'll be able to say something like: "oh ok, so that's how larry did his Ai...". If anyone decides to implement it, i'll be here to help.. also, if someone learns better by looking at code, FEAR's Ai is openly available to the public


Phemox, after you're done reading it, i'd be more than happy to talk with you about how you can make this work for your problem. The action planning system is actually ideal for AIs like what you need..

ok, so a couple more hours and i'm done!! 67 pages in total!

With or without neuronal nets in c-script?! ( )

lol..

well, here's what you could do. try to get the minesweeper example work in Lite-C.. that seems do-able.. and if you do, make sure to post! i'm sure a lot of people, including me would love to see that! that demo was veeery impressive, i used to keep it going for days, just to see how good those minesweepers are gonna get

Where is this minesweeper example??

Damn i know what you mean now ^^
i'll try my best ^^