I thought some of you may be interested in the 20-line &#106avascript programming contest over at Ozone Asylum this month, as the topic is A.I.

I'll be entering, so if anyone else fancies it, jump on in. Read more in the DHTML/JS forum here:

http://www.ozoneasylum.com/7808

It's all fun thing - the winner gets to choose the topic for the next month's competition. Nothing too serious ;o)

Hope to see some of you there!

Dan

hey!

seeing as this has turned into a competition promotin' free-for-all... why don't you hop on over to ai-junkie and help us settle on a theme for the next AI programming competition. ;)

see this thread:


There'll even be prizes! Yee haw!

Why don't you simply set up some webserver and declare the post/get to some document on it as the interface for all participants. The environment state is feed into AI via the returning page (XML). So everyone can make use the language what the whish to use. I wouldn't take part if I had to choose java, no matter if IBM, SUN or what else ... but you did not tell it's a contest for java but a contest for AI, didn't you?

Quote:

Original post by Sebastian Gruber Why don't you simply set up some webserver and declare the post/get to some document on it as the interface for all participants. The environment state is feed into AI via the returning page (XML). So everyone can make use the language what the whish to use. I wouldn't take part if I had to choose java, no matter if IBM, SUN or what else ... but you did not tell it's a contest for java but a contest for AI, didn't you?

This would seem fair until I create an AI agent which forwards all of the information to my distributed network with 3,000 hosts in order to have more processing power to determine the best choice to make. Tying this kind of contest to a specific language is the best way to do it.

Besides; to a computer scientist the limitation of language is not a real limitation; so long as it is a reasonably robust language (and no one can argue that Java is not), then you should be able to accomplish everything you need with only a trivial investment of time. This is especially the case if you know C++ or C#, both of which have similar syntax. The differences exist, but they are nothing that can't be overcome in a weekend of studying.

i think Java is OK, because with java you can protect it against hacking without having to setup the server, etc (and server idea is bad because not everyone have equal processing power).
Also, really many AI contest arenas is avaliable. I even have one installed on my PC.

[google]

[google]

Why don't we have a GUN-TACTYX competition? It's an AI programming game where you write small programs for your bot and pit him against other programmed bots in a 3D arena. The scripts use the SMALL scripting language.

also, with client-server model, one can connect really HUGE neural network [grin] . _{i mean brain, dudes.}

Quote:

Original post by ChaoticCanuck Quote: Original post by Sebastian Gruber Why don't you simply set up some webserver and declare the post/get to some document on it as the interface for all participants. The environment state is feed into AI via the returning page (XML). So everyone can make use the language what the whish to use. I wouldn't take part if I had to choose java, no matter if IBM, SUN or what else ... but you did not tell it's a contest for java but a contest for AI, didn't you? This would seem fair until I create an AI agent which forwards all of the information to my distributed network with 3,000 hosts in order to have more processing power to determine the best choice to make. Tying this kind of contest to a specific language is the best way to do it. Besides; to a computer scientist the limitation of language is not a real limitation; so long as it is a reasonably robust language (and no one can argue that Java is not), then you should be able to accomplish everything you need with only a trivial investment of time. This is especially the case if you know C++ or C#, both of which have similar syntax. The differences exist, but they are nothing that can't be overcome in a weekend of studying.

Well, if you are going to build a good interference machine that can solve a number of rules sets in a reasonable time in java, then I will be the first bying it ;-)

But apart from this about limitation of languages, I do not understand how you get from the use of 3000 distibuted computers, where it is not clear to me, how you come to the conclusion to assume such number from less 20 posters to this thread. It is absolut unclear to me, how this results in a conclusion, that for this tying to one language is the best solution. Best solution in what? In networking? In distribution?
I also do not understand what the knownledge of some specific langauge leads to ease of use of some different language, certain problems are not resonably solveable with?!
From my little opinion this is certainly not done with a weekend studying! Or can you make any proof of such assumption?!

Well, I am much too simple person for any such artistic stuff anyway, since I am already 41 and a little slow in thinking sometimes.

But all of your answer did not help in what others already have questioned... >> How can such a contest make sure anyone can participate in it no matter what language might be used?
Because some already thought about if it could be possible to make use of lisp, prolog, or whatever else prefered for the contest...
And my posting was one of the possible answers for this questioning, regardless of if you personally would prefer this or not.

But perhaps you might wish to provide some better solution in this sense?

[Edited by - Sebastian Gruber on October 13, 2004 6:56:05 PM]

Apart from my latest posting above, the cheating question is a good one. But since the challange is not defined so far, there are many, that humans can not solve because of complexity but are also difficult to solve for AI.

So I get to the point to ask the very questions...

* What is going to be solved by some AI in the contest?
* And how many participants are expected to be in the contest?
* Are the AI's competing with each other?
* Or do they soupposed to solve something any better than other AI's but on their own?
* If they are competing with each other, are there only two competing at a time or all together at the same time?

Perhaps there are some more questions ... ;o)

You are right Anonymous Poster, I was false in spelling and didn't notice that at this late hours (gmt+1)... :-/

My fist posting to this thread was following the idea that anyone might be able to participate no matter what language or architecture might be used. My comment about java at this time was all thought to be a provocative comment in this direction.
I personally would prefer else, but do not mind others to use what ever they might wish for a contest, if it is upcoming ever anyway?!

The question of cheating is from my little opinion more one of the type of challange, and for this not much an argument not to allow people using their computer at home feeding some environment/arena in the internet.
But sadly it didn't get that far with the discuss right now... instead it got more sort of the endless 'what is the best language' thing I guess...

It's nice you have various source lying around, but actually I already have what I need right now and can write myself most stuff I might want to have in addition... But thank you anyway. Well, of course I would give my 2 cents for it, as I said I would buy it, but do not have a need for such right now. ;o)