Hi there,

I am always watching the ongoing open-source-AI projects, but I wonder why noone has come up with something like "hunting-grounds" yet.

So what is "hunting-grounds" you ask? Well it's something that came to my mind while playing with golem@home and DaliWorld. I've seen most fascinating implementations of evolutionary life simulations and AI based on neural networks, but I'm always missing something: an application that really appeals to the average user.

People seem to enjoy virtual aquariums a lot and deerhunter was one of the most successful games in 1999, but none of them feature AI or digital life.

What I'd really like to see would be a game-oriented environment which combines elements of player-interaction with artificial lifeforms. And I am not talking about dots in a raster here, but about 3D-terrain and actually moving/walking lifeforms. Couple that with a p2p network where lifeforms could roam and experience all different kinds of, well, experiences and an editor where people could add, breed and train lifeforms and you'd have an interesting project.

A big attraction could be to hunt down these lifeforms (although I'd probably just watch them most of the time) which would put some real pressure on evolution as a side effect.

Why do I consider this to be such a good Idea? Because evolutionary AI software really depends on the amount of interactions it is exposed to. Considering the vast amount of users and interaction that could be gained from a software as described above I'd expect to see some rather surprising and fascinating lifeforms pretty soon. And the field of AI could also gain some publicity by this. I can't stress how important the appeal to end-users is, especially for open source projects. Once something is heavily used by normal end-users, development get's a whole new vitality.

I'm a java-programmer specialized in web-based business applications, so I am at least capable of understanding the principles of the software which would be required to create something like "hunting-grounds". I am sure it'd be possible to create it right away with todays technology, but none of the projects I know of go into that direction.

I'd love to here some opinions about this. Greetings,

Konrad