Game AI Roundup Week #15 2008: 5 Stories, 2 Videos, 1 Demo, 1 Quote

Novack on April 13, 2008

Weekends at are dedicated to rounding up smart links from the web. This week there are a bunch of great blog posts, Flash demos and videos of Google Talk lectures.

This post is brought to you by Novack and Alex Champandard. If you have any news or tips for next week, be sure to email them in to editors at Also Remember there’s a mini-blog over at (RSS) with game AI news from the web as it happens.

Game AI is Like Parenting

Ted Vessenes writes about his recent experiences on his BrainWorks, a Quake 3 based bot AI project:

“Writing artificial intelligence is a lot like being a parent. It requires an unbelievable amount of work. There are utterly frustrating times where your children (or bots) do completely stupid things and you just can’t figure out what they were thinking. And there are other times they act brilliantly, and all the effort feels satisfying and well spent.”

Also be sure to check out Dave Mark’s follow-up blog post too.

Starcraft: Bot Fight

On his blog on videogames and geek culture Twenty Sided, Shamus Young posted an article where he relates the results of a curious experiment on AI: a typical Starcraft scenario “The Hunters” — where only AI players actually play the game, and a human observer is there to watch the results:

“I’m not sure who will find this interesting. This is an AI analysis of a ten year old videogame. This entire endeavor will sound absurd to people familiar with the game in question, and hopelessly esoteric to those that aren’t. Still, I’m putting this up in case there is someone else out there who is just as peculiar as I am, in that I find this sort of thing intensely compelling.

About a month ago I wrote a Starcraft scenario which allowed you to observe a game between AI players. I’ve been curious about the quirks in the Starcraft AI and I’ve wanted a chance to see them do their thing in a deterministic environment. I learned some surprising things about this ten-year-old gem. While the races themselves are very nearly balanced in the hands of humans, it turns out the AI is a lot better at using some races compared to others.”

Strategic Victory for Computer Go

The game was played over nine rows and nine columns, however Catalin Taranu (the nine-stone handicap Go Master) beat the computer in a 19

Discussion 1 Comments

jamesford42 on April 14th, 2008

Excellent video. I was intrigued by the final statement he made, "can a system learn which abstractions of a state-space are worth learning?" In explanation for those who didn't watch... For a system to learn it must have abstractions of the otherwise too-large game-state space -- and the abstractions must be ones actually useful for the type of learning we want to allow. Currently in Game AI ( and his research ), these abstractions are hand made. Now, if an AI could "learn the state-abstractions valuable for learning" ... THAT would be something.

If you'd like to add a comment or question on this page, simply log-in to the site. You can create an account from the sign-up page if necessary... It takes less than a minute!