Game AI Roundup Week #25 2008: 9 Stories, 1 Paper, 1 Video, 1 Event

Novack on June 22, 2008

Weekends at are dedicated to rounding up smart links from the web relating to artificial intelligence and game development. Good material this week; you’ll find many insightful articles, a video, a paper, more data on the Paris Game AI Workshop, and interesting news; not forgetting random thoughts from Twitter.

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

BrainWorks: Give It A Go

Recently, a wave of interest has arise around the AI development og the ancient game Go. This time, Ted Vessenes, in his blog BrainWorks, wrote about his perspectives on the matter.

While my primary area of expertise is First Person Shooter AI, I still get a number of questions about other AI applications. One of the hardest game AI problems to date, if not the hardest, is the game of Go. What’s so surprising is that Go has simpler rules than Chess, but Go AI is several orders of magnitude more complicated than good chess AI.

Rainbow6 AI

In the Game Review column of Baller Status, Prateek Sanan and Rohit Loomba checked out Rainbow Six: Vegas 2, and wroted this about the game’s AI:

What does detract from this game, however, is the artificial intelligence. Computer controlled characters are repetitive and can be found in the same place doing the same thing if you die and pick up from a previous checkpoint. Computer controlled characters also seem unaware at times, making it possible to pass suspiciously close to them without them firing a single shot. More automation of the characters would make the game more challenging and fulfilling.

Paris Game AI Workshop Schedule Announced

GameDev.Net published the event schedule, which start on June 25th. Sign-up immediately as there are only a few extra places left!

The finalized schedule for next week’s game AI event, organized by and LIP6 University, has been announced (PDF). The day features five invited talks, two panels with top researchers and developers from industry, as well as an open roundtable at the end of the day. Attendees include programmers and designers from Crytek, Ubisoft, 2K Paris, Guerrilla, Ankama, and Arkane.

Intel researchers shine light on ray tracing

Brooke Crothers from CNET, interviewed two Intel personalities: Jerry Bautista (co-director of the Tera-scale computing research program), and Daniel Pohl (researcher). At some point, questioned about Intel’s upcoming Larrabee processor, this is what they commented on AI applications.

Finally, he talked about the application of Intel processors to game artificial intelligence (AI). He said people seek online games because they want intelligent interaction. “Online games are so predominant because the computer is boring. It doesn’t do things that are interesting and peculiar and unpredictable (in the same way another person is unpredictable). What if the AI engine in the game was sophisticated enough that it did things that were interesting and unpredictable? You would be happy playing that game and not going to the Internet.”

Sounds like a naive comment on the matter, or Intel is trying to recreate HAL within its new processors.

Real racing in the virtual world

Jonathan Fildes, Science and technology reporter of BBC NEWS, wrote an article on a new gamming technology of iOpener Media, which offers the chance of playing “within” real matches. Very interesting reading; this are the comments on he AI system:

The company does not intend to develop its own games; rather it will provide the backbone for games developers to build on to.

But it will provide some software; specifically an artificial intelligence (AI) program to make sure that the virtual and real worlds blend seamlessly.

“If Hamilton is driving behind you he can’t see you [in the game], so he would drive right through you,” explained Mr Lurling.

“So the AI takes over at that point and you see a very realistic overtaking.”

The system also handles the results of in-game collisions between real and virtual drivers.

Lara returns

Video Gaming Faqs published a preview on Tomb Raider: Underworld. Some short comments on AI, that are interesting nevertheless:

Combat looks extremely vicious, with panthers attempting to claw Lara to shreds and poachers taking a break from hunting to make her life difficult. The a new enemy artificial intelligence won’t let them stupidly run into gunfire either. It’ll react based on the situation, running for cover or attempting an ambush.

Ninjitsu AI

DeVOTeD 2 Gaming published a review on the XBox title Ninja Gaiden 2, where the AI is -surprisingly- praised.

The AI of the enemies was refreshing to see though. Instead of them sitting around waiting to be attacked…they willingly initiate combat with you and attack with everything they’ve got. You actually feel like you are defeating enemies and forget that they are just mere artificial intelligence drones walking around the game.

It is better to be alone…

Steven van Hemert, from Nudjit, wroted for the Games section a review on Battlefield - Bad Company. Interesting comments on the game’s AI:

Too many titles suffer from easy AI, which ultimately just makes the game a matter of point and shoot, with few strategic requirements. In Bad Company, enemy AI is sharp and creative, making the need for foresight and clear strategy absolutely essential in securing survival.

Enemy combatants will also adapt to your strategy, and set ambushes for your advance, or move tanks forward should you take cover and dig in. There seems to be a counter strategy for just about any move you might make, and an approach for every type of soldier you deploy.

Though you are part of a 4-man squad in Bad Company, the mechanics make it an easy style of play. You do not control or provide orders for your company – they seem to figure out your intentions and either follow you, or assist your efforts by flanking enemies or laying down cover fire. The frustration of playing with 3 other AI characters has also been minimized with a slightly jarring twist in game time; when getting into a vehicle, your squad will all instantaneously enter the vehicle with you, negating the time wasted waiting for squad members or driving around collecting them all. In the few hours I played, I experienced none of the usual headaches with companion AI, a massive relief.

Reddit debates

A very interesting Reddit discussion on Fast Ranking Algorithm.

New Paper

Interesting work, shows how to rank many different solutions in O(N) instead of O(N^2).

A fast algorithm for learning large scale preference relations.
Vikas C. Raykar, Ramani Duraiswami and Balaji Krishnapuram
Download PDF

Intelligent decisions published an AI middleware roundup:

ROUND-UP: AI middleware

Always interesting, at present the market for games-related artificial intelligence technology is even more fluid than usual

Video: Flash AI

A video tutorial (first of a series) introducing to the practical devlopment of AI in flash applications.

Stay tuned next week for more smart links from around the web!

Discussion 0 Comments

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!