Game AI Roundup Week #39 2008: 9 Stories, 1 Video, 1 Quote, 1 Game

Novack on September 28, 2008

Weekends at are dedicated to rounding up smart links from the web relating to artificial intelligence and game development. This week, as always, there are some good articles and blog posts for you to read. Remember, there’s also lots of great content to be found in the forums here! (All you have to do is introduce yourself.) Also don’t forget the Twitter account for random thoughts!

This post is brought to you mostly by Marcos Novacovsky (aka “Novack”). 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.

F.E.A.R. 2 AI “second to none”

Interviewed by, F.E.A.R. developer Monolith has boasted that the enemy AI in its upcoming sequel, Project Origin, is “second to none”.

Speaking to XBW, associate producer Eric Studer said the AI in the second is going to “keep you on your toes at every turn”. After playing against the impressively smart baddies in the first game, we believe him.

“The AI in Project Origin has been significantly expanded upon from the first game, and reacts even more realistically to how you play,” claimed Studer.

Pure artificial intelligence

Chad Sapieha from Controller Freak (, wrote praises for the Disney Interactive’s ATV racer game’s AI.

With few exceptions—such as the occasional rider who would obstinately try to follow a particular line despite my locking wheels with him—I could hardly tell that my opponents’ intelligence was rooted in silicon chips rather than grey matter.

And as the game progressed, they got even smarter. By the last of the ten stages in the World Tour mode, making just a couple of mistakes would send me plummeting to the back of the pack, with virtually no hope of ever overtaking the leaders. This is as it should be in a racing game’s final events, when the player ought to know the tracks and possess skill with the controls.

Indie side of the force

This week, we have a post on the James Ford’s developer plans, who have beeing working on a behavior trees editor for Torque Game Builder.

If you are interested in game AI and haven’t heard of behavior trees you might want to check out In short they are the new, sexier finite state machines, basically a good way of laying out behaviors for your game agents in a nice modular and reusable way. I felt like messing around with them and maybe a few other AI techniques for some experience, and ended up with this…

Memory-efficient pathfinding

In his blog, Christer Ericson posted yet another insightful article on pathfinding.

Third time’s a charm they say, and as I’ve talked about pathfinding twice before (in Don’t follow the shortest path! and Aiding pathfinding with cellular automata) I thought I’d charm everyone with a third pathfinding article. This time I’ll talk about how we can reduce the memory consumption on a commonly used pathfinding algorithm by about a magnitude!

Giveaway: 20Q Prize Pack

Through I found that setted a prize to bringing players to the Star Trek version of its “telepath” AI game. is putting its artificial intelligence (A.I.) into warp speed to bring players 20Q™ STAR TREK, the latest version of the hugely popular game that can “read your mind.” By choosing a person, place or thing from the Star Trek Universe, you use your wits to keep the system from figuring out your answer in twenty questions or less. If the system can’t figure it out, you win! But first, the game needs to bone up on Star Trek trivia to be a worthy opponent.

20Q™ needs your help to assist the game in “learning” the world of STAR TREK, and all you have to do to contribute is play! With each play of the game, the system gets “smarter” and more challenging. By simply testing your STAR TREK knowledge, you can help bring the game to life as a real 20Q™ home game!

Think you can stump the system and win big?

New World Strongest Chess Program?

Susan Polgar, on the Chess and Information Blog posted an article about the latest news on the Chess AI field: humans debunked, now its time for the machines to compete between them for the title.

Deep Delta is being designed to run on eight 64-bit processors and will be able to analyze millions of positions per second. Programmers are using ultra-fast games against rival programs, to create “an extensive opening book that will leave Delta a fraction of a pawn ahead when it leaves book.” Combined with extensive alpha-beta pruning and endgame logic, VersaGlobe hopes to beat any competitor, human or machine, with a low probability of draws.

Flash AI Game: Light Bot

Another addictive game, where we need to program the AI of a robot. Interesting for first approachs to programming, the flash development also delivers lots of fun. Only 12 levels though…

Artificial Intelligence is hard to program. Not every bot ever created can maneuver and function on it’s own. Rather some bots run along a path that the programmer presets for them for various situations.

You’re job is to light up all the blue tiles in the factory by the commands you issue to the light-Bot. Good luck.

Funny Quote of the Week

In a review of the EA Sports title Facebreaker (Xbox 360) Jakob Paulsen wrote for Boomtown:

“However, there’s a catch. The AI is very uneven and it almost feels like it’s cheating. With practice you can get the first two knockdowns done pretty easy - but after that the AI speeds noticably up and you have to fight like a madman. If you get cornered, you’ll hit the floor faster than Mike Tyson can bite an ear off.”

AI & Xbox 360

In the blog VG Wire the author posted some sort of rant on the AI capabilities of the console.

Let’s leave the format war between Sony and Microsoft and explore a gaming phenomena that unquestionably improves current gaming trends — Artificial Intelligence, or AI to many. Touted to make new breakthroughs and trends in the world of gaming, AI is the enabling of intricate decision-making capabilities in a software. Speaking of AI, the Xbox 360 boasts of an AI-capable system designed and developed by many brilliant and innovative engineers. Though the Xbox also has this feature, it is more pronounced and better in the new console combined with the HD DVD technology.

Roy Taylor on game physics, AI & piracy published an interesting interview to Roy Taylor (Nvidia and PC Gaming Alliance).

At this year’s Nvision conference in San Jose, I managed to catch up with Roy Taylor to talk about some of the current issues surrounding the gaming industry. Roy is vice president of content relations at Nvidia and is also head of the marketing sub-committee in the PC Gaming Alliance, which means he spends a lot of his time talking with gamers, developers and some of the big industry decision makers.

One of the biggest movements in the industry at the moment is of course the advent of game physics – the technology has been around for a while, but its awareness has sky-rocketed over the past twelve months and we’re now starting to see GPU-accelerated physics appearing in games.

We also talked about artificial intelligence and how games are going to become much more immersive as and when technology allows developers to do more before finally touching upon one of the biggest issues facing PC gaming today: piracy.

What follows is an condensed version of the thirty minute discussion we had about the various issues described above.

Why Game AI?

The Niss LiveJournal (a journal of a game programming project, an offline role-playing game with a huge world and advanced AI) strikes again, this time published an opinion on the roles that the game AI could have in an RPG.

As I said last entry, I’ve been questioning what role better game AI could have in an RPG. In an action game, an NPC could have better combat tactics, so that’s one area for improvement; but what about RPGs? Most potential advances seem to be things that players won’t notice or that could be replicated with a stupid random system. For instance, I could determine price fluctuations in a trading game by having a fancy AI-driven economic simulation, or by just having some hard-coded start values, random fluctuations, and a couple of story-related modifiers. The second option would probably look about the same!

Polyworld: Using Evolution to Design Artificial Intelligence

Note: the video seems to be working intermittently on embedded, if you can’t see it, please go to the youtube address, the video up there.

This presentation is about a potential shortcut to artificial intelligence by trading mind-design for world-design using artificial evolution. Evolutionary algorithms are a pump for turning CPU cycles into brain designs. With exponentially increasing CPU cycles while our understanding of intelligence is almost a flat-line, the evolutionary route to AI is a centerpiece of most Kurzweilian singularity scenarios. This talk introduces the Polyworld artificial life simulator as well as results from our ongoing attempt to evolve artificial intelligence and further the Singularity.

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!