Game AI Roundup Week #13 2008: 6 Stories, 1 Quote, 1 Video, 1 Seminar

Novack on March 30, 2008

This holiday week has been rather quiet on the game AI front, but here at we’ve managed to dig up a bunch of fascinating Smart Links from around the web, including mainstream news articles and developer blog posts relating to artificial intelligence in games.

This roundup 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.

Gunning with Imaginary Friends

Clive Thompson of Wired, wrote commentary in the magazine’s Games Without Frontiers section:

“Can a machine think? That’s the question that mathematician Alan Turing posed in 1950, when he posited his famous Turing Test. He argued that artificial intelligence could be thought of as intelligent if it passes a social test — if it can fool a human into believing it’s real.

Alas, critics agree that no machine has passed the Turing Test. We’re never fooled by chatbots for very long, as the annual Loebner Prize contest proves. The thing is, we humans are awfully good at decoding social cues and detecting humanness; we can instantly tell when a preprogrammed “conversation tree” is repeating itself. That’s why many philosophers say machines will never pass the Turing Test.

Except, of course, for videogames. They’re filled with AI characters — enemies we confront, and teammates we play alongside. And the truth is, we often develop complex emotional and social relationships with AI characters inside games. I pretty much fell in love with Alyx Vance in Half-Life 2; whenever I play any Star Wars space-flight sim, I get enormously agitated at the fates of my teammates when they’re under attack.”

GTxA Link Madness

Andrew Stern from Grand Text Auto posted a two page compilation of links semi-related to AI and games, some of which have featured in AiGameDev’s weekly roundup already:

“I occasionally make posts composed of link dumps, to help GTxA readers find articles they might enjoy and may have missed. This time I need to split the dump into two parts, the first part being a set of articles ranging from the slightly over-the-top to the truly hyperbolic. I will gently attempt to challenge, refute or debunk each as I go. :-)

[The second part] is a more grounded series of links. First, pieces from three of my favorite game journalists: Clive Thompson, Chris Dahlen and Kieron Gillen.”

Novamente’s Virtual Pets

Celeste Biever from New Scientist Magazine, wrote an article on synthetic characters in virtual worlds from an AGI startup:

“If first impressions are anything to go by, synthetic pets like Novamente s dog will be a far cry from today s virtual pets, such as Neopets and Nintendogs, which can only perform pre-programmed moves, such as catching a disc. The problem with current virtual pets is they are rigidly programmed and lack emotions, responsiveness, individual personality or the ability to learn, says Ben Goertzel of Novamente. They are pretty much all morons. In contrast, Goertzel claims that synthetic characters like his dog can be taught almost anything, even things that their programmers never imagined.”

What’s missing from this story is a cynical game developer to keep those claims in check!

More Expressive Processing

Grand Text Auto

This week also featured even more content from Noah Wardrip-Fruin’s experiment of previewing of his book Expressive Processing: Digital Fictions, Computer Games, and Software Studies online. The the 8th chapter in particular gets into the juicy details about Façade (a Top 10 AI Game):

“I first met Andrew Stern and Michael Mateas at a 1999 symposium on “Narrative Intelligence” sponsored by the Association for the Advancement of Artificial Intelligence. The symposium was organized by Mateas and Phoebe Sengers, two of the final Oz PhD students. They managed to bring together a number of their mentors, colleagues, and friends with a wide range of people pursuing different facets of the intersection of narrative, character, and AI. The Zoesis team was present, showing off their most advanced demo: The Penguin Who Wouldn’t Swim. Bringsjord and Ferrucci discussed active development of Brutus. Stern described his company’s newest commercial product based on believable agent work: Babyz. Mateas and his collaborators premiered Terminal Time. It felt like the field was blossoming with new projects, pushing the state of the art to new levels.”

“The surface experience produced by the Façade’s processes and data is shaped by a series of choices that have clear impacts in terms of the Eliza and Tale-Spin effects. The results are instructive.”

Impersonating Superstars

GameSpy editor Gerald Villoria wrote an article for the XBox360 section entitled WWE Smackdown vs. Raw 2009, where he comment something intresting. It seems like the game AI its not only pointing to mimic human behavior, as now also try to imitate the manner of acting of specific individuals.

“Improving Smackdown vs. Raw’s artificial intelligence is also a focus this year, and the publisher’s goal is to release a game with an AI system that features more intelligent opponents that actually act like their real-life counterparts. Part of this system will revolve around Superstar-specific A.I. The Undertaker was utilized as an example. The plan is for him to use signature moves in key situations, so that when the fans would expect him to use his old-school move, the A.I. will do so.”

Anyone here notice these features while playing the game?

Unangband Monster AI

Andrew Doull published the final article of a series of posts analyzing his “Unangband” Monster AI:

“Well, what conclusions can I draw from my experience so far with the Unangband monster AI? I’ll stick with listing a set of principles that have proved to be a useful guiding framework for anyone else working with roguelike (and other game) AI.”

He goes through lots of practical tips that should be useful at almost any stage of development.

Teaching AI

In his last editorial, Videolamer talks fondly about Sins of a Solar Empire. However, the editorial also points to one more emergent consequence of a poorly designed AI: the game being harder to learn.

“For various reasons, the game is very difficult to learn to play properly. In no particular order, let me throw a few out there:

1. The single player AI is too easy, to the point you don’t need to utilize counter units, so you don’t learn much about units

2. There’s no campaign to make you learn to use different units in different situations to make up for the general AI being terrible”

Game AI in Zox Universe

This video was recently posted in the increasingly busy forums by the developers of Zox Universe. Here’s the video caption:

“5 computer players playing :). We are currently working on AI so we can make single player for our ZOX space.”

The State of Procedural Content

Create. Power. Play. is the tagline of this free webinar (online seminar), sponsored by HP and Intel and hosted by Gamasutra. The call is on Wednesday April 2 and lasts 1h, featuring Chris Delay from Introversion Software:

“The State of Procedural Content: In this in-depth technical Q&A;, Introversion Software co-founder and lead architect Chris Delay (Darwinia, Defcon) will discuss the use of procedural algorithms and techniques to create assets in some of the UK developer’s intriguing upcoming games such as Subversion.”

Let’s hope HP’s sales pitch isn’t very long!

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

Discussion 1 Comments

Andrew on March 30th, 2008

A shame Sins of a Solar Empire has hardcoded AI. Due to the less-micromanagement, more-macro-strategy, it'd be a good one for modding or improvements. Although lack of a campaign is a real sore point with me TBH :( The Unangband stuff is interesting - the final post is a nice summary if you want to skip (or don't know enough about) the rest.

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!