In the build up to GDC 2008, there are quite a few smart links relating to industry news. This roundup is brought to you by Mark Wisecarver and Alex Champandard. If you have any news or tips for next week, be sure to email them in to: editors at AiGameDev.com.
Grand Theft Auto IV and Euphoria
IGN UK has an article about the animation system in Grand Theft Auto 4. It’s overly sensationalist with regards to Natural Motion’s technology, but the article is worth a read, notably because of its references to the AI:
“When you combine this reactive AI with the bustling street life, you get fascinating crowd mechanics. Everyone seems to be doing something unique; there’s much less of the repetition of character types as compared with past entries. People carry bags of groceries, read magazines and hold conversations, changing the pace of their walk and their task on the fly. They all react to what Niko does around them; opening fire within earshot of pedestrians causes a panicked flurry of activity — something improved since San Andreas thanks to more circumstantial and conditional AI.”
It sounds like an unavoidable title for 2008!
Procedurally Animated Spiders
Here’s a little demo of procedural animation specifically for 8 legged insects, built using Newton as the physics engine and Ogre 3D for graphics.
Evolving PacMan AI Using Neural Networks
In the GameDev.net forums, Kirk DeLisle announced a release of his latest project on Sourceforge. According to the project description:
“A Pac Man clone in which the AI for PacMan is evolved using Neural Networks based on NEAT.”
This version in particular features increased stability and configuration options for playing around with the system.
EMotion FX Animation Middleware v3.6
MysticGD, a leader in real-time character animation middleware, today announced the release of its latest version of EMotion FX, the real-time character animation and model export system.
“EMotion FX v3.6 provides intuitive and efficient tools for artists, increasing runtime performance and pushing greater detail to game audiences. EMotion FX v3.6 also showcases a powerful level of detail (LOD) system that allows the developer to tune their model skeletons for optimum runtime performance, while giving artists a workflow pipeline in which they can quickly setup complete in-game characters.”
The Core of Game AI?
A game development team of students from the Nanyang Technological University (NTU), Singapore wrote about their take on artificial intelligence in games:
The core of the in-game A.I can be broken down into 3 smaller parts: 1) Path-finding, 2) Triggers and 3) Combat
Artificial Intelligence at 3F Studios, 3fstudios.wordpress.com
More Details about Neurotic AI…
As reported a few months ago, neurotic AI seems to have surfaced as the winner in a somewhat contrived RTS experiment.
“The Austrian researchers want games to be more engaging by having emotional, not just coolly calculating, computer players. Instead of just challenging your rational planning and decision skills, you?ll have use your emotional intelligence too. They created aggressive, defensive, normal and neurotic versions of the AI software in the war strategy game Age of Mythology, drawing on “the big five” emotional dimensions to personality recognised by psychologists.
The bots are able to switch between states of pleasure, pain, clarity, and confusion in response to events. The strength of particular emotional changes is related to the overall personality. The neurotic bot was more likely than the others to distort hard facts about resources — like the amount of timber around — and flip between extremes of behavior. And it was better than the rest.”
You can also find the PDF slides from the original research below.
Emergent AI in Crysis
Crysis isn’t selling too well, and certainly has room for improvement in its AI… But it got many things right:
“We build emergent AI for Crysis. It means that if the player chooses different tactics in order to try to beat the enemy, the AI will respond to each tactic differently. Every play session is essentially different. As Crysis is a realistic game, we tried to model the AI perception as realistically as possible. The player can create sounds which will make the AI to investigate the source and the AI really needs to see the player before it fights back. That allows the player to have great fun playing stealth. Another challenge for Crysis AI was the dynamic and breakable world. We put in a lot of effort that the AI is able to fight back in the midst of the player created havoc of broken huts and fallen trees.”
Have you played the game? What do you think?
Kynapse 5 Set for GDC Release
In the build up to GDC 2008, Kynogon is making the most noise with a new version of its middleware:
“Middleware developer Kynogon has announced that it is preparing to release the latest version of its AI middleware, Kynapse 5, at GDC 2008.? The latest version will also offer a new parallel processing framework for multi-threaded and multi-core console and PC architecture, offloading AI pathfinding calculations to secondary threads and processors.”
Kynogon Preps Kynapse 5 For GDC Release, gamasutra.com
Factor 5 Selects xaitment for AI
Game developer Factor 5 turns to xaitment for artificial intelligence middleware.
“Interactivity is at the heart of games and distinguishes them from film. But the ability to simulate thousands of characters both graphically and with accurate physics is meaningless if their behavior is unconvincing,” said Julian Eggebrecht, President of Factor 5. “The step up from mindless zombie to a clever human or alien adversary is the challenge for games of the future. To address this aspect of game design one of our main efforts is the development of a cutting-edge AI system.
Jobs in Industry and Academia
There are also many jobs going for people interested in game AI. Some positions require more experience than others:
Stay tuned next week for more smart links from around the web.