Fast-Forward

Thirty years after games were widely introduced to the public, electronic games are now "fully entrenched in popular culture," Perry says. Games today have recognizable brands, such as Mario and Metal Gear Solid, with franchises such as Madden Football returning with the regularity of a birthday. These are big, successful games. They are also generally thought to be good games. But do we think about games any differently today than we did three decades ago?

Georgia Tech has an Experimental Games Laboratory, which Michael Mateas, Ph.D., of the School of Literature, Communication, and Culture and of the College of Computing, runs. Mateas' focus is artificial intelligence-based art and entertainment, melding research with art, thus creating what he calls "expressive AI." Mateas' research reaches back through the annals of game history to determine where games might be headed, specifically related to AI.

"A lot of what I'm really interested in is genre innovation, so I'm trying to create experiences that aren't clearly action adventure, RPG (role-playing game), RTS (real-time strategy), FPS (first-person shooter), or any of these fairly well-defined marketing categories for commercial games." Mateas defines experimental games as "radical genre innovation that requires both design and technical innovation."

Eric Zimmerman and Katie Salen’s The Rules of Play.

One theory, at least among gamers, is that true innovation is thwarted by developers' needs to build games that fit within the ordained genres for sales purposes. Murray thinks that new genres are forming, however. "I think it's very telling to me in Eric Zimmerman and Katie Salen's new book, The Rules of Play, where they come close to excluding Sim City as a game, because it doesn't have a winning condition," says Murray, suggesting that by the same standard, The Sims might also be excluded. "If you're going to exclude one of the most popular games of all time...then it's because 'game' is too narrow a word. And I think that we'll start to think about games the same way we think about movies--as a sort of metacategory where there are a lot of different categories. I think there are things that are emerging that are new forms, and it challenges the boundaries of what we're used to thinking of as games."

Pac-Ghosts exhibit early (and spooky!) AI.

"AI" is another term that gamers and those interested in selling games throw around loosely, as compared to academic definitions of the term. "I do believe in a sort of consistent idea underneath the term AI," says Mateas, "which is this idea of a psychological project. What [marketers] mean is smarter enemies--maybe characters that you can find more engaging. Many of the techniques that are used in games that are called AI aren't really recognized by academics as being AI," Mateas explains. "I still think it's fair to say the game AI drives the part of the experience that you, as a player, can sort of project an inner life onto," as opposed to physics, which is commonly mistaken for AI. Psychological projections, according to Mateas, are indicative of true AI. Does your character have goals? Is another character doing something to you? Is a non-player character (NPC) mad at you? Assuming that AI is the child of more-advanced technology is a misnomer, because AI is not necessarily reliant upon complex lines of code. "You could have a really beautiful and complicated water effect in a game, with really nice ripples and shading effects, and, of course, there's some complicated stuff under the hood to make that happen," Mateas says. "But the player, when he or she is experiencing that water, doesn't have to project any internal state onto it. Whereas with the ghosts in Pac-Man, where the code is much simpler than the water effect, you do project a state onto them."

How can we learn from this to make deeper game experiences in the future? Mateas believes, coming from an AI background, that some of the more complex techniques from academic AI can now be "reexamined, reexplored, changed, and used to achieve even richer psychological projections." The future of games might be less about stunning physics and more about meaningful character experiences, or perhaps a balance of both. In this sense, The Sims can tout more AI than a first-person shooter, wherein enemies know when to approach you and when to retreat. "At the end of the day, trying to create this illusion of sort of a living world is really what AI is about," says Mateas.