Artificial Intelligence and mad computers hell-bent on universal destruction, a good ole’ staple of sci-fi, and for some reason computer games and computer gaming tend to feature prominently when movies are made on the theme. Tron, perhaps the first of its ilk, and WarGames, the seminal specimen, both feature computers dominating and threatening humanity by playing computer games – I can’t help but loving it! Back in the 70’s and 80’s, AI was one of the most promising not to forget frightening aspects promised by computerisation. It was hailed as saviour or destructor of our future, and though most people had an opinion on the topic, few held the view that “nah, nothing much will happen”. But this is what did happen.
Having studied a fair bit of Philosophy as well as CompSci, I’ve witnessed most traditional academic bastions for evolving artificial intelligence (Philosophy of Mind, “conventional” computer science and robotics), and really, their achievements were impressively absent. Rather, computer games was for a long time on the forefront of pushing the AI envelope. However, game development seems to have lost being the driver it was, and now nothing much really happens here neither. Robotics might be showing some progress, however, this seems to be in already established fields like specific pattern recognition (face recognition and voice recognition), and there’s basically nothing indicating quantum leaps or conceptual revolutions. Boring, and more than that, perhaps one of the biggest problems of contemporary computer games development: While our graphics grows increasingly realistic (whether this is for the better is another topic of debate) and ever-increasing computational potency allows fantastic physics and whatnot, the contents of the same games have, at large, stagnated or (arguably) even regressed. One common argument, which I with some reservations subscribe too, is that graphics, being the most immediate and obvious signpost of the game, assumes increasing focus and budget allocation, with detrimental effects on premises, story, narrative mechanics, game play, etc. etc. However, another side to it, which at least I haven’t heard mentioned is that today’s Artificial Stupidity is too insufficient to drive today’s massive games. Or is it simply that with being a larger business, with increasing number of titles to be (mass-)produced, there isn’t enough budget to develop AI; there aren’t enough skilled people to do the highly complex tasks of developing AI; the increasingly important middle-ware solutions and frameworks do not cover, or afford integration with, IA; or has the industry, in it’s collective growth euphoria simply forgot about AI? Or has it been convenient to forget about it when we can produce yet another MMORPG where we can replace dynamic content and intricate NP-characters with live human beings? Why is it that the online revolution, among other things, seems to have legitimised ignoring, even devolving, one of the most critical areas of computer games evolution?
I personally believe that game development could and should be one of the most important drivers of the development of artificial intelligence. Not only for the ever-stimulating reasons of Mad Science, but also because it would mean the potential for Actually Better Games, not as defined by graphic splendour, but as in better, more interesting game contents. The more graphics evolves, the larger, more intricate and environmentally impressive out game words become, the more salient and obvious is the discrepancy of these and that which populates them.
However, moving on on a wild tangent, I really do not doubt that eventually, AI evolution will have progressed to the point where we can actually speak about real, actual, no-bars-held artificial intelligence, rather than the Artificial Stupidity we’re tortured with today, and then games will land in science-fiction-land, because we’re going to have to start asking the Star Trek questions, those about the ethics of what we’re doing because we’re going to populate our creations with aware entities. Then the question for tomorrows game developer might have turned from “why is AI so lacking?” to “where does artificial intelligence end and virtual individuality begin?” and “can we really inflict violence on virtual individuals for our amusement?”. Nonetheless, I’d rather have that debate any day than suffer through another gorgeous game with retarded game play.