The success in Von Ahn's idea is in its simplicity. To make gamers compute means making games that exploit our inherent love of puzzling and communication. Without another person to play with, The ESP Game would not be particularly interesting, but as soon as we're competing against and working with other people, we are interested, no matter how abstract the process might be. The web makes this kind of application all too easy to deliver, and there are dozens more applications for this kind of practical, computational game design in the future. Von Ahn, for example, suggests that his principles could be applied to language translation, where a game could be designed to facilitate people in translating phrases into other languages on a competitive basis. It could make for a fun, cerebral language game, as well as a potentially useful service. The biggest limitation, as with games generally, will be the imagination required.
However, games could have other applications within computation - applications based on the mass of human traffic, rather than the individual cognitive abilities of human brains. They could, for example, be used to demonstrate economic theories. Infamous space-war MMOG EVE Online has long emphasized the way in which its virtual markets operate similarly to those in the real world. EVE's marketing team has even begun to describe the game as an instance of so-called "smart gaming" - extrapolating the way in which its open economy and complex financial markets mimic trends we see in the real world. It's not only a useful case study for those people wanting to understand economics, it's a working economic sandbox in which business theories can be deployed and tested. It's all very well running tour ideas against a computer simulation, but why not get EVE's 200,000 real people to test it for you? There are other possibilities too: viral spreads in virtual populations, marketing test-beds, surveys - almost anything that needs a large decision-making aggregate to function.
Sharper readers will have spotted that all this ties into another trend: the general notion of human computing and the rise of the wiki. Yes, it's that "Web 2.0" stuff that's been coming out of our ears for the past three years. The sheer number of people is the most creative force on the net, and enabling them to individually take on tiny amounts of work that they're interested in is what makes projects like Wikipedia or Second Life possible.
While "Games With A Purpose" might be making life easier for Google's search technicians, the overall trend of encouraging gamers to compute is already feeding back into the one thing gamers really care about: playing games. And I'm not talking about games with a built-in level editor, I'm talking about games in which the play itself generates new content for other gamers. Just by playing you're going to end up generating the game for yourself. Take Spore's touted capability to "pollinate" the galaxy of a single-player game with the cultures created by other people playing the game, or look at how Little Big Planet gives players access to a world-morphing toolset at all times.
Perhaps, as games mature even further, we'll see multiple computation applications for the same game. We'll be crafting the game world, processing images, modeling social science theories and maybe even handling some of that distributed computing stuff during our AFK periods. But, of course, that's a thinking man's game.
Jim Rossignol is an obsessed gamer and a contributor to the PC gaming site Rock, Paper, Shotgun.