Tuesday, July 05, 2011

Crowd-sourcing, video games, and the world's problems

This past weekend, I caught a snippet of a rebroadcast of this NPR story about Jane McGonigal and the thesis of her recent book. In short, she points out that as a species we have spent literally millions of person-years playing World of Warcraft, an online game that involves teamwork and puzzle-solving (as well as all the usual fun silliness of videogames). Her point is that in the game environment, people have demonstrated great creativity as well as a willingness to keep coming back, over and over, to tackle challenging problems (in part because there is recognition by the players that problems are pitched at a level that is tricky but not insurmountable). She wants to harness this kind of intellectual output for good, rather than just have it as a social (or antisocial) outlet. She's not the first person to have this sort of idea, of course (see, e.g., Ender's Game, or the Timothy Zahn short story "The Challenge"), but the WoW numbers are truly eye-popping.

It would be great if there were certain scientific problems to which this could be applied. The overall concept seems easiest to adapt to logistics (e.g., coming up with clever ways of routing shipping containers or disaster relief supplies), since that's a puzzle-solving subdiscipline where the basic problems are at least accessible to lay-people. Trying this with meaty scientific challenges would be much more difficult, unless those challenges could be translated effectively into problems that don't require years and years of foreknowledge. Hmm. Still very thought-provoking.


Amy Blum said...

There is already a pretty interesting game for RNA folding done by CMU and Stanford called EteRNA (http://eterna.cmu.edu/content/EteRNA). The winning designs for the week are synthesized and tested in the lab, with the results posted to help players in future weeks. I saw a really interesting talk on the game a little while ago. So far, the data shows that the mostly non-expert players score better than all of the predictive algorithms currently in use and even better than many experts asked to solve the same folding problems.

Guru said...

I am sure you are also aware of the polymath blog and its success!

Vincent said...

I think she is over rating the quality of problems solved in world of warcraft (WoW). One of the larger problems Blizzard faces in developing the game is that extremely simple scripts (typically refered to as bots) can emulate a player and automate the tasks of the game. The concept seems good on the surface, "millions of person-years", but what if you exchange the action to something else which is also very trivial. We have as a human race spent more then millions of years cleaning dishes, what if that action could be harnessed? Well it seems rather silly now doesn't it. Even though there are many years, there just isn't all that much challenge to the task. The same is true with the very linear simple tasks of wow. "Go to location A, Do task B 10 times, then go to Location C" Of cause typically phrased more epic for example "Go to the evil castle of Mordacai the wizard and slay his 10 minions of evil, and return to the queen of the forest to receive the sword of the sun".

I would argue that the triviality of the tasks is exactly why WoW is popular, and why the game keeps getting easier every time they update it. They don't want to challenge players as they will lose subscriptions.

Of cause you can make challenging online games, even where the tasks are relevant for science, but you won't attract the mass of WoW players. They are not in it to be challenged, but for social reasons, and for the rather trivial task/reward cycle which almost demands the task be insanely easy, very repetitive and therefore not very useful for real problem solving.

Surveillance cam said...

She wants to harness this kind of intellectual output for good, rather than just have it as a social outlet.