When University of Washington biochemist David Baker needed help predicting the structure of proteins, he did not turn to his colleagues. Rather, he decided to let the whole world participate.
Increasingly, scientists are relying on such “crowdsourcing”—calling on ordinary citizens to volunteer their help in addressing complicated problems. In Baker’s case, he helped develop Foldit, a computer game that challenges players to wiggle and shake protein chains into stable structures. In August a paper in Nature revealed that Foldit players, most of whom had little or no biochemistry education, surpassed or matched the performance of a sophisticated protein-folding algorithm on 8 of 10 puzzles. “People are better at analyzing the whole situation,” Baker says. “Computers just approach problems randomly.”
Volunteers for the Galaxy Zoo project have classified a million images from the Sloan Digital Sky Survey, leading to about 20 scientific publications and one genuine enigma: a peculiar green intergalactic blob. Other crowdsourced projects include labeling aerial photos of Mongolia in a quest to find Genghis Khan’s tomb and improving climate models by poring over World War I ship logs for weather information.
Government agencies are getting in on the action too, listing projects on a new Web site, challenge.gov, and offering prizes. In July a retired engineer from New Hampshire won $30,000 from NASA for a model that forecast solar activity with 75 percent accuracy. “There’s a huge appetite from people who aren’t scientists to actually get involved in science,” says Galaxy Zoo principal investigator Chris Lintott.
