With enough data at your command, a large group of humans can be coaxed into accomplishing practically any task. Even, as Alexander Pentland recently proved, when that task involves driving around every town, city and suburb in the U.S. searching for enormous red balloons.
Late last year the Pentagon's mad-scientist research wing, Darpa, announced the Network Challenge, a $40,000 prize for the first group to find and report the locations of ten red weather balloons that the agency would set aloft one day in secret locations around the country. Most of the thousands of groups that signed up quickly realized that crowdsourcing was the way to find the 8-foot spheres. So, naturally, they offered bounties to balloon hunters.
But Pentland's crew at MIT's Human Dynamics Lab--part of the MIT Media Lab--took their crowd control a step further. In their incentive structure, anyone who reported a balloon to MIT received $2,000 if the team won. The person who recruited the finder earned $1,000. The person who recruited that recruiter bagged $500, and so on. The maximum payout with such a formula would add up to $4,000 per balloon, just affordable out of the prize money.
Video: Young, Cool and Techy Females
MIT's simple system meant that participants were just as eager to enlist their friends as they were to search solo. In the two days leading up to the competition the group recruited 100,000 volunteers.
Six hours after the balloon release, that horde of searchers reported the tenth balloon in a park in Katy, Tex. Three hours later MIT had sorted through its thousands of submissions and won the competition.
"It was trivial for us to slap together the balloon thing," says the 58-year-old Pentland. That's because other groups' tactics were based on guesswork, he argues. His were based on lessons learned through data-mining research. "We won because we understood the science of incentivizing people to cooperate."
Since 1998 Pentland has been engaged in an unusual blend of sociology and data mining that he calls "reality mining." His researchers place sensors that he's dubbed "sociometers" around hundreds of subjects' necks and install tracking software into their cellphones, capturing the movements of every individual in a group, whom he or she interacts with, even body language and the tone of his or her voice. Then they mine the resulting reams of data to identify facts as elusive as which member of the group is most productive, who is the group's real manager or who tends to dominate conversations.
"Data mining is about finding patterns in digital stuff. I'm more interested specifically in finding patterns in humans," says Pentland, who has a Ph.D. in artificial intelligence and psychology from MIT. "I'm taking data mining out into the real world."
That tag-and-trace method means Pentland doesn't have to rely on imperfect measurements culled from social networking websites or, worse, self-reported surveys. "Computer scientists understand data but not social dynamics. Sociologists understand social dynamics but lack instrumentation and analytics," says Sinan Aral, a professor at New York University's Stern School of Business. "Sandy is a visionary because he has both."
Last year, for instance, Pentland's lab put sociometers on 80 employees at a
So the call center tried its own experiment. Instead of staggering employees' coffee breaks as it had previously, it aligned their breaks to allow more chatter. The result, Bank of America told MIT a few months later: productivity gains worth about $15 million a year.
The lesson of the call center is the same one that Pentland's balloon hunters used to motivate thousands of followers to spend a Saturday assembling search posses: create incentives that lubricate information-sharing and teamwork. "The sociometric stuff told us what the facts really are, independent of the sociology and cultural clutter," says Pentland. "And some of the facts are surprising, like the fact that gossip improves productivity."
The red balloon victory and the call center experiment also demonstrate that Pentland's sensor studies can do more than fulfill his data fetish. Reality mining can teach us how to change reality. Upcoming projects will aim to encourage better energy use and health habits. "How do you get people to stop smoking or find a lost child? You leverage social networks," says Pentland. "We've studied human behavior, and now we're learning how to shape it."
