Pentland bases these conclusions on big data analytics, looking at social network interactions, cellphone usage behaviors, and other large-scale human activity datasets. He presents them in a new book, Social Physics.
"You do things for people, they do things for you, and that stuff is actually more powerful than economic incentives," Pentland said in an interview with InformationWeek. His work traces what he calls "idea flow" -- the way social groups spread ideas and transform them into behaviors.
On a micro level, this kind of information is useful for employers who are trying to get maximum productivity from their workers. Rewarding peers who encourage one another to adopt a new work tool is more effective than instructing them to adopt it, for example. Similar models work well for retailers (think viral marketing), and we've talked about the value in finding influential customers and letting them evangelize for a particular product or brand.
On a macro level, things get even more interesting. Pentland imagines a world where whole cities, transit systems, and societies are designed more effectively based on individual insights, not collective averages. He summed it up in a 2012 interview with Edge:
If you could see everybody in the world all the time, where they were, what they were doing, who they spent time with, then you could create an entirely different world. You could engineer transportation, energy, and health systems that would be dramatically better. It's this history of thinking about signals and people together, and how people work via these computer systems, and what data about human behavior can do, that led me to the realization that we're at a phase transition. We are moving from the reasoning of the enlightenment about classes and about markets to fine grain understanding of individual interactions and systems built on fine grain data sharing.
In an op-ed in Time this week, Pentland talks about how social physics can improve health insurance sign-up rates among the 18-to-34 age group, a demographic critical for the Affordable Care Act's viability. He says a financial incentive that rewards everyone in a neighborhood or group -- but only if everyone participates -- may drive young people to "sign up in droves."
In one community experiment, my MIT research group offered people financial rewards for healthier behavior. The result was a small improvement, but one that disappeared as soon as the experiment ended. However, in a second community, we gave people rewards only when their neighbors or members of their workgroup improved. The improvement was up to eight times greater, and perhaps just as important, the pattern of healthier behavior continued even after the experiment’s money ran out… Social network incentives raise community awareness and create social pressure to work together.
What do you think, members? We're conditioned to believe in rugged individualism, but the behavior patterns encoded in our big data seem to indicate otherwise. Is this kind of social science going to usher in a new era of human civilization, or does it smack of collectivism and military systems? Let's discuss!
— Michael Steinhart, , Executive Editor, AllAnalytics.com