M.I.T.’s Alex Pentland: Measuring Idea Flows to Accelerate InnovationThe New York Times
By STEVE LOHR April 15, 2014, 4:54 pm
Alex Pentland has argued that not all data in the big data world is equal, by any means.In the 1970s, when he was a student at the University of Michigan, Alex Pentland worked on a government-financed research project that used satellite imaging to study animal life on earth. The focus was the beaver population and the animal’s movements in Canada.
“That’s not a bad model for everything I’ve done since,” Mr. Pentland said in a recent interview. “But I’ve been looking at people, not beavers. It has been using sensors to understand people.”
Today, Mr. Pentland is a computational social scientist at the Massachusetts Institute of Technology and director of the Human Dynamics research group at the M.I.T. Media Lab. For decades, Mr. Pentland and his graduate students have been attaching sensors to people to study their patterns of movement and communication in work and community settings.
In the early days, the sensors tended to be big and clunky. Over the years, the mobile sensors became steadily smaller and more versatile. The current generation, called sociometric badges, are only a bit thicker than a credit card, but they can detect location, movement and a speaker’s pitch and volume (words are not recorded).
Added to the mix have been the most ubiquitous personal sensors of all — cellphones, which became popular in the 1990s. “We understood what cellphones meant,” Mr. Pentland said. “Everyone was going to run around with sensors on them.”
That has led to a flood of information, a kind of big data. And in recent years, Mr. Pentland has been identified with concepts — and terms he has coined — related to the collection and interpretation of all that data, like “honest signals” and “reality mining.” His descriptive phrases are intended to make his point that not all data in the big data world is equal.
Reality mining, for example, examines the data about what people are actually doing rather than what they are looking for or saying. Tracking a person’s movements during the day via smartphone GPS signals and credit-card transactions, he argues, are far more significant than a person’s web-browsing habits or social media comments.
But Mr. Pentland argues that even the less valuable information in current flood of personal data could help open the door to what he calls “social physics.” That topic is the subject of his new book, “Social Physics: How Good Ideas Spread — The Lesson From a New Science.”
Central to the concept of social physics is the ability to measure communication and transactions as never before. Then, that knowledge about the flow of ideas can be used to accelerate the pace of innovation.
The best decision-making environment, Mr. Pentland says, is one with high levels of both “engagement” and “exploration.” Engagement is a measure of how often people in a group communicate with each other, sharing social knowledge. Exploration is a measure of seeking out new ideas and new people.
A golden mean is the ideal. High engagement alone, Mr. Pentland argues, can be perilous. “The great horrors of human history are places with lots engagement and no exploration,” he said. Ethnic violence and genocide are prime examples. Less dangerously, Mr. Pentland noted, the herd behavior that feeds stock-market bubbles is another example of a community with high engagement but without the new ideas that come from exploration.
In one experiment, Mr. Pentland and his graduate students tracked millions of trades on eToro, an online trading service for day traders. EToro also has a social network feature, in which traders can look up another person’s trades, portfolio and past performance. This feature, called OpenBook, allows for “social trades,” or copying someone’s trades.
The M.I.T. researchers found a spectrum of social trading behavior, which is a form of social learning. They ranged from isolated traders, with few or no social trades, to the traders who were the most densely connected, often to others with similar strategies to their own, a trading “echo chamber,” Mr. Pentland writes. Those with a balance of diversity of ideas in their trading network — engagement and exploration — had returns that were 30 percent ahead of isolated traders and well ahead of the echo chamber traders, too.
Mr. Pentland summons evidence from a range of experiments to support his thesis. The new data and measurement tools, he writes, allow for a “God’s eye view” of human activity. And with that knowledge, he adds, comes the potential to engineer better decisions in a “data-driven society.”
A better world through data. But Mr. Pentland does not ignore that powerful technology cuts two ways, including the possibility of a surveillance society beyond anything George Orwell imagined. The ability to measure, monitor and even control behavior is, Mr. Pentland writes, a classic example of “Promethean fire: It can be used for good or ill.”
Mr. Pentland is an adviser to the World Economic Forum on its initiatives on big data and privacy. He has proposed what he calls a “New Deal on Data,” which calls for individuals to own their data and control how it is used and distributed.
Big data, Mr. Pentland says, holds the promise of delivering major benefits in many fields, if privacy is protected, trust is established and the data is allowed to flow. “I see the public good that can come from using this data,” he said. “That’s why I got involved in this privacy stuff.”
http://bits.blogs.nytimes.com/2014/04/15/m-i-t-s-alex-pentland-measuring-idea-flows-to-accelerate-innovation/?_php=true&_type=blogs&partner=yahoofinance&_r=0
New SMAC quizzes available.
