Social Physics: How Good Ideas Spread-The Lessons from a New Science

Social Physics: How Good Ideas Spread--The Lessons From a New Science Alex Pentland (New York: Penguin Press, 2014) The impact of social media on us and on our institutions is a popular subject. We all have read about Twitter feeds that affected stock markets or nurtured movements that overthrew governments--the rise of the Arab Spring, for instance. Facebook has experimented with news feeds, exploring their influence on mood in a controversial 2012 study and, in an experiment during the 2010 US Congressional Elections, their power to encourage users to vote. More recently, Facebook invitations to take the "ALS challenge" (dumping a bucket of ice water over your head) accompanied by daily video posts from celebrities and friends made ALS the hottest cause of the summer, raising more than $115 million for ALS research. These phenomena raise compelling questions, and there is a pressing need to increase our understanding of them and of the effects that social pressure via social media may have on behavior. There has been much speculation about these possibilities, but Alex Pentland's Social Physics is the first book to offer concrete answers, based on mathematical modeling and living-lab experiments conducted at an unprecedented level of data intensity, with large numbers of measurements per person per minute. Pentland, who is a professor of Media, Arts and Sciences at MIT, conducted the research with his students at MIT and collaborators in various organizations. He also appears to have put his results into practice, as the book often offers examples from companies he founded or cofounded. The title of the book reflects Pentland's guiding analogy: the goal of traditional physics is to understand how the flow of energy translates into changes in motion; social physics seeks to understand how the flow of ideas and information translates into changes in behavior. Just as physical relationships can be modeled mathematically, social physics is also amenable to modeling. In fact, Pentland has constructed a mathematical model of social influence, expressed in terms of a conditional probability between each person's (hidden) state of mind at any fixed point in time (t) conditioned on each other person's state of mind at that point in time and a previous point in time (t--1). This conditional dependence represents the "influence strength" of one person over another. The influence model describes how an idea spreads, for example through the many digital and interpersonal social networks that permeate an organization. Using measures of the amount of social interaction to estimate social influence in this way produces accurate predictions of future behavior. Thus, the model can be used to determine ways of designing or tweaking social interactions in order to increase influences for desired behavioral changes. The modeling process is well documented; the mathematical details of the model are provided in an appendix, and Pentland references an extensive set of papers published in research journals and popular scientific magazines. A central concept arising from the model is idea flow--how ideas are generated or discovered as a result of individual social explorations and propagated through social networks via individual engagement. Ideally, these social explorations lead to the best ideas in a diverse group, which are then adapted and developed through social engagement (or social learning) in smaller peer groups. …