Quick overview: Sandy Pentland subscribes to the kind of ecological view that a lot of my recent sources have espoused- an emphasis on the relations between objects, instead of on the objects themselves. He argues for a “computational theory of behavior”, using Big Data and a system of collection/observation that he calls “reality mining”: the point is to grow the fullest, richest models of social behavior possible in order to capture all of the details, the micro-patterns, that traditional methods would overlook in favor of averages and stereotypes, and to capture them in “real life” as it is lived, instead of in a sanitized environment. Causation can be worked out through “quantitative, time-synchronized predictions” to make other explanations implausible, and then we might seek supporting quantitative lab experiments.

He has also demonstrated, as my previous installment notes, how he has applied social network incentives (as opposed to traditional, individualistic market incentives) to construct and improve complex organizational behavior often with a relatively light nudge.

Some more specific applications of this information can be found in the previous installment of notes.

This is part four (of presumably five?) of Pentland’s Social Physics. I’m roughly following the sections that the book lays out, in order.

  • Part 0: Some immediate personal/social context, what I’m thinking about before reading.
  • Part 1: A broad summary of the book, the methods involved, etc.
  • Part 2: Heavier details on the foundations of Social Physics (Social Learning, Idea Flow, Engagement)
  • Part 3: On Social Network-Incentivized Organizations
  • This part will be about “Data-Driven Cities”
  • Part 5 will be about Policy


 …We need to radically rethink our approaches. Rather than static systems that are separated by function- water, food, waste, transport, education, energy, and so on- we must consider them as dynamic and holistic systems. We need networked, self-regulating systems that are driven by the needs and preferences of citizens instead of ones focused only on access and distribution.

In other words, we need mobile sensing to create a “nervous system for cities, enabling them to become more healthy, safe, and efficient.”

This is obvious and I won’t dwell on it.

“Many of the sensing and control elements required to build a digital nervous system are already in place. What is missing, though, are two critical items: the first is social physics, specifically dynamic models of demand and reaction that will make the system function correctly [this chapter], and the second is a New Deal on Data, an architecture and legal policy that guarantees privacy,stability, and efficient government [next chapter].”

Already, this is a very different approach than the centralizing, “market-oriented” [not an oxymoron, folks] philosophy that Against the Smart City railed against.



Pentland’s spin-off company, “Sense Networks” allows for the movement and purchasing behaviors for tens of millions of people to be analyzed in real time. He introduces the languages of “tribes”, which here refers to distinct subgroups that tend to behave along similar lifestyle patterns (travel, taste, etc.). “These choices place them into a behavior demographic, becuase the behaviors of the tribe members selectively reveal their underlying preferences.” In addition, because they spend time together, social learning spreads through tribal lines and drives behavioral norms within the tribe. Tribe members may not necessarily know each other- that’s not the point. They have similar food habits, financial habits, entertainment preferences, political beliefs, and will tend towards similar health and career trajectories.

Sense Networks not only digs into preferences and behavior likelihood, but also the “rhythm of their daily habits”, which at the macro scale becomes the rhythm of the community or the city at large. Most of people’s daily patterns are the “workday”, the “weekend”, and the far less structured “exploring” (eg. shopping trip, special outing). These habits in time combined with the behavior demographics of the tribe allow for information to manage a city far more intelligently through forecasting and proaction.

Eg. Traffic flows, likelihood of car accidents based on other car data, bus routes. We can use transportation networks to increase productivity and creative output of cities by using the same kinds of analysis that earlier chapters have established. [Note: Tony Hsieh of Zappos had a good talk at the Long Now Seminars recently on this, and how he’s using this kind of social science to revitalize the area of Las Vegas that his company influences.] There are also public health examples Pentland suggests (sick people behave differently).

“Current city systems designs typically rely on financial incentives.” There are drawbacks to that approach- eg. tragedies of the commons; incidentally empowering the rich disproportionately.

Pentland identified three types of “social network interventions”:



Today we have a digital nervous system of sensors and communication already in place, ready to transform our cities into data-driven, dynamic, responsive organisms. Great leaps in health care, transportation, energy, and safety are all possible. In the Data for Development project [he will get into later], we will see that with only low-resolution, anonymous, and aggregate data, researchers were easily able to find transportation improvements of more than 10 percent [sic], health improvements of more than 20 percent, and make important contributions to the problem of reducing ethnic violence. The main barriers to achieving these goals are privacy concerns and the fact that we don’t yet have any consensus around the trade-offs between personal and social values.

“Urban areas use resources more efficiently and produce more patents and inventions with fewer roads and services per capita than rural areas.”

“[Pentland et al.] developed a mathematical model for how social ties drive idea flow within cities based on the number of people within face-to-face meeting distance. As we described in Nature Communications, this model gave us a simple, bottom-up, robust model that quantitatively predicts GDP and creative output.”

“Social tie density and idea flow offer simple, generative links among human interaction patterns and mobility patterns and the characteristics of urban economies without the need to appeal to hierarchy, specialization, or similar social constructs.”

Pentland focuses on urban exploration: people who live in the same place for a long time don’t stop “exploring” even after knowing the area well (here, we’re talking about purchasing from stores outside of their nearby, heavily favored shopping destinations.) Curiosity is a driver in decision-making, not just economic or even necessarily taste considerations. As cities grow wealthier, the ecology of opportunities available become more complex and exploratory behavior increases as a proportion. Wealthier people tend to have a vastly greater proportion of exploratory shopping, increasing the diversity of people and things they interact with. Families who used to be wealthy but are downwardly mobile had lower levels of exploration compared to familiar destinations. (That is, it isn’t that the wealthy have a tradition of greater exploration). Pentland later demonstrates that this effect can be used “to accurately map the wealth of neighborhoods, since their patterns of explorations are reliable signals of their disposable income.”

Pointed: “Unlike what might be expected if the “strength of weak ties” model were true (i.e., the idea that having more social ties results in more wealth), exploration does not seem to translate into greater wealth in the short term. Instead, it is the other way around: Wealth allows people to invest more in exploration. […] the fact that cities with more exploration tend to have greater growth in their wealth suggests that harvesting new experiences and meeting new people does pay off, but it just takes a while. Exploration benefits the city as a whole, and that increase in idea flow within the city is bound to help both individuals and their families, even if only indirectly.”



“Designing Better Cities”