Social Physics, book review: Innovation, interaction and data-driven governance

MIT professor Alex Pentland outlines his theories on how good ideas arise and spread, and how sensors and big data analysis can deliver better social outcomes.

Social Physics: How Good Ideas Spread - The Lessons From a New Science • By Alex Pentland • Scribe • 300 pages • ISBN: 978-7-922247-55-1 • £14.99

There's a certain, perhaps peculiarly American, mythology that brilliant new ideas come from lone geniuses. In a way, this is a convenient theory; lone geniuses may be thin on the ground, but they are a zero-sum game. Once you have one, you have the secret recipe to success — and, better still, no one else can have it.

The reality, as Steven Johnson wrote in Where Good Ideas Come From and MIT professor Alex Pentland writes here in Social Physics: How Good Ideas Spread — The Lessons From a New Science, is that to really get ideas flowing you need informal settings in which people engage with each other spontaneously.

Johnson argued that what makes great ideas is the collision of half-formed ones and gave examples. Pentland, who helped create and direct MIT's Media Lab, saw that funding agencies, university administrators, CEOs and government leaders at the World Economic Forum failed to understand this principle.

Frustrated, he set out to create a scientific framework that includes social interactions. The ten-year effort produced many academic papers, a number of start-ups — and this book.

Pentland begins with one of the top American research groups: Bell labs, where internal studies showed that the stars were the ones who interacted most, who sought out the widest array of individuals to learn from, and who eagerly embraced new sources of new ideas. Moving on from there, he notes that influence is social: people are more likely to be motivated to change behaviours when they see their friends are doing so than when they are targeted individually through less personal means.

Reality mining

Pentland is one of the early promoters of big data — or 'reality mining', as he calls it — and part of his proposition here is that by widely deploying sensors and taking advantage of today's myriad data trails we can identify the mathematical underpinnings of how people interact.

Sensors and a 'digital nervous system' can, he writes, solve many of the challenges facing us: climate change, maintaining the energy supply, ensuring access to food and water. However, a sustainable future society requires a new 'nervous system'. Social physics will make that nervous system run correctly to support cities that, according to Pentland, are not markets or classes but idea factories.

Anyone who has seen the BBC series The Secret History of Our Street or read the associated book knows of the Victorian social researcher Charles Booth, who spent many years studying poverty in London by drawing a series of maps. That was the data-driven approach of the time, and it resulted in, among other things, the creation of the old-age pension when his maps made the link between poverty and old age.

Pentland believes that the ability to safely share data "will inevitably produce governance and policies that are more driven by data" — and he believes the result will be better social outcomes because we will have near-instant feedback on which policies are working and how well. This needs more research, though: can social physics can create political will?

Read more book reviews