Big Data and Social Physics: Reflections on Society, Ethics, and Algorithms

As Internet ubiquity has grown, so has data collection. Our ability to sort, understand, and derive insights from data raises questions and creates opportunities for public relations, and also raises more profound societal questions around the quantification of human behavior, privacy, and ethics. Upon completing a course on Big Data and social analytics at MIT, Kevin McCann, a NATIONAL partner based in Halifax, prepared this reflection. (Also available through his LinkedIn page)


Near the end of 2016 I completed an 8 week course from MIT on big data and social analytics. It was fascinating, challenging, timely, and thought-provoking. Here are some reflections.

Societal Shift

Saying everything is disrupted is the conceit of our time, but it is a truth that we are experiencing a societal shift. As we create ever more wonderful and fascinating devices to distract us, drive our cars, modulate our temperatures, track our fitness levels, we will be creating more and more data. Data that we can use to tell us more about ourselves and our world. Alec Ross describes the implications succinctly in his book “The Industries of the Future”:

  • Who owns big data is as important a question as who owned the land during the agricultural age and who owned the factory during the industrial age. Data is the raw material of the information age.

The inference is clear and stark: How well data is used and understood will have a great influence on organizational and enterprise success. Thus the rightful fixation on data-driven PR in marketing and public relations.

Social Physics Coming Into Its Own

An older term than you might think, social physics is a way of understanding human behaviour based on the analysis of Big Data. Whether the term “social physics” sticks or not, it represents the collision of social sciences with the data-driven disciplines and the logic of engineering, physics, and computer science. This is very relevant for public relations, marketing, and communicating in general, because looking at behaviour and what people actually do will always be more truthful than what people actually say. MIT Professor Sandy Pentland puts social physics in context and why it is relevant to our present day:

What’s happened now that we have the big data, this rich data– things from social media, things for cell phones, from credit cards– all those other sorts of things, is that we finally have enough data to really answer, in a quantitative predictive way, the questions of social science. How do ideas move from person to person? How do you get tipping points? How does all this stuff work? That’s what social physics is. It’s a mathematical, predictive science of how ideas move through society. Source:

Social media threw gasoline on specific areas of analytical study, moving them from obscure and specialized to commoditized and highly relevant. If we could look at a group via who they call, who they spend time with, where they spend money and where they go, and who their social media friends are, we can get a richer picture of what inspires and resonates with an individual beyond what they say on the internet. One of the by-products of the internet has been to truly expose and normalize the data of our species, and we are just at the beginning of understanding this.

Privacy and Ethics in a Data-Driven, Algorithmic World

Advances in technology force an ethical pause and more often than not, a reckoning. We need look no further than our Netflix or Amazon recommendation queue to see the role that algorithms are playing in our lives today. No human devised these lists, other than crafting the logic that fills them with things you will spend your time and money on. Consider:

  • MIT Media Lab Professor Kevin Slavin points out that the stock market crash of 2010, where 9 percent of the entire U.S. stock market disappeared in five minutes, was algorithmically driven. “Nobody ordered it, nobody asked for it, and nobody could control it. The algorithms took over, and humanity was locked out”.
  • The fake news controversy (even describing it this way seems to weaken how important this was in 2016) is the result of algorithms doing their job, and these algorithms only exist because of the massive amounts of data exposed in Facebook and other networks that capture online behaviour. Influencing algorithms is practically the new media relations, and understanding this reality is an essential part of modern marketing. And unfortunately modern politics.
  • Zettabytes of data and the ability to understand it enabled things like self driving cars and autonomous vehicles. Now we have to think about the ethical decisions that come with these technologies: Should the artificial intelligence of a vehicle protect passengers or pedestrians, if faced with the decision? As with any advance in technology, new ethical considerations come into play, and this is truly more pronounced in our age.
  • Marketers have been conducting experiments on human subjects for as long as they have been able to do so—from taste tests to email A/B testing to social media custom audience advertising and retargeting. As we continue to add digital data insights to who we reach, how we reach them, and what data we use to market to them, privacy quandaries will rise and become normalized.
  • In the hands of a data scientist with the right data sets and algorithms in place, only four geospatial data points are needed to identify 90% of individuals from a seemingly anonymous data set. Privacy is an illusion a digitally networked world, even when we think our data points do not have identifiers.
  • Facebook Safety Check is an algorithmic-based notification service that detects when a disastrous crisis is happening. In the June 2016 Orlando nightclub shooting, Safety Check reflected that the crisis was happening 11 minutes before police officially issued a release. This decision was algorithmic and not human-based. Algorithms in our daily lives  have changed crisis communications forever.

The application of ethics to a big data and algorithmic world are still undiscovered lands where we are just finding our way. Communicators must help organizations navigate in ways that are truly trustworthy, lest they fall victim to cynicism and apathy among their publics. This charge is part of the mandate of public relations today.

Code Thinkers: Skill Sets That Embrace Technology

Taking this course took me back to the fun and challenge of thinking in code. The Python exercises and deliverables were both humbling and empowering. And thinking as a communicator and a developer was perfectly embodied with Jupyter Notebooks—a simple but ingenious way to combine writing and programming in one intuitive place (a tool I wish was in place for every other programming source I’ve ever taken).

Beyond my own little reverie, this is relevant as it applies to the talent needs of the future. Observers and predictors are seeking specialists, but beyond that organizations and leaders are seeking those who can think in multiple disciplines and understand the implications of data, the realities of an algorithmic world, and apply the logic of programmatic thinking to the art and creativity of social sciences. Beyond lucrative specializations that the market simply will pay for, a mish-mash of thinkers that understand technology, data, and humans in a changing context will continue to be in demand.

As Tim O’Reilly put it,

Instead of the binary question of which jobs will be eliminated…it is what tasks that are being automated, and automation doesn’t simply destroy jobs. It changes them.

This is where we are, and where we are headed, and many prepared to handle this future will be able to apply technical and social-sciences thinking to new contexts. This is especially true for communications and public relations, and is underpinning jobs and the need for talent across industries:

The Top Skills of 2016 on LinkedIn Canada

Source: LinkedIn Unveils The Top Skills That Can Get You Hired In 2017.

Just the Beginning

When the web and personal computing started to become accessible, vocabulary that is common today was new and inaccessible then. Even small words, acronyms, and phrases like HTML, script, pixel, database, SQL, app, api, social graph, have become common concepts for anyone thinking in a communications context, or their relevancy is at least understood and appreciated.

This normalization of terms will become true for data analysis and network analytics. Terms like eigenvector, centrality, betweenness, aren’t new but their meaning is becoming more relevant, as their specialized fields are more applicable in an algorithmic, data-fueled world. Those outside of specialized fields won’t actually use these words, but many will internalize the concepts and be able to apply to their jobs and context. This I believe, forgiving my own bias, will be particularly true for anyone charged with a communications role. We are just at the beginning of the applying true data analytics and social physics to communicating in the modern world.

Links and Sources

Some of the links from above are captured here, or added to below with articles that caught my eye in the last weeks of 2016.


Kevin McCann is a Partner with National Public Relations, where he focuses on organizational strategy, technology, and advocacy.