Futurithmic is moving to nokia.com! Read more here.

Social data and sentient AI

Share on facebook
Share on twitter
Share on linkedin
Share on reddit
Share on email

Within just a few years “people data” will be 90% of the world’s collective data.

Jeff Nick, CTO of EMC, personal communication

Our society has reached an unprecedented level of global connectivity.

We live in a world peppered with countless digital interfaces, devices and platforms. This, in turn, makes our online and offline space a behavioral-data-goldmine. Further development of cutting-edge technologies – from 5G to deep learning and ubiquitous computing – means the ability to gather and decipher behavioral patterns and anomalies in real-time and real-space will be boundless.

Though the term ‘social physics’ was coined two centuries ago, it represents a relatively new scientific field enabled by advancements in Big Data analytics. Through social physics, we can understand and predict preferences, attitudes and behavioral patterns of individuals and groups in a much more accurate way than traditional marketing segmentation like gender, race, education level or profession.

“We can write equations about it. We can predict it. We can influence it.  We can engineer it,” says Alex (Sandy) Pentland – an MIT professor, data scientist, author and social physics pioneer. Pentland’s work provides behavioral analysis and context and enables organizations and governments to become more cooperative, productive, and creative.

Pentland’s research is used to advance collaborative problem solving, mobilization in urban areas and even correcting communication signaling and cultural biases.

Some examples include observing workplace interactions. Pentland’s data analysis indicated that women and men interrupt each other just the same. It might have been a sign of power equality if it wasn’t for the fact that the nature of the female-to-male interruption tends to be positive/affirmative, while male-to-female interruption has more of a dismissive nature.

Another example looks at job mobility success factors. For instance, social connectivity has a much higher value over actual STEM skills. And among the strongest success factors for startups is having cultural diversity.

We could cure global warming tomorrow if we could all just decide what to do and get about doing it. And the impossibility of that makes us not even consider that as an option.

Alex (Sandy) Pentland, “Why Social Physics?”

Data for Good

When everything we do can be collated and quantified, we want to aspire for meaningful application of data.

We are already witnessing the deflation of the “going viral” hot air balloon, partially due to the inflation of fake online engagement and following generated by click farms or bots, but also due to the realization that beyond the immediate attention-grabbing, achieving consumer “reach”, “influence” and “engagement” metrics do not guarantee lasting or qualitative value.

Pentland and his students’ research looks for ways in which ideas are diffused and transformed to new behaviors, hooking short-term and screen-deep digital engagement to intrinsic motivation and converging it to actionable commitment.

Having a socially-positive application of data analysis is essential for a data-driven society to thrive, especially since Big Data, civic data and biometric data gathering are not only available to lab researchers but to a wide range of governments and private organizations, as well as commercial entities. This represents a new value system where behavioral data is used as currency, shifting financial models and entire markets and societies.

Such ability to shape social behaviors creates a new power paradigm and raises ethical concerns. Beyond issues related to data privacy, it manipulates notions of agency, influence and

Data is a powerful tool in shaping social behaviors, especially as we develop and apply more sophisticated data-gathering systems. With the help of neuroscientists and behavioral psychologists, tech platforms and devices are embedding tools that can capture and decipher subconscious cognitive and physiological signals (such as eye-tracking and facial-coding). This increases our capabilities to achieve tools (or weapons) for mass behavioral influence.

Such ability to shape social behaviors creates a new power paradigm and raises ethical concerns. Beyond issues related to data privacy, it manipulates notions of agency, influence and governance. Social scientists like Shoshana Zuboff have expressed concerns about the rise of the “surveillance capitalism” model.

Economists such as Kate Raworth are looking into a holistic approach for human progress such as the “doughnut economics” model, designed around economic thrive versus growth.

One of Kate Raworth’s seven ways to think like a 21st century economist is to change the goal.

We require better legislation, regulations, frameworks and, ultimately, a global consensus that ensures that the application of data analytics would be primarily applied towards social progress.

As the founder of MIT’s Trust:Data Consortium and the co-lead of the World Economic Forum’s Big Data and Personal Data initiatives, Pentland is promoting and advocating for data distribution systems that ensure privacy protection and accountability. He also initiated The New Deal on Data that defends the right for people to own and have control over their data.

Humanized & Sentient AI

Muslim friends enjoying smartphones

Pentland’s goal is to achieve a humanized application of AI through social physics. This involves rebuilding meaningful social action, interaction and networks, and to ultimately reduce the current social divide that is reinforced by AI-enabled echo chambers.

He advocates for a notion of a digital society that works for humans. One where the outcome is safe, efficient, scalable and inclusive. His vision and research explore modules of sustainable digital ecology. His work and his students’ work at the MIT Connection Science and Human Dynamics lab focuses on improving policymaking, healthcare and crisis management; improving decision-making and collaborative problem solving; and elevating community-based interaction and idea-sharing to sprout diversity, innovation and access to opportunity.

As we approach the realization of a sentient AI, NOT establishing a socio-positive application of data analytics means unleashing a tremendous computational force with no social responsibility and guidance. This is an irresponsible act at best, and one with a devastating outcome to humanity at worst.

The ways in which we will manage the tension between the seductive potential of applying social physics for personal gain, versus serving the social good, will be a determining factor on how we will shape the future of technology and our society.

Want to see more? Watch the second episode of Futurithmic “Will Big Data Affect Government Policies?” here.


Share on facebook
Share on twitter
Share on linkedin
Share on reddit
Share on email

Subscribe to the newsletter: