Ali Jadbabaie, the JR East Professor of Engineering in the Department of Civil and Environmental Engineering, associate director of the Institute for Data, Systems, and Society (IDSS), and director of the Sociotechnical Systems Research Center (SSRC), explores research topics as varied as his titles — using the foundation of systems theory to examine everything from the dynamics of sociopolitical change to the impact of fake news. His approach to his work is a holistic one, and represents a relatively new way of thinking about systems engineering: one that sees it as applicable to many of society’s most important, complex questions, and that is robust enough to include the human element in the equation.
Ali is a control theorist by training. He received his undergraduate degree from Sharif University of Technology in Tehran, Iran (a school modeled on MIT), and moved to the United States to get his master’s degree in electrical and computer engineering at the University of New Mexico. He went on to get his doctorate in control and dynamical systems from the California Institute of Technology, and took a faculty position at University of Pennsylvania in 2002 (following a stint in between as a postdoc at Yale). Continuing on the path he chose while a postdoc, Ali began his career at Penn focused on problems of formation control and swarming – that is, how you can control a large group of spatially connected agents to complete a task, such as forming a specific configuration, as in a school of fish or flock of birds. However, it was while he was at Penn’s General Robotics, Automation, Sensing, and Perception (GRASP) Lab in 2008, working mostly on getting groups of robots to move en masse, that he found himself increasingly interested in how collective behavior emerges in networks of human agents that are interacting to form collective opinions and actions. This led Ali to questions with broad implications for the kinds of problems systems theory might be used to solve: “What can you learn about the behavior of the overall system from the structure of the network and from individual dynamics? How do people make decisions in groups, how do they update their opinions, and how do they make decisions by combining their private observations and opinions, and the actions of others around them?”
With this shift in perspective, the next logical step was for Ali to incorporate social sciences into his engineering work — specifically the quantitative aspects of disciplines like economics or sociology, where exciting new research opportunities had been (and continue to be) opening up, thanks to the massive proliferation of data and major advances in data science.
Now studying collective autonomy alongside social networks and related phenomena like viral marketing, Ali found himself “in the right place at the right time” when, in 2014, he arrived at LIDS for a sabbatical. “That year [there was] discussion around creating a new entity that combined information and decision systems, statistics, and social sciences” in a new and intriguing way — a way that mirrored the direction he had been taking his research. This entity would become IDSS, which nine months into his sabbatical, Ali was asked to stay on and help launch, an opportunity that he enthusiastically accepted.
Ali has officially been at MIT since 2016, and thanks to strong connections to LIDS faculty throughout his career, he has found MIT a natural home. The LIDS connection is particularly relevant as LIDS is a major research lab within IDSS, making its faculty much of IDSS’s core, as well as key collaborators in Ali’s work. Having LIDS faculty in the same space as statisticians and social scientists has allowed Ali’s — and many other IDSS faculty members’ — research to be truly and more easily interdisciplinary, leading to more openness and innovation.
Describing some of the many projects on which he’s currently working, Ali says, “We have a large, multi-investigator, multi-disciplinary project that is in its final year on the evolution of social norms and the dynamics of sociopolitical change.” This project includes him, several LIDS faculty, and faculty from economics and political science. “The purpose of it is to combine the mathematics of social interaction and interacting systems and [things like] collective behavior with sociopolitical phenomena like uprisings, the Arab Spring, and political unrest and so forth.” One method of gathering data on social interaction that the team has used, for example, is studying call records from Yemen. These records are used “to establish behavior of social networks” to see if and how communication patterns change during a crisis; researchers can examine who talks to whom, how frequently, and for how long, and then that information is used to make inferences about that country and what life there is like by looking at a piece of the digital footprint.
Other recent and ongoing projects in which Ali is involved include topics such as decentralized and online optimization, research on the fundamentals of learning for DARPA, and nonlinear and non-complex optimizations. He considers himself lucky to have been part of so many different research projects, and calls it “an exciting time” to be at MIT. With so much work to do and so many concurrent projects, he tries to unwind by playing classical guitar, a hobby he hasn’t lately had as much time for, and by watching movies. His other hobby is following politics and the news, and as a self-described “political buff,” his newest project, which is in its beginning stage, matches up perfectly with this passion.
Along with multiple LIDS and MIT colleagues, Ali aims to study and “understand issues related to group polarization and information, and the spread of fake news.” The researchers are hoping to find and interpret hard evidence of the impact of fake news and other problems related to the fragmentation of online social networks. Ali points out that when they first started to form, online social networks were meant to democratize information and thus help the world become more equal. “The world would become flat, [we would become] closer to each other.” Instead, as has become obvious in the last couple of years, these networks “allow people to create these bubbles around themselves of like-minded individuals.” For instance, he explains, although people may involve many others in a particular discussion, their desire for and focus on social media “likes” colors the conversation significantly, leading to less nuance and more polarization.
Measuring in a “rigorous, scientific way” how this affects our social and political discourse could prove crucial to how we move forward in our interactions with each other through our online social networks. “The first step,” Ali says, “is to understand and measure the phenomenon [of fake news], to see, if any, what type of impact it could have had. The next step is to investigate what policy prescriptions we might be able to provide, what changes need to be made in our social discourse, in how we spread information, what incentives one can provide to rectify some of the issues that have been brought to bear by what we’ve seen in the past few years.”
This is exactly the kind of important, impactful, and complex societal problem IDSS was created to solve. Powered by the fundamental methodologies explored at LIDS, Ali and his collaborators are well-positioned to lead the way.