Alexander Peysakhovich is technically a behavioral economist, but he bristles a bit at being defined that narrowly. He’s a scientist in Facebook’s artificial intelligence research lab, as well as a prolific scholar, having posted five papers in 2016 alone. He has a Ph.D. from Harvard University, where he won a teaching award, and has published articles in the New York Times, Wired, and several prestigious academic journals.
Despite these accomplishments, Peysakhovich says, “I’m most proud of the fact that I’ve managed to learn enough of lots of different fields so that I can work on problems that I’m interested in using those methods. I’ve co-authored with economists, game theorists, computer scientists, neuroscientists, psychologists, evolutionary biologists, and statisticians.”
Peysakhovich’s interdisciplinary work is driven by his deep interest in understanding decision-making—both human and machine—and by his desire to figure out how artificial intelligence can improve our decision-making processes. He builds tools that help people make better choices, and machines that can turn data into, as he puts it, “not just correlations but actual causal relationships.”
To that end, Peysakhovich looks at how to combine multiple types of data to go beyond just determining the average effect, but also to figure out, say, which segments of a population a particular policy works for, or doesn’t. One of his papers (co-authored with the Massachusetts Institute of Technology’s Dean Eckles) is about how to put together many randomized trials, such as different medical studies, to derive more nuance about what causes what. “This is something that’s important for building machines that can help us make smarter decisions,” he says.
In Facebook’s artificial intelligence research department—he says it “has a pretty broad mandate”—Peysakhovich works on developing technology that can generate good, data-based decisions. Before that, Peysakhovich worked on Facebook’s News Feed, improving “the algorithm that tries to help people find the most relevant content.”
His intellectual idol is Herbert Alexander Simon, a 20th-century Renaissance man who won both the Nobel Prize in economics and the computer-science Turing Award, besides writing 27 books, including a few about philosophy. “If I’m lucky, I can get a tenth of the way to [Simon’s] productivity,” Peysakhovich says. “He made so many path-breaking contributions to artificial intelligence, social science, and behavioral science. People say terms that he coined without realizing that he’s the one who coined them.”
Peysakhovich was born in Moscow and moved to Philadelphia with his family when he a small boy. “When things collapsed after the fall of the Soviet Union,” he says, “I think most people who could have left would have. This is especially true for many Jews.”
As a kid, he spent lots of time playing board games, then simulating, with dice rolls, what would happen in various situations in those games. “It should really be no surprise to anyone that I’m doing all this computational and statistical work now,” he says.
He went to New York University and fell in love with game theory while majoring in math and economics. His mentor there, Adam Brandenburger, urged him to pursue a science career. In 2009, the same year he graduated from NYU, he headed to Harvard to get a Ph.D., intending to work on “very theoretical problems.”
His Harvard advisors, though, had different ideas for him: Al Roth, a Nobelist, and Drew Fudenberg, a Guggenheim fellow, nurtured even his craziest notions but pushed him in an empirical direction. “Their support was really important, because most of my graduate school colleagues didn’t take my ideas very seriously,” he says, “Without my advisers, I think I definitely would have dropped out.”
These days, much of Peysakhovich’s work focuses on finding ways for scientists to use data to generate theory from hard fact. “It’s really important to be critical of things that are held as definitely true, by yourself and by others, because most of the time they’re not,” he says. “There’s lots of theory in science that sounds reasonable but is completely false.”
Making good decisions, as Peysakhovich sees it, is all about “building more-accurate representations of the world.” If we imbue machines with this ability, he believes, it would empower us humans to make smarter decisions, especially toward the betterment of society.
But how, exactly, can we build the type of technology that helps us make better decisions? That’s what he’s closing in on. “Recent advances in artificial intelligence,” he says, “are really giving us so many new tools to try to tackle this question.”
“The rapid explosion of computational power, data, and knowledge,” Peysakhovich continues, “is creating all sorts of new fields and opportunities that we’ve never seen before. The field of artificial intelligence right now feels like a huge open frontier to explore, and I’m lucky that I come to work every day at a place where I can be around the people that are exploring all sorts of different parts of that frontier.
Explore the complete list of this year’s 30 top thinkers under 30 here.
