Where Many Different Networks Come Together: Jennifer Tour Chayes

What do phase transitions, online advertising rates, human genetics, and children’s use of Facebook have in common? They’re all about networks, and standing at the center is Jennifer Tour Chayes, managing director and distinguished scientist of Microsoft Research New England, located in the Microsoft New England Research & Development Center (or NERD) in Cambridge, Mass. Each of these areas can be better understood by learning about the network of connections and decisions involved: how people interact and make buying decisions online, how the gene regulatory network in a cell breaks down and starts fostering the growth of a tumor, how the economic incentives in a healthcare system influence the decisions that doctors and patients make about treatment.

“What’s interesting is how widely applicable some of the techniques we come up with are,” says Jennifer. For example, to understand a complicated network whose structure isn’t fully known — like the Worldwide Web — researchers at the Cambridge lab developed algorithms that would help them reconstruct a network’s overall structure based on limited information available about individual users or sites. It turned out that those algorithms were also useful in helping to understand the gene regulatory system and how it directs a cell to produce more of one kind of protein or break down another, which could ultimately help researchers develop more precisely targeted drugs to fight cancer or genetic disorders. “That’s the thing that’s so wonderful about basic research; you never know where it’s going to take you,” says Jennifer.

Research has taken her a lot of places, from mathematics and physics to the intersection of computing and economics. Jennifer began her career in mathematical physics, earning a Ph.D. at Princeton, doing postdoctoral work at Harvard and Cornell, and teaching at the University of California, Los Angeles. “I always believed in doing multidisciplinary work, but I wasn’t at all focused on computer science,” she says. But as it happened, she went to graduate school with Nathan Myhrvold, and ran into him again later when she was teaching mathematics at UCLA and doing research on phase transitions and other computationally difficult problems. He lobbied hard for her to leave academe, come to Microsoft and start a theory group.

“When I came for my first interview, I kept saying to people, I don’t understand why you want me,” says Jennifer. “And Nathan said don’t worry about it, and later Bill Gates said don’t worry about it, because they knew much more than I did how deeply theoretical the questions are that Microsoft has to answer.”

So she joined the Microsoft Research team in Redmond and brought in the first theoretical computer scientists and physicists. “There’s a huge theoretical component in everything Microsoft does,” she says. There’s basic research that can lead to the development of previously unimagined products. Or investigation into current systems that helps make them better — for example, cryptography, an essentially mathematical domain that’s absolutely vital to the creation of secure, robust software products.

After a few years in Redmond Jennifer began to realize that while the greatest computer science in the world was being done at Microsoft, there was a lot of research opening up at the boundaries of technology and the social sciences — and that to really bring those worlds together she needed to be in a location where there was a lot of great social science research already happening. So about three and a half years ago she opened the Cambridge lab, building connections with the Microsoft team and researchers at Harvard and MIT. The work being done in Cambridge reflects those connections; for example, an empirical economics program led by Susan Athey of Harvard, looking at issues such as the effect of news aggregator sites on local news sites (contrary to popular belief, it turns out they bring more readers to the local sites), or work by danah boyd on how young people use social media — and how many parents will help them lie to get around age restrictions on sites like Facebook. Jennifer recently announced the addition of three new researchers to danah’s social-media research team, which will result in even more cutting-edge interdisciplinary discoveries.

The lab has a vibrant energy, says Jennifer, with lots of researchers in their 20s and 30s, with permanent researchers and short-term visiting faculty and postdocs, and with a constant buzz of different ideas coming together. “It’s the personality of the lab,” says Jennifer. “We work off and with each other in such an effective way. We have weekly meetings in which we all get together and discuss the work, and it’s often very surprising — we’re using different language but looking at the same problems.” Research runs from the highly theoretical to the immediately relevant or even newsworthy, such as mapping out the Cairo protests as they played out on Twitter. The brilliance and background draws on the best traits of a university, but the Microsoft structure enables new investigations to start and progress rapidly, without the delay of bureaucratic approval processes.

Jennifer points to a lot of new work under way: computational biology that will draw on the Cancer Genome Atlas and high-powered computing to identify genetic variations that could ultimately point the way to new drug treatments for cancer; cryptography for online networks that could help protect people’s data even when their passwords have been compromised; pricing and scheduling energy supplies to make maximum use of the cleanest resources to power data centers. Basic research is under way that could lead in any number of unexpected directions, which is the entire point of starting at the theoretical end and seeing where you end up.

“I’m really lucky to be in this lab of incredibly dynamic people who are so passionate about their work and about having an impact on society,” says Jennifer.

“It’s really just amazing to be around them. It’s essentially impossible to be a slacker here.” Of course I highly doubt there’s much chance of that for Jennifer wherever she is.