Three months later, Suresh Venkatasubramanian’s May 17 tweet (“It’s time for a transition…”) is still getting shared. Announcing a new post at the White House Office of Science and Technology Policy as well as his decision to join Brown CS and Brown’s Data Science Initiative (DSI), as of this writing it’s just shy of 1,000 likes.
Suresh, who starts this fall as professor, is the latest hire in the multi-year CS With Impact campaign, the most significant expansion in the department’s history. He describes himself as, among other things, a computational philosopher, saying that he stole the phrase from a undergraduate at one of his lectures: “Many people see computer scientists as builders, as engineers, but I think there’s a deeper intellectual perspective that many CS people share, which sees computation as a metaphor for how we think about the world. It’s a lens with limits, but it’s helpful in many ways, and I like to think about how it collides with a humanist lens, a societal lens. That’s how I hope to engage with people. I’m not going to fix your computer, but maybe I can get you to think about these things.”
After the briefest pause, Suresh adds a postscript that’s both charitable and cheeky. “And maybe I will fix your computer,” he says, “but I don’t want to.”
Venkatasubramanian traces his first interest in computers to age eleven: noodling around, trying to make games in low-level assembly code in an era when even getting hold of a computer was challenging. Heading to the Indian Institute of Technology Kanpur for his BTech, Suresh says that he had no idea what being a computer scientist meant. “Thanks to books I’d read,” he tells us, “I was interested in two things, both of which had to do with Turing: artificial intelligence and Turing tests.” He also started learning classical guitar, and says that he’s been a perpetual intermediate student ever since.
Later, at Stanford University, what Suresh calls in retrospect a completely random walk (“I was a piece of wood on the sea”) led him to algorithms and computational complexity: “Within a month, I decided I didn’t want to do AI. Complexity was fascinating to me – I’d read Gödel, Escher, Bach and I was intrigued by recursion, what you could do with formal description. This was where my interest in computational geometry and in some weird sense, data, started.”
Almost two decades of exploration in both the theoretical and applied aspects of data science (computational geometry, sublinear algorithms, clustering, kernel methods) followed. And then in 2013, a shift. Between-the-lines readers of the “slightly more light-hearted” bio at Suresh’s website might infer that reading science fiction could have led to his current interest in the social ramifications of automated decision making.
“It was definitely influential in the questions I was asking,” he says, citing Neal Stephenson, Cory Doctorow, and the Turingesque imagined libraries of Jorge Luis Borges. “I was going on sabbatical and people were encouraging me to take a new view of the field. My first instinct was to look into what the geometry of music tells you about its aesthetics, but I had just found Cory Doctorow’s ‘Human Readable’ and I was thinking about what it might mean when algorithms are embedded everywhere.”
Eight years ago, Suresh says, it seemed a little futuristic to be asking how we might know if our lives are or aren’t being controlled by an algorithm: “But very soon, we were there. I made a lucky guess, and I was right. One of the things I like about sci-fi is the worldbuilding: what happens if we pick one thing we take for granted and change it?”
Algorithmic fairness has been in the news almost daily as of late, making it part of even the layperson’s vocabulary. What can we expect next? Suresh sums it up succinctly.
“If you apply for a loan tomorrow,” he says, “you’re going to talk to a robot. Everything it can find about you, down to the random comments you made on Twitter about how you hate mango juice1, goes into a gigantic database, and then it decides if you get the loan. How do you feel about that? Imagine that you go to a different country, with a different culture, and the robot is still making decisions about you – that’s the world we’re in right now. Now replace ‘loan’ with everything: whether you get a job, whether you get surveilled by the police, whether you go to college, whether you go to jail. When you ask why, imagine the robot answering all your questions with the math said so, the algorithm said so. Would that be ok with you? That’s the question.”
Returning to Suresh’s tweet from May 21, we focus again on the theme of transition. Is his upcoming role as White House policymaker part of a personal shift from theory toward practice?
Absolutely not, Venkatasubramanian says: “Theory and implementation aren’t opposites. They’re part of an ecosystem, and you can’t separate them and say that they’re opposing forces. If you have a rigorous way of thinking but can’t deploy anything, you’re talking into a void. I want to use all the tools at my disposal, and policy tools make sense, but part of me is also thinking about theory. In an ecosystem, you’re balancing things that work together. Computer science has a bad habit of trying to binarize these questions, which is a waste of time. We all need each other.”
The other half of Suresh’s transition is his arrival at Brown, where Venkatasubramanian will co-found the forthcoming Center for Computing for the People. With gratitude to Brown CS Professor Seny Kamara, Professor and Chair Ugur Çetintemel, and Professor and DSI Director Sohini Ramachandran, Suresh says he’s looking forward to creating a vision and a space for what it might mean to reimagine computer science, attuning it to what people want and need instead of serving the wants and needs of whoever’s paying for it.
Of all the elements that might make up the new center, from teaching to research to mentoring, which does Suresh find most interesting?
“I don’t have that answer yet,” he says, “but what I will say, and this is Seny’s vision as well as mine, is that I want to make our value systems more explicit. With the success of our field has come a realization: how we frame our questions has a non-neutral impact on the world. If we can all make more informed judgements about the effects of how we frame problems, we can choose framings that are in line with our beliefs. We don’t all need to be committed to the same issues, but I want the Center to be a place where we can figure out the trade-offs between the different ways of framing a question. And to put my own cards on the table, I want to help people who are not being represented well by technology. That’s my stake in the ground, and I want the Center to do that as well.”
Forming connections across Brown and doing interdisciplinary work, Suresh says, is key. For students with more technical training, this means learning to appreciate the ways in which technology intersects with society, reaching out to colleagues across campus who might have a very different view of how technology might affect them. As with his work at the White House, this might lead to policy-making, but the conversations, the bringing of our own CS questions, and the listening to the concerns of others all come first.
“One of my soapboxes,” Venkatasubramanian says, “is that for a long time we’ve had metaphors for talking about human society from the humanities, physics, math, economics. But what we now have are computational metaphors: ‘The algorithm will do this, the AI will do that.’ Like any metaphors, they have advantages and limitations. I’d like to expose where these metaphors are effective but also show their limits. If I can bring a sensibility that language is coded in certain ways and that we’re not aware of the way that computational metaphors are creeping into the way we talk about the world, I think that will be helpful.”
Curiosity piqued by seeing the name of a well-known Brown CS alum, danah boyd, in Suresh’s list of publications, we ask him about techno-optimism.
“In a weird and twisted way,” he says, “I don’t think I’m a tech pessimist. I may have a realistic view of how people use technology, but I have an optimistic view of tech itself. And yet that optimism isn’t automatic: I was teaching a class on the ethics of computer science, and my young students were saying that social media is dominated by Facebook – they couldn’t imagine any way around it. I said, ‘Well, what about open source, which worked for the Internet, for email, and a lot of other things?’ To have that optimistic view of technology, you have to allow yourself to go beyond standard frames.”
There’s a lot of possibility for optimism, Suresh believes, but we have to move beyond our conceptions of what technology looks like right now and ask what a tech-enabled world should be. “Computer scientists,” he notes, “often talk about the need for the social sciences and the humanities, and they’re vital, but a degree of imagination is what’s most important.”
It’s exactly why we benefit, says Suresh, from what he describes as the more enlightened kind of science fiction: “The key thing is to imagine better.”
1 Suresh tells us that he likes mango juice, personally.
For more information, click the link that follows to contact Brown CS Communication Outreach Specialist Jesse C. Polhemus.