Michael Sipser reflects on time as School of Science dean
Sipser hopes the MIT community ‘continues to ask the really big questions in science and makes progress in getting the answers’
Michael Sipser intends to step down as dean of the School of Science on June 30, “assuming a suitable successor is found by then,” according to Provost Martin Schmidt PhD ’88 in an email to the MIT community. Sipser has served as dean since 2014, after several months as interim dean and 10 years as mathematics department head.
The Tech sat down with Sipser to discuss his time as dean and his thoughts on what is ahead. This interview has been lightly edited for length and clarity.
The Tech: Why are you stepping down as dean at this time?
Michael Sipser: This is my sixteenth year doing administration. The short answer is that there are other things I want to do. I've loved being dean, but I want to get back to research, and I want to devote more time to teaching. I love the students here.
TT: Which accomplishments as dean are you most proud of?
Sipser: Supporting the activities of the community of people in the School of Science has been, for me, the most rewarding and most important thing I’ve done.
The thing that comes to my mind is the Aging Brain Initiative. Neurological diseases of people when they get older affect a very large fraction of everyone worldwide. It’s such an important problem, and I think MIT could do more, so I put some of my own resources and time as dean into helping that grow.
TT: I think that for a lot of students, what a dean does is sort of mysterious. Using the Aging Brain Initiative as an example, can you go into more detail on your role?
Sipser: Being a dean has many sides to it, so this is only going to be one piece.
First, I made my own sense of the importance of this problem clear to the people who were already working on it. When somebody’s work is recognized by higher levels in an organization, that gives them more confidence that they’re going to be supported.
I talked about that work when I spoke with other members of the administration, so people knew that I believed that this was going to be a priority for the school. I spoke to donors about raising money.
So it’s a combination of all of that: helping ideas to coalesce and become more visible.
TT: How would you describe the current research landscape of the School of Science?
Sipser: Science at MIT is very healthy. There’s a tremendous amount of exciting things going on, whether it’s detecting gravitational waves, finding exoplanets, or editing genes.
Broadly speaking, you can think of science as breaking down into curiosity-driven research and solutions-driven research. There’s some pressure from various funding sources for solutions-driven research, as opposed to just expanding knowledge as dictated by the tastes of the people we hire.
MIT has historically had a very good mix of those two approaches. It’s important that we maintain that and not shift too far toward the more applied side because fundamental research is what leads to new applications down the road and helps us understand our world.
TT: Do you think MIT is moving in the right direction with its growing emphasis on computing research, especially in artificial intelligence? How do you envision the School of Science fitting into that picture?
Sipser: Artificial intelligence has clearly been a game-changing technology over the last decade. MIT has to be invested in it because it’s affecting almost everything.
More and more of science [involves] accumulating vast amounts of data, and the question is how do you understand all that data? How do you use it to make predictions? One approach is to use machine learning. AI has proven that it can discern patterns. It’s only a tool, but it still helps you see things that you wouldn’t necessarily be able to see otherwise. I think we are going to be engaged with the College of Computing around those sorts of things.
There are other things besides artificial intelligence. Quantum computing, for example, has become a very popular field these days, and that touches the School of Science in other ways.
TT: You’ve been dean of the School of Science since 2014, and before that you were the mathematics department head since 2004. What drew you towards administration?
Sipser: I ask myself that sometimes. I guess the best answer is that first of all, I love science. And I also really like scientists.
Even though I had never imagined being in any of these roles before I started as math department head (I was very surprised when I got asked), I found that I was having some success, and so that led me to feel encouraged and to continue.
When the opportunity to become dean came up, I realized that the job is different because you’re somewhat removed from any particular community. Of course you have the broader community of your entire school, but that’s a more distant relationship.
With that recognition, but still an interest in broadening the kinds of scientific activities that I would be overseeing, it was something I wanted to try.
TT: So do you think you made the right decision to be dean?
Sipser: [Laughs.] It’s hard to say. You only get to live your life once. I think there’s much that I look back on with pleasure, and I feel some personal satisfaction, but I do—
I have no regrets at all about the time I was math department head. That was, for me personally, an unqualified success. Being dean was a bit of a challenge for me. It’s a very hard job. And I think if I have a regret around that, it’s that I had very little time left over to do anything else.
I continued to teach. [Sipser teaches 18.404 (Theory of Computation) each fall, which in its most recent offering had 227 students.] I think some other deans might teach as well, but I don’t think anyone teaches a large class. I like interacting with the students. I refused to give that up, even though it made my schedule crazy.
But time to do research was almost nonexistent. The fact that I was giving up years of my career where I could do essentially no research — that was a big sacrifice.
TT: Do you have specific post-deanship plans?
Sipser: I have research ideas that I want to explore of a mostly mathematical, theoretical computer science nature. But I have to get back into it. There’s a certain amount of rebooting to start to think about mathematics again.
When I have some time, I think I’d like to work with the department and also more broadly within the Institute on rethinking what the math GIRs should look like. I haven’t done the analysis yet, but I know there’s a need, for example, for students to know more probability and statistics. I’m not sure how that need balances with our current offering in 18.02, where we spend a good amount of time on vector calculus.
TT: What do you think are the biggest challenges that your successor will need to address?
Sipser: There are many challenges around funding science. Science is often expensive, and MIT itself is expensive. If you want to have a group of students in your lab, it’s a big challenge to raise the necessary funds. Federal funding is not keeping pace with the need.
TT: In what ways do you think the School of Science may look different in, say, 10 years?
Sipser: These are more aspirations rather than predictions. I hope that our efforts at increasing diversity succeed. It’s slow, but we’re making progress. I also hope that MIT continues to lead in science, and that our community continues to ask the really big questions in science and makes progress in getting the answers.