In June, Sethuraman Panchanathan was confirmed by the U.S. Senate for a six-year term as director of the National Science Foundation (NSF), an independent agency of the U.S. government that supports fundamental research and education in all the non-medical fields of science and engineering
In order to take on the directorship, Panchanathan had to take leave from his positions as professor and Foundation Chair in Computing and Informatics at Arizona State University, where he also served as executive vice president and chief research and innovation officer.
Panchanathan is the author of more than 425 refereed and conference publications in human-centered multimedia computing, haptic user interfaces, computing technologies for individuals with disabilities, machine learning for multimedia applications, medical image processing, and media processor designs.
Following is an edited version of a conversation with Panchanathan in early August, when he had been at NSF for little more than a month.
The position of director of NSF historically has been dominated by physicists and engineers. You are the first computer scientist to hold this position. One exception might be Erich Bloch, who was NSF director in the 1980s, though he was more of an engineer than a computer scientist. What does your being chosen as NSF director say about the position of computer science in the landscape of research today?
It's a personal joy to serve as the first NSF director from the field of computer science and engineering, though you rightly mentioned Erich Bloch, who did unbelievable work in bringing computing and engineering thinking into the foundation.
Computing and its advances are an integral part of how we work, how we live, everything. The Covid situation has probably brought to light even more intensely how humans and machines working together in synergistic and symbiotic ways often leads to tremendous breakthroughs.
People in other disciplines who have acquired computational expertise and the computing spirit are becoming even more innovative, and their creativity is being expanded. This is a very exciting time to be at the helm of the NSF, to be able to help advance computer science as a discipline, but also as an enabler of all the other disciplines, to come together in solving problems to uplift humanity and society and the economy.
At the NSF, computer science is primarily supported by the Computer Information Science and Engineering (CISE) directorate. In the 2021 budget request from the Trump administration, the NSF is slated for a decrease of 7%, though that decrease might not happen when Congress actually appropriates the NSF budget. The request includes a 7% increase for the CISE directorate. Can you talk about the priorities behind the fact that CISE did well in the request?
Budgets are a complex process I won't go into, but I will talk about the priorities, which is what you alluded to. It's always gratifying to see that each year NSF does get invested in more than the previous year.
Computing supports the advancement of other disciplines, so this is not an investment for CISE only, but an investment in CISE for the uplifting of all the disciplines. Whether it be big data or machine learning or AI or human-machine symbiosis, what the investment enables is tremendous partnerships, with industry, across the NSF directorates, with other agencies, and across the globe. The outcomes far exceed what we could do by looking only at individual, discipline-oriented investments.
I am confident that the benefits to humanity, to society, to the economy that are demonstrated by this investment will bring more desire to invest.
A bipartisan bill being considered in the House and the Senate would quadruple the NSF budget, add a big technology arm, and change the name to the National Science and Technology Foundation. What do you see as the advantages and disadvantages of such a transformation? How can you best protect the basic research arm from becoming merely the servant of the technology arm?
You are referring to the Endless Frontier Act. I cannot comment on the Act because it is pending legislation. However, I can talk about my view on the latter part of your question.
From my own research and experiences as a computer scientist, I don't believe in the premise that the basic research arm could become a servant to the technology arm. Future and emerging technologies require—in fact I would even go further and say they mandate—fundamental research. Most often fundamental problems get solved because you find that something is not working, therefore you have to reframe your fundamental assumptions and ask different basic questions. It is not a servant relationship, it's a master-master relationship.
Emerging technologies like AI and machine learning have caused concern about the impact on individuals, on privacy, on fairness in society. How does NSF integrate these concerns in the research it supports?
We have a directorate called Social and Behavioral Sciences (SBE). The collaboration between CISE and SBE has grown significantly over the past few years because of the kinds of problems you are talking about. How do we ensure that policy is not just in after-the-fact thinking, but that policy and technology are interwoven, so that as you are developing technology, you are thinking about the policy implications and thereby developing better technologies?
This goes back to the previous question. These concerns are challenging the fundamental assumptions that we are making in the design and development of technology, so that it is evolving in a much better way. There are many examples of this in projects that CISE and SBE are working on.
Can you tell me about one?
There is a project looking at implicit biases in AI. The people who are programming have implicit biases, and the systems they program become biased. How do we jog ourselves out of the implicit biases? The first thing is to recognize that there are implicit biases, before we can find a solution. SBE and CISE are working on ways to bring out understanding of implicit biases and to see how we can build better algorithms that are bias-agnostic.
There has been a lot of concern lately about national security and intellectual property. How does NSF balance that concern with the need to foster open international collaborations?
NSF's priority is to maintain a scientific ecosystem that is both secure and open. It is not one or the other. To do this effectively, we are reaching out to the U.S. science community and our partners abroad and emphasizing the value of talent from everywhere and of our partnerships. At the same time, we are intensely focused on ensuring that appropriate measures are in place to protect intellectual property, for a two-way symmetrical collaboration. We have a new office at NSF, created for addressing research integrity and security, headed by Rebecca Keiser.
We strongly believe in the American values of being transparent and open and having scientific and research integrity. We want to partner with those who share these values. International collaboration has been a priority for NSF's mission right from the beginning. Our commitment to international collaboration is solid, but we are very mindful of security issues. Therefore we are focused on partnering with those who share our values and are willing to work with us so that we share information, we are transparent and forthcoming, and we do this with scientific and research integrity.
Building a diverse science and technology workforce has been a challenge in the US for a long time now and is acute in computer science. What kinds of approaches have you seen that really work to promote diversity in STEM [science, technology, engineering, and mathematics]?
To me, diversity enriches innovation. The best science is shaped by different perspectives and experiences and is informed by different cultural and racial backgrounds. Therefore, for the US to maintain global leadership and cultivate a robust science and engineering enterprise, accessibility in STEM must be a priority.
In my own work with people with disabilities, I have seen how unbelievable talent can come out, empowering and enriching them. I have had students who are visually impaired who have designed solutions that would never have been designed otherwise, because they had first-hand knowledge of what the impediment was. This brought out their creative spirit and solved the problem in a way that benefits the entire population.
You were born in India in 1961. Was there something in your background—a book, a mentor, a course—that set you on the path to a life in science and engineering?
My father was the prime inspiration for me. He was a scientist, and his work was in atmospheric physics. His quest for scientific exploration, for discovery, for academic achievement, for solving real problems—I watched this in action every day. My mother was very focused on education and ensured that we valued education. So the combination of my mom and dad was the ideal incubator for me to express my calling.
You also get inspired by the people you encounter, teachers, mentors, friends, people you look up to. Then you build your career by learning from every opportunity. You can learn from everybody in every instance, if only you will give yourself the opportunity.
In the short one month that I have been at NSF, I have met people with unbelievable commitment and dedication. I am so gratified to have the opportunity to work with such a team. I know I will be learning from them, and I am hoping they will also learn from me. It is going to be a wonderful future at NSF, and hopefully through NSF for our country.
The Science article about you said you are an optimist. I see that.
I will take that accusation! I really wonder when people are not optimistic. We are the greatest country on Earth. We are fortunate to be here and fortunate to have this unbelievable nation. We should be optimists so we can build this nation into the future and develop humanity and global society as a whole. Why would you not be an optimist? The future is great.
Allyn Jackson is a journalist specializing in science and mathematics, who is based in Germany.