Information Sciences Institute

Jim Kurose, Distinguished University Professor in the College of Information and Computer Sciences at the University of Massachusetts Amherst, will be a speaker at the Symposium on the Future of Computing Research held by USC’s Information Sciences Institute on September 12-13, 2022. He gave us a preview of his thoughts on this matter.

The role of computers in society is changing and evolving right now. Knowing this, what are the topics that we should be researching to fit this shift?

We have moved noticeably in the last five years in particular from a discipline that has focused on theoretical foundations, and artifacts and systems at scale, to now also considering the use of these systems in society. Our field has gotten really good at knowing the kinds of things that we can build. And now it’s more about how we shape these things that we’re building to really work in our society, and to work with people at a high level. To me, that really is the future.

What is a societal issue that you foresee coming from computing research that’s happening now?

AI systems are playing a more and more important role in our world. In AI research, there are issues of fairness, accountability and transparency. We have AI systems that have been trained on certain sets of data, but there can be inherent biases in the operation of these systems. For example, we can build a system that’s really good at identifying faces, but it’s only good at identifying the kind of faces it was trained on. Facial recognition is a great example of where computing intersects with society. Civil society, technical societies and others are stepping forward and asking, “What kind of policies, what kinds of AI systems, do we want?” These are national and international concerns. For example, in 2019 the OECD released its “Principles on Artificial Intelligence” that are aimed at promoting AI that is innovative and trustworthy and that respects human rights and democratic values. I think the issue of the fair, accountable and transparent use of tools, and how you demonstrate and ensure that, is a tremendously important social and computing challenge.

Who should be in charge of regulating this?

Regulation almost always implies government involvement. In the case of facial recognition, our US government doesn’t yet have policies in place. A few years ago, we saw industry actually calling for the government to address regulatory issues in this area. I was at the Office of Science and Technology Policy and the NSF during that time and my sense was that government wasn’t quite prepared to take that up at that time. It’s my personal hope that this will be taken up in the future. One would also hope that technologists and civil society organizations would come together and have discussions that would contribute to such an activity.

How do you think the government should act to help foster innovative and meaningful computing research?

There’s been lots of pending legislation focusing on making very significant investments in both basic and use-inspired computer science research. The Endless Frontiers Act was a bipartisan bill coming out of the Senate in 2021; the House had the National Science Foundation for the Future Act. Both aimed to significantly increase the research budget of the NSF. The Endless Frontiers Act also called out 10 key technology areas of focus, many of which were computing related.  Significant parts of Senate and House bills were then included in the CHIPS and Science Act.  That Act also calls for—and this year, it was established—a new directorate in the National Science Foundation on Technology, Innovation and Partnerships, the TIP Directorate. The other directorates in NSF are science verticals: computer science, engineering, mathematical and physical sciences, biology, etc. This new directorate is more of a cross-cutting one.  It’s now received a significant amount of new funding in order to increase, across all of science and engineering, use-inspired research. The authorizations in the CHIPS and Science Act represent once-in-a-generation increases in investments in both basic and use-inspired research, in particular, in computer science.

In all these computer science fields, should only computer scientists be involved? And if not, who should be involved in computer science research? 

Computing research nowadays has a significant focus on what computing means to people and to society as a whole. That’s not something that we as computer scientists have typically been trained to do. We need to think about how to evolve the training of computer scientists, or perhaps better said “co-evolve” that training with other disciplines, the social sciences in particular.

Thinking about how humans use computing, that’s bringing together computer science with the psychological and the brain sciences, and other traditional social sciences, such as anthropology. The social sciences and computing – that’s, in my mind, one of the most exciting interfaces we have with other disciplines.

If we’re bringing other fields into computer science research, should we have new expectations? How should the criteria for evaluating the research change?

That’s not just a computer science question. For a long time, one important measure of research impact was having highly-cited papers a researcher published in prestigious journals in and conferences their field. I’ve certainly noticed that our students and young faculty increasingly care about the impact of their work on people. So I do hope that universities will start evaluating faculty more broadly than publications alone – because that’s just one way of many in which faculty can contribute. There are researchers who are building things and doing civic engagement experiments that are engaging people. They tend not to write as many papers, but that work is equally important. That sort of outreach, the human and the societal impacts, the economic impact of your work—these are other dimensions of impact that, historically, universities have not recognized and rewarded as much. I do think that universities are starting to broaden their notion of what it means to be successful and how impact is measured.

How do you convince people in other fields to become part of the team? 

There’s a move afoot, called X + CS, where x is a variable. It’s any discipline that you want to think about. For example, University of Illinois, Urbana-Champaign has a CS + X (they put the X second) with agriculture, economics, philosophy. They’re establishing a digital medical school there. It’s not that you go out and say, “Become a computer scientist and join us.” You say, “Hey, your field and my field, working together, can do these amazing things.”

There’s a type of research, known as Convergence Research, that is problem- and challenge-driven, and requires inherently inter-disciplinary solutions. Convergence research problems are often grand challenges, human-health-related, for example, where now we need to bring together multiple disciplines to solve that challenge. We’ve got to bring together many disciplines and players – medical schools, basic biology, computation, engineering to fabricate devices, and more. And we cannot solve such a grand-challenge problems unless we bring all these disciplines together. Convergence was a really important theme while I was at NSF, and I think it will continue to be an important thread in the future. Research challenges in computer science plus social/behavioral/economic sciences are often an instance of that.

What are some research areas that will have exciting developments in the future?

In almost any research area you can think of, somebody is looking at the applications of AI. An area that I find really exciting is the notion of fusing human computation with chip-based computation. People are putting silicon chips in neural pathways to help the unsighted be able to see better, or exoskeletons and smart prosthetics that help and enhance physical mobility. To me, that human augmentation, both physically and mentally, will be transformative.

Another important research area is quantum computing and quantum networking. Industry – Google, IBM, Microsoft and others – have started up quantum computation R&D efforts; in particular, being able to solve problems that were computationally not tractable before. It’s a fairly limited set of problems so far, but I do think that quantum computation is something that we haven’t experienced even close to the mainstream yet, and that’s absolutely going to be a growing area.

How will today’s discoveries change the world in 10 to 20 years? 

The future of work at the human technology frontier and how people interact with cyber will be vastly different. Many people will be working in their jobs with computation right alongside them. In some cases, it will be disruptive. It will create new jobs, but other jobs will go away. So as a society, we will have to adapt to that and train for that.

Computing has also shown recently how important it can be in areas such as drug discovery. We’re now able to design drugs in silico, rather than just in the wet lab. When we’re doing basic drug or vaccine design in simulations – computational immunology – we can build and experiment with them in computer simulations, to determine approaches that have the highest likelihood of working best in practice. With COVID, for example, it was less than six months from the initial sequencing of the coronavirus to initial trails with candidate vaccines. I think the application of computing in medicine is already important and will continue to be really important.

What is a piece of advice you would give to somebody who’s just entering the field of computing? How should they prepare for a growing and changing research field?

As a piece of actionable advice, I would say to students that they should study principles. They should study foundations, because much of what that they’re going to be doing in the future will be done a lot differently than it is today. The OECD and World Economic Forum have cited estimates that 65% of kids entering grade school will work in jobs that don’t exist yet.  That a hard number to pin down, but it speaks to the challenge of how do you study and prepare for a job or career that doesn’t even exist when you graduate from school? In my mind, the only way you can do that is by studying foundationally, having strong principles, and understanding the basics.

What fields do you see students most interested in?

I sense a strongly increasing trend among our students towards doing socially impactful and socially relevant work, including research. This makes me so optimistic about the future. If I think back to the Internet bubble in 2000, my sense then that what was driving students to take computer science courses then was that it would result in a secure and high-paying job. I see much less of that kind of thinking now, and much more of “Yeah, okay, I’ll earn a good living, but I’m really doing this because it’s a socially good thing to do. And I’m going to feel good about the work I’m doing.” That’s a really noticeable trend that I’ve seen in the last five years, and it’s a wonderful thing.

Published on August 23rd, 2022

Last updated on August 29th, 2022