“Imagine living without the internet, search or email”

Professor James Larus © 2021 EPFL

Professor James Larus © 2021 EPFL

After eight years at the helm, Professor James Larus steps down as Dean of EPFL’s School of Computer and Communication Sciences in September. He reflects on his tenure and his career in “the most influential field in the last 50 years”.

James Larus was at high school in the United States, in the 1970s, when he saw his first computer. “They were very rare in those days and my school had a minicomputer, which was the size of a large cabinet, but the great thing was, there was an actual computer that we could program. It was very hands on, I remember it had lights and switches on the front and you actually used the switches to enter things into it. And I thought ‘this is an immense amount of fun’!”

Fast-forward half a century, and a lifetime of research in a field that is continually evolving at lightning speed, and Larus is stepping down after eight years as Dean of EPFL’s School of Computer and Communication Sciences (IC). He has immensely enjoyed the role but feels like now is the time to make way for someone with a fresh perspective. Professor Rüdiger Urbanke, an expert in powerful channel coding methods and who has been with EPFL since 1999, will begin as Dean in September.

Theory versus practice

After being captivated by the minicomputer at high school, Larus studied computer science at Harvard but found it very theoretical, contemplating a complete change of direction to become a lawyer. Fortunately for the field, he chose to complete a Masters in Computer Science at Berkeley and found the hands-on approach much more interesting. “They built things! It was a totally different perspective on computer science – there were great theoreticians there but also people that built systems. I remember super exciting projects like ‘RISC’ a new kind of computer system that proved extremely influential,” he says.

After becoming a Professor at the University of Wisconsin, Larus went on to work at Microsoft, where he is most proud of his work on the Singularity Project, a new operating system constructed in safe, high-level programming languages, as well as building a hardware accelerator for the Bing search engine.

The luxury of time

Back in academia, at EPFL, Larus says it’s obvious that universities don’t have the physical or financial resources to do research on the same scale as today’s tech giants, but what they do have is the advantage of time. “In companies for most people the deadline is immediate, you are worried about the next product that you are going to launch and the next quarter’s results. You don’t have time to think about what comes afterwards,” he says. “In academia we are trying to work out which way the future is going and think ahead three, five, ten years to try to come up with solutions to hard problems. We have the luxury of trying to find the real solutions as opposed to the most expedient ones.” So where does Larus think the future is going?

AI, machine learning, quantum computing and sustainability

He says that deep learning and artificial intelligence will have a profound impact on how we live but have yet to see anywhere near its full potential. “Part of the reason for this is that deep learning is somewhat of a ‘dark art’ and while students are taught the underlying math and lots of examples, we can’t yet give them principles. We don’t know what they are and we literally don’t know why it works the way it does. IC fortunately has faculty members who are looking at these fundamental questions.”

Quantum computing is another area that Larus believes is finally starting to gain momentum. “After years of talking about it, companies are now building quantum computers. They aren’t very capable, they are extremely expensive and we don’t really know how to use them that well. But, it’s just the beginning and the potential for solving problems that we can’t solve today is huge.”

“Last but certainly not least, I think sustainability is a pervasive issue for all fields and computer science is certainly one of them. There is a real question as to whether combatting climate change is compatible with the ever-expanding energy needs of computing, but I think we’re also going to be deeply involved in the solutions because we will just not be able to build complicated systems and infrastructure without advanced computation,” he continued.

A forward looking legacy

After handing over to Urbanke, Larus will remain at EPFL and is looking forward to getting back to research, but he has been reflecting on his achievements in eight years as IC Dean.

“I’m proud that I have hired a large number of young faculty members. They have lots of ideas, they are ambitious, they want to do things so it changes the way in which the department operates. It’s a lot more exciting!” he said. “I’m also particularly proud that I have hired and promoted more women. We now have seven women all on the tenured faculty track and more offers out to others this year. It’s not as large a number as I would have liked but it’s closer to where we should be,” he continued.

“Finally, I have seen the IC school build important partnerships with other parts of EPFL. This collaboration is good for EPFL and it is good for science in general,” Larus says.

His final message is one of encouragement that he hopes will inspire the thousands of students who have chosen to study computer science at EPFL and who make the IC School such a vibrant place, “You’ve made a great choice! As a field we have had an enormous impact, and this impact will continue to grow with AI and machine learning only in their infancy. You’re lucky to be in a field that is continually evolving. Jump in and make the change that you want to see happen!”