New computer science chair Michael Franklin discusses the past, present, and future of computation.
Data science scholar, entrepreneur, and software developer Michael Franklin has begun his appointment as the Liew Family Chair of Computer Science and senior adviser to the provost on computation and data science.
Franklin joined the University this July from the University of California, Berkeley, where he chaired computer science and led the Algorithms, Machines, and People Laboratory. At UChicago he leads the department in a major expansion of faculty, educational programming, and outreach, increasing the scope and impact of computer and data science and building collaborations on and off campus.
Tell us about your experience in data science.
I’ve been in the field of databases and data management about 30 years now, investigating how to work with data across different types of computing environments. Most recently my work has involved scalable analytics, where you build systems that can expand as the amount of data needing to be analyzed grows.
I also work with processing data on small devices—how to do data processing in a highly distributed, highly unreliable network. Human computation in crowdsourcing is also of interest—how to get groups of people connected by a network to collect data to solve analytics problems, like reporting traffic, pointing a telescope at the sky, or carrying air pollution sensors.
What exactly is data science?
The term “data science” grew out of industry—web-based companies like Facebook and LinkedIn who were gathering increasingly rapid streams of data. It was obvious that there was value in this data, but they needed a way to extract it. The software, hardware, and even theory that traditional analysts had been using weren’t adequate for the volume and diversity of data and the types of questions these companies wanted to answer.
So a new field arose that involves concepts from computer science and statistics, as well as applied math and social science. It spread from industry to academia, but because of its breadth, it’s a challenge to figure out exactly where data science “lives” on a university campus.
What are your plans to bolster the CS department?
What’s exciting about the University of Chicago is an eagerness to engage with computer science, statistics, and applied math and to work on and solve computationally intense problems.
We need to continue this process of scaling up the computer science department in core areas as well as outreach to other fields. We also need to build on the interdisciplinary work that’s traditionally been done at UChicago, particularly projects housed in the Computation Institute. We have a committee working to define the future of the CI and how it will relate to computer science and other departments and divisions as an intellectual nexus for computation and data-oriented research on campus.
Beyond campus, CERES—a center for unstoppable computing—is a project involving UChicago faculty and corporate partners. The city of Chicago has a growing tech ecosystem, and UChicago is poised to increase interaction with city government and business. A strong computation effort will be key to those engagements.
How has computation changed research in fields like social sciences and humanities?
Computation has changed research in almost every field. It’s easier to collect interesting data about the behavior of individuals and large groups. The challenge is finding the signal in all that noise. If you can find it, you can see things that weren’t visible before.
And it’s a two-way street; social science is also going to have a huge impact on computer science because computing has become deeply ingrained in everyday life. Advances in computing will depend on understanding how people interact with technology. My plans for the department include building bridges and training students to move comfortably between traditional, technical questions and also social science questions.
How far are we from integrating computational devices directly into our bodies?
I don’t know, but the trend is clear. Most people already have a cell phone stuck to the side of their heads.
Should we be worried?
With any technology there’s potential for great benefit and fear. Part of computer and data science education is teaching students to think about the broader impacts and ethical implications of what they’re doing and the technology they’re building.
Was there fear during the early days of computers?
There’s always been nervousness around any sort of automation. Industrial Revolution workers feared machines would displace them. Mechanical looms were a popular target for British Luddites, who destroyed machines in protest over low wages and poverty. The “programmable” Jacquard loom—one of the inspirations for early computers (see “A Pattern of Progress”)—was fiercely opposed by Parisian silk weavers.
And for computers, absolutely. Look at science fiction. Isaac Asimov—who wrote dozens of robot-based short stories and novels—defined three laws of robotics to make sure robots didn’t harm people. There’s always concern, but it can and must be managed.
How did you get involved with computer science?
When I was a senior, my high school got a new computer. It wasn’t a fancy school, so it was a big deal. I signed up for the class, and it turned out I had an aptitude for it. One day my computer teacher made an offer: if I would write a certain program for him, he’d give me As the rest of the year and I could do whatever I wanted.
He had a friend teaching at another school who was bragging about their new computer and the programs they got for free. Evidently my teacher said, “Well, one of my students could write those programs.” And they made a bet.
When I think back to that program, I cringe because I now know how I could have written it so much better. But it worked, and my teacher won the bet.