Interview with Daniel P. Siewiorek

Daniel Siewiorek

Daniel P. Siewiorek
Director, Human-Computer Interaction Institute
Buhl University Professor of Electrical and Computer Engineering and Computer Science

Where are you from?
Cleveland, Ohio

I’ve heard of a Siewiorek motto, can you elaborate on "I reinvent myself every ten years" and what this means from your perspective.
Have you heard of Moore’s law? The amount of memory, processor speed, bandwidth, doubles every 12 to 18 months. In 10 years… it’s a factor of 1000. So if I have that much more capability and that much more opportunity, I have the possibility to do something previously thought to be completely unrealistic a decade before. You want to be at the forefront and do research that will have a big difference.

You’ve produced 58 PhD’s so far in your career, which is a very high number. How does teaching compare to research in your eyes, and what kind of equilibrium do you have between the two?
It’s something like a pyramid... think of breadth verses depth. We use education projects such as the Rapid Prototyping Course to explore the breadth and find interesting systems. When we find a gap in a technology, we then have a great area for more in depth research that PhD students can explore.

How do you balance family and academics?
Well one is the compactness and family oriented environment of Pittsburgh. One example is when both of my daughters were in high school and played soccer. So I went over to watch their game, and then came back to Carnegie Mellon. The biggest thing is making the family part of what you do. For example I had my daughter and her best friend help assemble our first wearable computer. My wife deserves the credit for maintaining the family focus such as having the whole family home for dinner, and to make other family times. Now my girls are into ultimate Frisbee, so my wife and I go to their games. So basically we carve out some time each day, and each year try to go on family trips, and learn what entices our daughters to go on family vacations...exotic locations!

Your work has transitioned from reliability and multiprocessors, to design & rapid prototyping, to human-centric design. How did these progressions occur, and what was the motivation behind them?
Each one inspired the next. There’s a trajectory that I’ve gone through… that in hindsight makes sense. When we first started we were working on multiprocessors when Raj Reddy observed that if he needed a single board computer with twice the memory of an existing computer, it would take a designer six months to produce a new design. Why couldn’t the design be done in 24 hours? After 3 PhD theses, 8 masters projects, a dozen undergrad student projects, and about a decade, we were able to meet Raj’s challenge. The MICON system, based on artificial intelligence techniques was able to create a design from high-level specifications in ten minutes and, coupled to a layered deposition manufacturing system, fabricate the design in 24 hours. The results from that project inspired work on wearable computers, and then after about 10 years, that inspired even further work in HCI. By building systems we realize there are gaps in our horizon. Over the years, what becomes important changes, and that directs our research.


What is your academic background? Why did you choose those areas, and do you feel that they prepared you for the breadth of your work?
In elementary school I decided to go for electrical engineering. I didn’t know what they did, but it sounded good, so I put it down on my college application. Electrical engineering gave me a very broad background. You could be almost anything... and I liked building things. Throughout my undergrad years, I had a number of summer jobs that taught me what I didn’t want to do. Computers weren’t that big at the time, but digital design came easy to me, so I pursued courses in computer engineering – before it was called that.

Where did you study and what steps did you take after graduation?
I went to University of Michigan for undergraduate, and Stanford for graduate school. Angel Jordan came to Stanford to recruit saying there was an opening in EE and CS here, and that they were looking for a joint appointment. I found that Carnegie Mellon had a very strong relationship with Digital Equipment Corporation, and was doing a lot of work in the computer field. I really didn’t know any other university that was building systems and putting them out into the field to get evaluated as Carnegie Mellon was doing.

How long have you been at Carnegie Mellon?
35 years!

I’m sure you've had so many options and opportunities in the past 35 years, why did you choose to stay at Carnegie Mellon for the majority of your academic career?
Carnegie Mellon is very collegial, very innovative, and there are hardly any barriers. If I want to work with someone in fine arts, I can do that. People are very receptive to engineers, and if I want to do something, I do not need to ask permission. There’s very little red tape.

I also love the students, they keep me young. They are always full of energy, and the neatest thing about them is that they don’t know what can’t be done. Students are not constrained by preconceived notions that are often found in industry. On one hand, it's hard to see them graduate, but it’s great to watch people grow. In our Rapid Prototyping course you can see people grow and expand their horizons in one semester.

What other academic directions do you still want/plan to undertake?
The human side of things is very important. In HCI we are now working with and learning from cognitive psychologists, social scientists, and designers. It’s great to have the opportunity to learn new areas and new ideas. I want to build systems that work, and methods to measure their impact. So basically exploring HCI from an engineering perspective.

What is the most interesting HCI project you have worked on?
I’m going to throw out a couple: Certainly, there have been some pivotal ones. Maintenance of aircraft was an early driver in wearable computers. The exciting projects are those that take you from school into the actual workplace. You get to see how people currently do their job and develop ways to improve on it.

Another project is trying to develop a cognitive assistant for office workers - another system with a human in the loop. We are attempting to bring together computer systems, and human interaction with AI. The human isn’t interacting with the AI system – like in a chess game. Instead, the AI offers human advice –without becoming an annoyance. A partnership develops rather than the user trying to hide or get rid of the assistant. I like projects that bring together groups that weren’t together before. So the exciting thing for me is doing something advanced that couldn’t be done with one field alone.

What major leaps do you think the HCI will take in the next 5 years? If any..
I’m really fascinated by Marcel Just, who has been working with functional MRI’s. He's been giving people tasks, and measuring what part of the brain is affected. He found that there’s this capacity in the brain, where if you have the brain do 2 simultaneous things, they only get 70% accuracy, compared to 100% when doing them separately. So going back to the multi processor, the brain doesn’t seem to know how to take advantage of parallel processing. So maybe there are things we know or can learn about processing that can help us learn how to use our brain better. So when the brain is in cognitive overload, we can coach the human to get out of overload and better use their time. But I think that might take a little longer than 5 years, maybe a lot longer… I think psychologists are calling this the century of the brain.

What are your hobbies or other interests?
Well first, I really enjoy what I do. I am an amateur history buff. I like looking back at the World War Two generation and trying to see what I can learn from them, it was a very generous generation. I really like looking at historic technology. I was in Florence recently, and looking at all the sculpture and diversity in artists in one place. It’s a concentration of expertise, all the great artists in one spot… and in some ways I feel like that’s what Carnegie Mellon is like. And you can’t really see that research concentration anywhere else. It’s a concentration where people learn from each other. I can borrow something from ECE or Robotics and create a new system. It’s a very rich place of ideas and individual pinnacles of excellence so you can do so much more that by yourself.

Favorite Quote?
When encountering a new technology that is less than intuitively obvious to use - “The engineer who decided this should be shot”.