acm - an acm publication
Articles

An Interview with Richard A. Demillo

| BY Ubiquity staff

|

To view the full citation in the ACM Digital Library please click here

Richard A. DeMillo is the Dean of Georgia Tech's College of Computing. He previously was Hewlett-Packard's chief technology officer and served as director of the Georgia Tech Information Security Center. Under DeMillo's leadership, Georgia Tech's College of Computing has replaced the core curriculum for undergraduates with an ambitious and innovative Threads program, as he explains in this interview with Ubiquity's editor-in-chief John Gehl.

UBIQUITY:     Why don't we start by having you tell us a little bit about Georgia Tech's College of Computing, and how it evolved to be where it is today.
DEMILLO:     Certainly. The College grew out of a much older academic School of Information and Computer Science. At about the end of the 1980's, Georgia Tech and Carnegie-Mellon were really the first two research Universities to recognize information technologies as the driving force not only for higher education but also for the economy in general, and so CMU formed a school or college of Computer Science in the late '80s, and just about that time Georgia Tech hired John Patrick Crecine, the provost of Carnegie-Mellon, to be the new president of Georgia Tech, after Joe Pettit died.
UBIQUITY:     And that was the spark of the new program at Georgia Tech?
DEMILLO:     Yes. Since Pat Crecine had been through this experience at Carnegie Mellon, he brought with him a knowledge and love of technology, and so Georgia Tech became the second university to create a College of Computing. We didn't call it Computer Science, we called it a College of Computing. So there is an intellectual tie from the very earliest days of the College of Computing at Georgia Tech with Carnegie Mellon's program in computer science, which at that point already had a rich and lustrous history in this area.
UBIQUITY:     Why was Georgia Tech's activity called computing rather than computer science?
DEMILLO:     At about that time there was much discussion as to whether or not computer science, which focuses really on underlining technology, was really too narrow a view of what information technology should be in a university, and so Georgia Tech decided to call it, not the College of Computer Science, but the College of Computing - self-consciously, I think. mimicking the terminology of engineering. In engineering, the underlining technology of an activity is extremely important, but equally important - and maybe more important - is what you actually do with the technology.
UBIQUITY:     So the activity and uses of computing were really built into the new College from the outset.
DEMILLO:     Exactly. That was in 1990, and over the next seventeen years the College grew dramatically, from its original beginnings with just a dozen or so faculty members and a small number of graduate students. What we are today is a college that has three departments, 2,000 students, 130 faculty members, and we're growing our student base at about 30% a year. That's a dramatic number, because if you look across the board at computing programs around the country, most are still in a steep decline after the dot-com bust. Yet at Georgia Tech we have been growing in all aspects of the program. Undergraduate programs will be up 40% this year, and we will have the highest number of women ever enrolled in computing programs at Georgia Tech, with the highest number of under-represented minorities ever enrolled at Georgia Tech. We just became, for the first time, a Top Ten ranked program by US New & World Report, and I attribute much of that success to the insight that Pat Crecine and Pete Jensen and others had in the late '80s that "computing" will eventually be considered a broad intellectual activity like "engineering" is.
UBIQUITY:     We should probably note that Pat Crecine just died a few weeks ago.
DEMILLO:     Yes. He was certainly a change agent at Georgia Tech. Among the legacies of his vision, in addition to the College of Computing, are his focus on humanities, and the fact that Georgia Tech has a well-respected and in many areas nationally ranked Humanities College, due to Pat Crecine's insights.
UBIQUITY:     What do you think explains the College's ability in just the last few years to become so dramatically competitive with computer science programs in the rest of the country?
DEMILLO:     I think a lot of it has to do with the determination at Georgia Tech and at this college to think "outside the box" and not be averse to taking risks. When I joined Georgia Tech in 2002 I had just come from many years in industry, and one of the things I was determined to do if I ever got back to Academia was to make every attempt to broaden the education of technology students.
UBIQUITY:     What made you so determined about this?
DEMILLO:     Because my experience helped me understand how difficult it was to get even top undergraduates from the best programs in the world to become interested in more than their narrow little specialty. I was seeing technologists become more and more Dilbertized, to become more and more caricatures of what I knew an Engineer and a Technologist should be.
UBIQUITY:     How did they become Dilbertized, to use the term you've coined for them?
DEMILLO:     Because they had been forced to endure outdated and narrow programs at the undergraduate level in the top universities. So when I came back to Academia I came with an agenda to help fix that.
UBIQUITY:     Let me interrupt you and to ask what you can tell us briefly about your industry background.
DEMILLO:     I was Chief Technology Officer at Hewlett-Packard just before I came to Georgia Tech. Before that I was head of computing research at Telcordia Technologies, a company that used to be known as Bellcore. I had also served several years in the federal government at the National Science Foundation, so I have seen the public sector from the federal level.
UBIQUITY:     What did your organizational experiences have in common?
DEMILLO:     I had been in companies and organizations that underwent massive transformations. For example, at H-P I had gone through a kind a wrenching transformation with the acquisition of Compac, and I had it in my blood (and still have it in my blood) to be a change agent, so this opportunity to head Georgia Tech's College of Computing was exactly the right position for me at exactly the right time. And I'd also say with the right administration and the right faculty. It gave me the opportunity to start with a blank sheet of paper and really redesign undergraduate education in computing in a very different way.
UBIQUITY:     What kind of an approach did you take to redesign computer science education at Georgia Tech?
DEMILLO:     It started with was an "offsite" - an off-campus meeting - that we had just after I took over, in which we as a faculty got together and made some decisions about what we were going to place at the top of our priority list - and, just as important, what we were not going to place at the top of our priority list. In the midst of those discussions one of our senior faculty members came up to me and said, 'You know, one of the big problems we have with computing is the low numbers of women that enter the field.' Well, that faculty member's observation was correct, and my theory has always been that if you want to get diversity of membership in an academic program you first have to have intellectual diversity. And that started me thinking about the whole idea of intellectual diversity, and about connecting that idea to the observations I had made in industry about the narrowness of people who were educated in current computer science programs. And once you start thinking about diversity, you start thinking about what diversity means beyond gender or ethnic or racial diversity - it means global diversity, it means national diversity, it means diversity between disciplines. The question then is how do you build that into a program in which there are layer upon layer of prerequisite requirements in order to get to a certain point in the scale of expertise that faculty members feel students have to achieve? But that path is a slippery slope, because maybe intellectual diversity can't be achieved by building vertical stacks of courses on top of each other; maybe diversity has to be achieved by putting together, essentially and academically, conversations that students can participate in that will carry them towards a goal. And you know that seems like an abstract kind of (and not very useful) observation until you start connecting it to the possibilities in the university for using courses as a vehicle for having those conversations with students. And just coincidentally at that same time we were considering this, Tom Friedman's book The World Is Flat came out, and gave us a vocabulary to talk about these ideas, and we realized that really what we are doing here at Georgia Tech is putting together people in ways that that can be successful and allowing Georgia Tech graduates to be successful in the new technological world. The Friedman book gave us an opportunity to step back and see that maybe what we need to do is put together a skills prospectus that allows people to decide where they want to go: globally, economically, intellectually, once they leave Georgia Tech. And over the course of time we made those abstract ideas very concrete we put them together in a curriculum that we call "threads." Incidentally Tom Friedman noticed what were doing, and visited Georgia Tech, and was so taken with the idea that he devoted a chapter of the second edition of The World is Flat to us. I think its chapter six of the book, called, "The Right Stuff."
UBIQUITY:     Congratulations, that's terrific.
DEMILLO:     It's really a great story, and I also have to give a nod to Daniel Pink, who wrote a book called A Whole New Mind. The chapter in the Friedman book is a throwaway reference to the Daniel Pink book, because Daniel Pink spends a number of chapters talking about the role of right-brain creativity, talking about the role of community, talking about the role of conversation and narrative of success at all levels of activity, including technological ones. Pink noticed that virtually all the value in a "flat world" is added by things we normally think of as right-brain contributions. Story-telling, for example is one of those. And that insight also had a big influence on how we put this "threads curriculum" together. So if you want to look for kind of intellectual fathers of this idea I would say it was my off-hand conversation about diversity with a faculty member, it was the Tom Friedman vocabulary, and it was the right=brain contribution that came from the Daniel Pink book. In the space of eighteen months we essentially took apart a curriculum that has been essentially intact for almost 25 years and replaced it with a brand new curriculum that does not have core courses. That allows students to peruse approaches to computing that are really tailored to what their interests are and at the same time emerge from the program as Georgia Tech graduates who can hit the ground running and be a credit to the Georgia Tech brand as a fine undergraduate program. Essentially what we've done is figured out a way of running 28 separate degree programs under one degree umbrella and to do it in a way in which each of those programs is an accredited computer science degree from Georgia Tech.
UBIQUITY:     I'm enormously impressed, I really am.
DEMILLO:     I never would have thought I would be so enamored with Higher Ed and undergraduate education, but I think there is a revelation here that American universities have to be innovative or r risk being isolated in a world that is going through these types of transformations. At this point in my career, towards the end of my career, I have become quite passionate about this.
UBIQUITY:     This is really quite a pioneering effort. How is Georgia Tech is viewed, not by potential students, but by competitive programs in other universities?
DEMILLO:     I think we are regarded as leaders in this area. We've been responsible for leading the computing research community, which has traditionally taken a very hands-off attitude toward undergraduate education - kind of leaving it in the hands of curriculum committees - so the research community has realized that its pipeline it at stake unless undergraduate is renovated by leading an effort to establish the first national committee of research universities to have oversight of undergraduate programs in computer science.
UBIQUITY:     Are there any similar programs to yours?
DEMILLO:     There are a number of clones of the threads program that are springing up around the country. There used to be I think nine programs that are exact replicas and then a number of other universities that have adapted the basic ideas behind the threads to their special circumstances. MIT for example has decided to look at a thread-like conspectus but at the university level, not just in computing or computer science. And then smaller universities which, frankly speaking, can't afford the overhead it takes to run a program like threads, have scaled it down but kept what I think the important concepts are. So, by my count there are something like 47 or 48 undergraduates programs around the country, that are going down the same path that Georgia Tech went almost four years ago.
UBIQUITY:     Are there no competitors that regard you as an odd duck doing a completely radical change of things by destroying the core curriculum?
DEMILLO:     It's been an interesting journey. I would say some of the most highly regarded programs in the country have been the least flexible. But I don't think there has been massive resistance. We get called in to participate in curriculum reviews for some of those programs reviews, for example, and frankly speaking they kind of a look with an envious eye at what's going on here at Georgia Tech and are trying to do within their own system what we did here. Georgia Tech has historically, I think, had the ability to be more flexible than many institutions, and the speed with which we adapted is evidence of how quickly this place can move if it wants to.
UBIQUITY:     Have you had to expand your advisement system substantially?
DEMILLO:     We are in the process of doing that. We recognize that when we started the program we were introducing considerably more complexity into administering an educational program. There is no longer a one-size-fits-all introductory course; there is no longer a single common core. We have students that can be spread out over any number of threads in a given course. The former academic advisement system used to be a very good but was very checklist-oriented.
UBIQUITY:     How does it work now?
DEMILLO:     Now the advisor has to have some depth of knowledge of what computing can consist of and what it is going to be used for by the student. And so establishing the right level of advisory apparatus, particularly for freshmen and sophomores, quickly became a priority for us, and it is something that we have worked very hard on. Actually, the delivery of course material when you have this kind of complexity turns out to be interesting, in that an instructor can be standing in front of a class of 50 students yet lecturing to 28 different classes because those are the number of threads. Therefore when an instructor assigns projects or gives an exam, that instructor has to be aware of who is in the class, and so there's an element of real time involved -- almost a customer-management layer that lets the instructor present the material appropriately to the different constituencies in the class.

We think that has to be automated in order to keep cost controllable for this kind of system, so we are spending quite a bit of time on that aspect of things and we have people who are really passionate about pursuing this. We are spending quite a bit of time thinking about what kind of software support we need to run a program like this, and we're going to start seeing those software products over the next couple of years. They will of course be made generally available for the community.
UBIQUITY:     Who are some of the people you would like to acknowledge in helping you?
DEMILLO:     Well, there is a whole bunch, and this was an effort of the whole faculty - but there was clearly a leader in getting this done. My first hire when I came to Georgia Tech was a fellow named Merrick Furst, who had been a research colleague for many years, and was at the University at California at Berkley at the time. He had been a faculty member at Berkley; he is an entrepreneur and had been an Associate Dean at Carnegie Mellon during the formation of CMU's college of computing. So, I hired him away from Berkley to Georgia Tech as an Associate Dean, and asked him to take on responsibility for putting together the pieces to it took to do all this. He was extremely instrumental in pulling the intellectual strings to make a coherent program.
UBIQUITY:     Others?
DEMILLO:     There is also a fellow on our faculty named Mark Guzdail, who has a Ph.D. in education as well as a Ph.D. in computer science. He had been experimenting for many years to develop a way of teaching an introductory computing conspectus to student from the humanities, who were more interested in media and narrative than they are in the nuts and bolts of engineering. And so he had put together a course that used media computing as the vehicle for teaching freshmen.
UBIQUITY:     And the list goes on?
DEMILLO:     Absolutely. A third fellow who really took the challenge of making and administering a program out of all these pieces is a fellow named Charles Isbell, who is currently my Director of Academic Affairs and Undergraduate Studies. Charles is just a force of nature when it come to moving all the pieces into the right part of the room and make things are exactly right. He is also looking at what the technology support has to be in order to make this a success. So those are our main guys, but of course there are many, many people involved now, and it's an effort that probably consumes a third of the college and let's us focus like a laser on the quality of undergraduate education.
UBIQUITY:     How would you like to sum up?
DEMILLO:     I would just like to say that in my view we're not done with this. As I said, I think this is something that all of this country's universities are going to have to face over the next generation. You know, Georgia Tech has a mission statement which says that we aim to be the defining University of the 21st century, but I don't think that the people who said that really anticipated what that means - because the defining University of the 21s century t is not going to look very much like the universities of the 16th century -- which is how universities in the US function today. So we are throw out a lot of bath water (baby and all), and to discard a lot of assumptions about how universities have to work in the future. We are just at the very earliest baby steps in doing that. But some of the same issues that came up in the definition of thread are going to come up in this broader contest.
UBIQUITY:     When we began the interview I had no idea I would be so impressed. I congratulate you, I really do.


END


********************************************************


News from Dean Richard A. DeMillo:

June 12, 2008

Dear Members of the College of Computing Community and Friends,

I have sent a letter to Provost Gary Schuster informing him of my desire to step down as Dean of the college next semester and in a very gracious response he has accepted my resignation effective no later than November 1, 2008. An Interim Dean will be named as soon as practicable and together we will begin what I hope is a smooth transition process. My intention is to turn as many internal administrative duties over to the Interim Dean as possible in the next few weeks to allow me to concentrate my efforts on the many external and international projects that now have to be concluded by November 1.

I leave the Dean’s office at a high point in the college’s history having achieved all of the goals that were set when I was named Dean in 2002. We are a financially and administratively sound college with the best finance, HR and communications organizations on campus. We have grown dramatically while many other top computing programs around the country have experienced steep declines and cut-backs. Our faculty headcount is up 40 percent, our operating budget has grown even more rapidly (allowing us to pursue new initiatives) and our research has grown 60%. We have hired well at all ranks. Our junior faculty are among the best in the world and we continue to lead in such areas as NSF CAREER awards. Most importantly, we are now ranked in the top ten and are poised to climb even higher.

We have formed three new schools and gained worldwide attention for our leadership in defining the discipline of computing, a goal that was set for the college when it was formed in 1990. We have launched in total seven new degree programs. Furthermore, we will continue to set the agenda for computing with our exciting plans to form two additional schools over the next few months, the School of Biomedical Informatics and the School of Information Science.

We have invested wisely in areas like robotics, algorithms, theory, and high performance computing and hired the best senior people in the world to lead those efforts. Signature centers like GVU and GTISC have been renewed under great new leadership and we have partnered well in cross-campus initiatives like the Health Systems Institute.

As you know we focused much attention on a dramatic turn-around in undergraduate education. I am happy to report today that our educational programs have never been better. We have become an international leader in undergraduate education. I hear from hundreds of faculty and students from around the world who are grateful that Georgia Tech took on the challenge of renewing how we teach computing. Our enrollments continue to rise and this year we will enroll the largest number of women and minorities in the College’s history. Our incredible instructors continue to be among the best on campus. Our international presence is strong and growing, fueled in part by our reputation as a leader in education.

As many of you know, I do not believe it is possible to be a great college without also influencing the computing industry. We are now among the top university collaborators for the IT industry. Our commercialization efforts have been aimed at helping us achieve greatness in this area as well. We have spun out four companies and in the process created an investor ecosystem that will allow faculty and student entrepreneurs to be successful. We have established the idea firmly that commercial success will be rewarded by the college.

Our community is also the strongest (and largest) it has ever been. The Tech Square Research Building, The Klaus Building, our growing presence in Metz and Oak Ridge National Labs, and our current College of Computing Building – which is undergoing a major renovation – give our community unparalleled campus-wide scope and geographic breadth. We now have an active alumni association, an incredible Advisory Board, and thriving student communities. We have kept our ties to the origins of the College with a network of emeritus advisors, including Peter Freeman, the founding dean and constant friend of the college. Our award-winning web presence ties all of our constituents together and our high visibility in the local and national media lets everyone know of our accomplishments.

The path forward is more challenging than the path we’ve traveled. We are now in competition and cooperation with the best programs in the world. The expectations are high and the next dean will be called upon to provide exceptional leadership.

My plan is to return to the faculty (something I tried to do in 2002) as a Distinguished Professor of Computing and Management. I will be taking a leave for a couple of semesters to complete some long-delayed writing projects and to restart my research agenda, a subject I will be talking to some of you about in more detail over the coming months.

I am proud of all that we have accomplished and I am in particular proud of all of you for your hard work and dedication to our shared vision of Computing at Georgia Tech. I look forward to working with you in my new role.

With Best Wishes,



Richard A. DeMillo
The John P. Imlay Dean and
Distinguished Professor of Computing

Source: Ubiquity Volume 9, Issue 25 (June 24 - July 1, 2008)

COMMENTS

POST A COMMENT
Leave this field empty