Home > News > Interview with Dr. William A. Wulf

Interview with Dr. William A. Wulf

From Computing Research News
September 1, 1997

O ver the summer, CRN interviewed Dr. William A. Wulf, President, National Academy of Engineering (NAE), Vice-Chair, National Research Council (the principal operating arm of the NAE) and AT&T Professor of Engineering and Applied Sciences at the University of Virginia, Charlottesville.

Dr. Wulf also served as assistant director of the National Science Foundation (NSF), was chair and chief executive officer of Tartan Laboratories, Inc., Pittsburgh and was Professor of Computer Science at Carnegie Mellon University. Prior to his election as NAE President in 1996, Dr. Wulf was chair of the NRC's Computer Science and Telecommunications Board (CSTB).

Summary: During the CRN interview, Dr. Wulf reflected on his first year as NAE President, and in response to CRN questions, discussed his views on a range of subjects including continuing education for engineers and government funding for computing research vis a vis other scientific disciplines. He also urged those in the computing research community to serve a few years in government. In commenting on the community's presentation of the proposed NGI initiative, Dr. Wulf noted it was a good example of why having additional government-savvy folks is important to future initiatives. Dr. Wulf also emphasized the need for coordinated leadership among computing research groups, and cited CRA's previous summit meetings as particularly helpful in this regard.

CRN: What is it like being the first computer scientist to occupy the top position at the NAE? How do you see your role?

WULF: I have a couple of responses to that.

The first one is that-except for other computer scientists-nobody has mentioned the fact. I find that fascinating. I take that as an indication that the discipline has achieved a degree of maturity and recognition that apparently doesn't strike any body as unusual that would be the case.

On the contrary-and we can see this in the context of CSTB as well-the centrality of the discipline to almost everything else makes it seem like a good idea to have a computer scientist in a position like this.

The other half of the response relates to what another computer scientist said: "Have you had any reaction to being a computer scientist" (emphasis on scientist) "and heading the National Academy of Engineering?"

Two years ago, I wrote an article for Computing Surveys in which I tried to tackle the question of whether we're scientists or engineers. We happen to carry the name computer scientist, but there's obviously also a lot of the character of engineering in what we actually do. I came down firmly on the position that we are neither and we are both. The fact that we happen to be called computer scientists should not be confused with the wonderfully broad spectrum that this discipline embraces; a spectrum of everything from profoundly deep mathematics to crafty programming and everything in between. That's a strength that we should celebrate, and not something that we should worry about.

So, I'm trying to fit into your question the fact that I feel strongly about the fact that I am an engineer and I am a scientist, and I'm a hybrid that's not quite either, and I'm very proud of what I am.

CRN: What areas of study would you like to see NAE-through the NRC-take on in the coming years? Do you have a particular agenda, or certain areas that you'd like to see tackled more than others?

WULF: There are two kinds of answers to that question. There's a generic answer, and there's specific content.

The generic answer is that I think there are a host of ways in which the NAE can more completely fulfill our 1863 Congressional charter. I suppose not everybody who reads Computing Research News understands that the Academy is a private corporation that operates under a Congressional charter. That charter calls on it to answer questions on science and technology for the government.

We also have this honorific character, but the reason for existence is the 1863 charter and providing advice to the government. I feel strongly that engineering is the activity that creates artifacts and processes that improve people's lives. The importance of engineering to the country is not as well understood as it might be. And so, a lot of what I want to accomplish in the next four years has to do with more completely fulfilling the charter.

I've got some tools at my disposal. I've got a fabulous membership, for example; the most accomplished engineers in the country. I've got a convening capability. And I've got a bully pulpit. I intend to use all of those tools.

As for specific content, let me focus on two topics: education and the changing nature of engineering. I think there are things that the Academy of Engineering needs to say about engineering education. Generally speaking, engineering education is the responsibility of the professional societies.

For example, the specific curricula for electrical engineering or civil engineering are not issues that NAE should deal with. But there are more generic issues I think we must tackle. For example, the question of the baccalaureate being the first professional degree. Engineering is the only profession for which the baccalaureate is a professional degree. Given the amount of information that one needs to successfully practice engineering, I think that's a question that we have to reopen.

Another issue is career-long education for engineers. We don't have culture in engineering that demands that we continually renew ourselves like, for example, physicians do. Yet, there are estimates on the "half life" of engineering knowledge that are in the 2 1/2 to 7 1/2 year range. I think we need to change the culture of engineering-and computing-to one in which it is expected, and it is the responsibility, of individual engineers to continually re-educate themselves.

I believe that the nature of engineering practice is changing fairly dramatically right now. Part of that is induced by computers, by new materials, by a variety of things that increase the "design space" in which an engineer has to operate. My definition of engineering is "design under constraint," and there are also a large number of new constraints. Functionality and cost were the primary constraints in the past, but now we have to be concerned with environmental qualities, safety and reliability, maintainability, ergonomic concerns-a whole long list of such things.

There are also the results of the restructuring of industry, the use of teams, for example, which puts engineers in a different context than they used to be. They have to be able to converse with marketers, with lawyers, with financial people; so that is another dimension of the change. There's another dimension of the change simply having to do with the globalization of the industry.

So, I see engineering as in flux at the moment and I think the Academy has a responsibility to understand that change and to put in place whatever mechanisms we have the wisdom to identify; not only to help engineers, but to help the country get the most out of engineers.

CRN: I wanted to pick up on something that you said a little while ago about the growing influence of the computing research community upon the government, or the government's appreciation, perhaps, of computing research. You mentioned the need to educate the public about engineering, but what about the government? Do you think it's come to better understand what computing research means to the federal infrastructure as well as to the day-to-day lives of citizens?

WULF: I remember when I was at NSF, people would ask me "what is the government's position on issue 'X'?" You have to understand that the government doesn't have a position on most things. The government isn't a monolith.

I want to break your question into two halves. There's no question that there's an appreciation of the value of computing. I think there's also no question that there's a diffuse, softer appreciation of the value of research in general. I don't think computing research has any special standing vis a vis other kinds of research; at least I don't perceive it....

CRN:....or should it? One of my follow-up questions (that I think this community wants to ask) is, given the set pool of federal dollars, can there be an expectation, should there be an expectation that funds be reallocated and put towards computing research, as opposed to some other type of science?

WULF: Unfortunately, virtually every discipline makes exactly the same argument; some are more convincing than others, some are more self-serving than others. But, in general, it's not a winning argument. The winning argument is in terms of the merit of specifics; not along some generalized notion that computing research is better than physics or something else.

Aside from dollars, there is another metric on the value accorded to computing research, and that is the number of individuals who recently have been in fairly senior positions in the research establishment; the fact that Anita Jones was Director of Defense Research and Engineering, the fact that Ed Feigenbaum is the chief scientist of the Air Force. I think those kinds of things are very explicit recognitions of the value that individuals in government positions place on having a knowledge of computing and computing research in particular.

I just have to throw in at this point that I'm still disappointed in the computing community for a failure to participate in the process. It is still like pulling teeth to get computer scientists to spend a couple of years at NSF as program directors. And, as a consequence, we don't have as many people who are savvy in the way that science policy is actually formulated as we should. There are opportunities that come along from time to time, and unless people have had some exposure to the process, we're not going to be able to seize those opportunities. And so-sorry-just standing on the sidelines and saying, "government, you ought to send money" is not going to work.

CRN: I think there's a segment of the computing research community that might argue "our sector is more important because it leads directly to social benefits, like better schools, or distance learning." Can other disciplines make a similar argument?

WULF: The argument that we can make and is fairly compelling is that the technology can be used in many, many different contexts. But the biotechnology community can make strong arguments about the value of their research, the materials community can make very strong arguments about the value of their research. Every one of the communities can make a case. Is our case more compelling? Well, I'm biased; I think so. But, I repeat, it's not the winning argument.

CRN: Given the increasing pervasiveness of technology in society, do you think that the computer science community has an obligation to make a statement or come out with any type of position regarding some of the ethical issues that have emerged such as intellectual property, or crypto, or privacy? Is there any obligation upon this community, or even the Academy, to do so?

WULF: The short answer is "yes."

One does have to ponder the question of who is the "community"? And who can speak for the "community"? One of the things that we have not done as well as we might is recognizing that there are multiple representatives of the community (the community is not monolithic). Specifically, we have not done well at effectively utilizing and coordinating those different representatives. The Academy can do certain things that a professional society cannot. A professional society is perceived as an advocate, and that will color the perception of whatever they say. On the other hand, they can say some things in a timely way, and from the point of view of representing their membership that the Academy cannot. CRA, being a representative of the research community, can say things that the professional societies can't because their members are individual practitioners.

A few years ago, CRA convened a couple of "summit meetings" of these groups. I thought those were very useful because they had the effect of letting us begin to coordinate our positions on ethical issues; on issues such as intellectual property, privacy, cryptography, and so on. I would like to see that happen more.

CRN: To some extent, you addressed this next question a little earlier when you encouraged the computing research community to do a tour of duty in public service at NSF, but do you have any other recommendations for CRA's membership? Is there any particular message you'd like to convey to our membership and readership?

WULF: The two that are top on my list are the ones that I've already mentioned. First, do a tour in Washington. It doesn't have to be at NSF. There's ASOFR, there's DARPA, there's ONR, there are lots of opportunities.

The second one is that I really would like to see the computing summit reconvened with an explicit attempt to cross-fertilize the leadership. Don't forget, the leadership of the professional societies changes fairly frequently. If they don't stay in touch, they will not be as effective. That's not to say they should coordinate on everything, but there are clearly questions of mutual interest and concern where they will be more effective acting in a coordinated way.

CRN: Are there any other ways they can be influential without directly entering service or coming together for the forums...?

WULF: There's a book that I would recommend to everyone. It's published by the AAAS and called How to Work with Congress. It has some wonderful suggestions. One of its suggestions for people at universities (and a suggestion that is strongly endorsed by Capitol Hill staffers) is to get your local Congressperson-when he or she is in your district-to come visit your university or your laboratory. Show him or her some interesting things. It's an entirely different kind of context than going and visiting somebody in their office. They are more relaxed. They are trying very sincerely to understand what their constituents want and need. And you will be introducing them to the fact that you have expertise that they can call upon, and that you have interests that they can help.

CRN: One final question-what are your views with regard to the NGI initiative; the proposed $100 million that the Administration keeps talking about?

WULF: First of all, I feel strongly that this is a very important idea. If we've learned anything from the Internet experience, it is that we really can't anticipate what will happen as we provide higher bandwidth and higher connectivity. But the evidence is that good things happen and so the idea of seeding some modest number of sites with very high bandwidth access is a super good idea!

However, it illustrates how careful one must be in doing these things. The NGI is in some danger at the moment; the reason is the way in which it was described. We started by focusing on the sites to be connected at a gigabit and at 155 megabits. Two Senators heard that much of the description. They didn't hear the part that this is a research project and that the reason we're doing it is to see what new ideas will emerge. They heard the first part and they said: "Oh! This is an infrastructure program; why is it just going to the elite universities? Why isn't it going to provide high bandwidth access to Alaska?" (Senator Stevens happens to be one of the individuals.)

Well there's a good reason, and it's the one that we understand. We can move the entire enterprise forward by learning how to use this higher bandwidth effectively. But that's not the way that we presented it, and it is therefore in some jeopardy.

This is a good example of why we need additional savvy folks who have had some experience dealing with Washington.

Louise Amhelm, who writes regularly for CRN, conducted and edited the interview.

[The CRA Leadership Summits, mentioned by Dr. Wulf, continue. The next one is planned for October 1997. Leaders of AAAI, ACM, CRA, IEEE Computer Society, SIAM, and USENIX will meet to discuss government policy and other issues of common concern.]


Original Article | Local Copy

 

More news about William A Wulf

 

Return To List