"Science in the National Interest: A Shared Commitment” - MIT Symposium (Session 2/3) 2/7/1995
MODERATOR: Since we're running about 10 minutes late, I'd like to start the afternoon session. After a exciting morning in which we laid out the general principles, we're now going to this afternoon-- move more into the particulars. And the afternoon session is divided into two parts, the first part of which we'll begin now, which is basic research and industry, perspectives on the life, physical, and information sciences.
And it's a great pleasure to introduce as the chairman of the first of the afternoon sessions. I was going to say my counterpart at Harvard that is Jeremy Knowles, except-- first of all, of course-- actually, I'm going to suppress sarcastic comments about Harvard. Secondly, he has not just the sciences, but the arts, which I understand are also a component of education at Harvard, although I'm never sure.
Jeremy has had a distinguished career. His undergraduate and graduate work and early part of his professional career were all at Oxford at Balliol and Wadham College. For the former and the latter, he then, however, made occasional trips to the United States.
He was a visiting faculty member at Harvard-- or pardon me-- at Yale in '69 and '71 and a visiting professor at Harvard in '73. Obviously, there was something on this side of the ocean that was drawing him here. And then, finally, in '74, he made the transition and joined the Harvard faculty permanently as professor of chemistry.
His research is on the borderline between chemistry and biochemistry. He is, in fact, made truly distinguished contributions in biochemistry, and specifically on enzyme catalyzed reactions, and on the evolution of enzyme function. In 1971, he took on the daunting task of being dean of the faculty of Arts and Sciences at Harvard.
He, like all of our other speakers today, has received an extraordinarily large number of honors, including being a fellow of the Royal Society. As a Canadian, the one I like best is that he was appointed commander of the British empire, which I guess means he's my commander, is what-- my counterpart. And the Queen's Birthdays Honours of 1993-- Jeremy Knowles.
KNOWLES: This morning, Bob, by being excessively gracious to his provost, tried to raise his salary. Why he's being nice to me-- I can't imagine, except I suppose for the possibility that he has some college-aged children wishing to go to a really distinguished institution.
MODERATOR: Actually, it's true-- not yet admitted.
KNOWLES: Everything comes to she who waits. After a reasonably stirring morning, I hope we're going to have a very stirring afternoon. What I hope will happen this afternoon is that we shall be able to discuss the shifting flow of ideas, of people, and information, and actually dollars, amongst the three places represented here today, namely government, industry, and academia.
And this concern about the flow of people, and of information, and knowledge, and ideas, and perhaps equally importantly, dollars is something that we must surely all attend to. I'm glad to say that essentially every one of our six speakers this afternoon are what I might call happy hybrids in the sense that they essentially have-- not mutant.
Hybrids-- that they have-- I'm causing a great stir in the front row, I'm afraid. But having touched in their lives government, industry, and academia. And our first speaker is Dr. Leon Rosenberg who is the only MD-- in case any of you are feeling unwell-- on our program, receiving the MD in 1957 from the University of Wisconsin. He then went more or less to government.
That is to say to the NIH as an investigator in the National Cancer Institute for six years. Yale seduced him away. And he stayed in that institution in a number of posts, founding the Department of Human Genetics and ending there as the dean of Yale School of Medicine. And as you all know, in 1991, he succumbed again to a different temptation and became president of Bristol-Meyers Squibb Pharmaceutical Research Institute, the post he now holds. Leon?
ROSENBERG: Good afternoon, everyone. For a former longtime member of academia like myself, it's always a great pleasure to return to the campuses of institutions like MIT, even if only for a few hours and even, I might add, on an uncommonly cold day in New England.
Whatever that temperature is here today, I have a feeling that if we correct for the wind chill factor of the Republican response to the president's budget methods, it is colder in Washington than it is in Boston. In fact, the idea that so many of us have chosen to brave perhaps Boston's coldest winter day this year is testimony to the importance of what we're here to talk about-- science in the national interest.
Or put another way, the need for a national rededication to basic research. I use the word rededicate, because events of the past several years raise serious questions in my mind about our national dedication to science in general and to basic science in particular.
I am and I remain deeply concerned that our national research effort is slowly being undermined, slowly enough, in fact, to go unnoticed for too long, slowly enough that only now are the threats becoming apparent, threats to our great university centers of basic research, and threats to the future competitiveness, economic vitality, and scientific preeminence of the pharmaceutical and other American industries, which depend and have depended for so long on fundamental discoveries made here and at the other research intensive universities that together should constitute for all of us a national treasure.
Because we are meeting today at an institution renowned for its mathematical prowess, it is perhaps appropriate that I offer some numbers to begin my remarks, specifically numbers which I believe indicate the unstable and uncertain status of basic research in this country.
And I suspect these numbers are well-known to most all of you. In 1993, the last year for which we have reliable data, total US support of non-defense research and development was about 1.9% of the GDP, an amount well below those in both Germany and Japan.
Even more sobering, the amount of federal funding, which supports basic research, the engine that drives our nation's industrial enterprise totaled only 0.2% of GDP. These figures alone should be a clarion call for anyone truly concerned about sustaining a strong basic research effort.
If we believe basic research to be the venture capital of America, we must ensure that funding levels are meaningful. This investment in the future must increase if we are serious about maintaining the nation's economic prosperity.
And most importantly, this investment must increase if the American people are to continue having the benefits of improved health and quality of life that our past investment in basic research has provided the people of this country, indeed of all countries. Improving health and the quality of life. The pharmaceutical industry has accomplished much toward the universal hope of all people to live longer and healthier lives.
The American pharmaceutical industry is the world's unquestioned leader in the discovery and development of life-saving medicines and over the last 50 years has generated a flow of new drug treatments, which have prolonged life, reduced suffering, and dramatically improved the standard of health for the world's people-- antibiotics, antivirals, anti-depressants, anti-hypertensives.
Yet, everyone here knows that the pharmaceutical industry alone is not responsible for these accomplishments. Rather, it has stood on the shoulders of this country's world-class basic scientists and academia. The US medical research enterprise is comprised of a unique collaboration of government, academia, industry, not-for-profit institutes, foundations, and voluntary health agencies.
No one major participant leads the enterprise. Rather, it is characterized increasingly by partnership and interdependence. Throughout the 50 years in which the national effort has developed, government and academia's relationships have taken many forms-- unrestricted grants, research contracts, strategic alliances.
The vast bulk of federal research funds appropriated to NIH and NSF have been granted to academic investigators who have, in turn, produced both a steady stream of exciting new information and the next generation of medical scientists. Over 50% of PhD students in the life sciences in our country are supported by NIH research grants.
These doctorate-level scientists, in turn, have trained a generation of research technicians and assistants. The pharmaceutical industry where I now work has benefited enormously from the discoveries these basic researchers have made and from their laboratories serving as training grounds for young scientists.
Ideas and people flow increasingly bidirectionally between universities and corporations. And the supply of qualified scientists who gain their training within the academic research community remains absolutely critical to America maintaining its scientific and industrial preeminence in the future.
Today, the pharmaceutical industry involves itself with academia in much the same way that government does through grants, contracts and alliances. More recently, interactions between government and industry have increased through cooperative research and development agreements known as CRADAs and other technology transfer instruments.
Thus, each of the major three participants in the medical research enterprise has a remarkably large stake in what happens to the others. The pharmaceutical industry has benefited from the discoveries in academic laboratories, which have transformed our understanding of living organisms.
Industry, for its part, has used this information to explore disease mechanisms and translate these discoveries into new strategies to diagnose, treat, and even cure disease. Taxol, one of my company's newest anti-cancer products, illustrates, I believe, the distinct role played by each partner within the medical enterprise and the interaction so critical to their individual and collective success.
Taxol, as some of you may know, was finally introduced to the marketplace in 1992 for the treatment of refractory ovarian cancer and was hailed quickly as the most important anti-cancer drug in 20 years. Bristol-Myers Squibb spent hundreds of millions of dollars developing Taxol, most importantly learning how to formulate, administer, and manufacture the Taxol molecule.
No small feat, I assure you. You may recall that the active ingredient in Taxol is derived from the bark of the yew tree. And concerns over depleting yew trees in the Pacific region for Taxol supplies initially caused a great deal of concern and controversy, controversy that was put to rest by the company's ability to extract Taxol from twigs and needles rather than from tree bark.
But what is important for this discussion is to remind ourselves that the work that led to our company developing Taxol began 20 years earlier at the Einstein School of Medicine in the early '70s through grants from NIH. Subsequently, Taxol's preliminary evidence of being clinically efficacious was demonstrated by NCI sponsored trials in academic centers.
Further, Bristol-Myers Squibb finally obtained access to Taxol through a government sponsored CRADA. Taxol today is approved to treat ovarian and breast cancer. And many other studies are underway to define its value in cancers of the lung, head and neck, stomach and pancreas. We still don't know, nor will we know for a number of years the full extent of Taxol's benefits.
But thousands of cancer patients around the world have been treated successfully, because government, academia, and industry have worked in partnership to see that Taxol's promise was fulfilled. Taxol's development followed a long, uncertain, arduous path from basic science laboratory to the bedside.
The NCI sent botanists together, plant specimens that might have medicinal uses. And chemistry work revealed how this taxane prevented cells from dividing. That work led the way to studies that demonstrated Taxol's potential and ultimately to its success in clinical trials.
Our national commitment to basic research must continue to cherish such unfettered research that often yields exceptional, yet unexpected advancements like the story of Taxol. And I have another brief story to tell you. It is about hydroxyurea, a drug recently in the headlines, which epitomizes the unpredictable value of basic research.
Hydroxyurea was first marketed by my company in 1968 for the treatment of leukemia, skin cancer, and other types of malignancies. We knew that it bound to DNA. And although considered a significant new treatment, it has surely never been a major commercial success. Nothing to be called a blockbuster or anything like it.
Today, 25 years later, this compound with no longer proprietary rights was shown in studies funded by the National Heart, Lung, and Blood Institute just two weeks ago to be the first important breakthrough in treating sickle cell anemia, a common, inherited, and often debilitating blood disorder afflicting people of African descent.
The story of hydrea and sickle cell anemia is to me a wonderful example of what Peter Eisenberger of Princeton University calls, quote, "the unanticipated developments of unrestricted research, which create new opportunities that could not be previously identified," end quote. Recently, the Battelle Memorial Institute announced the results of its annual survey on US scientific research and new product development.
The survey reports that industry R&D spending will continue moving toward product development and away from basic research. And that even the most basic research conducted in industry will have an application towards satisfying specific corporate product and process needs as companies attempt to make R&D spending pay off more quickly.
I have learned in the past two years how rapidly companies make changes in their R&D budgets when the bottom line demands that they do so, something that I recommend as an informative exercise for all of you who do science.
If this trend continues, this trend of R&D spending and industry needing to have a return on investment-- and I believe it will-- the need for more basic research sponsored by publicly supported government funds and conducted by academic investigators devoid of any dedicated national objective other than new knowledge will become increasingly imperative.
Such fundamental work must continue if we are to discover the next hydroxyurea, or the next Taxol, or the next AZT, or the next cephalosporin effective against resistant Staphylococci. Today's meeting was called in part to gain input from the industrial sectors, which rely heavily on the fruits of basic research.
As the head of a pharmaceutical company's research Institute, my job is to oversee the translation of exciting new science into safe, efficacious, and cost-effective medicines. Neither my colleagues holding similar positions in other companies, nor I, can accomplish this objective without a national commitment to sustaining basic research.
Our collective responsibilities here today, all of us, are equally difficult, determining how best to foster and sustain basic research at such levels that it will thrive and continue to provide the insights that lead to important developments in industrial sectors, pharmaceutical, and others. Let me offer some initial recommendations that relate specifically to medical research.
They come largely from Research!America's 1994 consensus statement in support of medical research. First, incorporate the promotion of medical research as an integral element of any health care reform measure. Second, increase the capacity of NIH, NSF, the VA, and the CDC to support medical research by growing the budgets for these organizations.
I applaud President Clinton's 1996 budget requests for NIH and NSF and his stated goal of increasing the national capability for basic medical research. I only wish that he had boldly demanded more. Fourth, encourage the discovery and development of innovative and effective pharmaceuticals, devices, and other agents by private industry.
Fifth, ensure a stable revenue stream to academic health centers for medical research and education of medical scientists. Provide these centers with incentives for heightened interaction with industry, voluntary health agencies, and other private sector sources of medical research support.
Events of the past year suggest to me that no matter how precarious the stream of funding is at places like MIT, it is far more precarious at the major academic health centers of universities like Harvard, and Yale, and Columbia, and Duke, and Stanford, and others.
As an incurable optimist, a valuable profile, I might add, for somebody heading pharmaceutical R&D, I'm encouraged by the recent signals emanating from Washington and by what appears to be the framework for bipartisan support of science between the administration and Congress. But we must see to it that these encouraging words translate into deeds.
Setting broad national goals, which has been suggested by the administration, may be one step toward that end. But goals won't be realized if policies and mechanisms do not exist which sustain our basic science engine and promote translation of fundamental science to the clinic in the marketplace.
This process requires healthy institutions and practical, coherent federal policies, which balance short and long-term research expectations. In considering any actions today, we must bear in mind the impact they will have on the future.
Medical science truly is on the brink of discovering new insights into the fundamental workings of the human genome, and the immune system, and the fundamental mechanisms responsible for major diseases, such as cancer, Alzheimer's, AIDS, diabetes, and many other afflictions of the aged.
The promise of medical science vastly exceeds its progress to date, a statement that I and many other of my colleagues have made many times. And I will say it again. The promise of medical science vastly exceeds its progress to date.
That is not a slogan. It is not hyperbole. It is not a cliché. It is the deeply held belief that has been educated and tempered by 35 years of direct involvement with a marvelous thing called the American medical research enterprise. Thank you.
KNOWLES: Thank you very much. We're going to move to the other half of life, as it were, of our live segment, before we have any discussion of Dr. Rosenberg's paper. So let me introduce with great pleasure, Phil Sharp. We come to a bunch-- a sequence in the program of chemists, or that is to say, ex chemists.
Professor Sharp got his PhD in chemistry and his AB, for that matter-- his PhD from the University of Illinois in Urbana. He then began a mild transformation by going to Caltech and then to Cold Spring Harbor before coming here in 1974 as a biologist, a not uncommon root for chemists to follow, in fact.
Since 1991, he's been head of biology at MIT. As you all know, he discovered independently in 1977 that you can hire genes that split. And there began an immense effort on RNA splicing for which he was awarded the Nobel Prize two years ago. That's his academic credential. What about-- because I wanted to see legs for him.
He is the co-founder and has been for many, many years the chairman of the scientific board of Biogen, a not undistinguished local company. And most recently, for his third leg, he is now appointed a member of the President's Advisory Council on Science and Technology. Phil?
SHARP: Thank you, Jeremy. It's a pleasure to follow Lee in this program. And I was particularly pleased by his passionate defense of research and basic research. And what I'll try to do in the next few moments is just expand on some of the opportunities in basic research related to life science and then add a little detail to those points.
But before I begin, I want to make a point about life science and basic research. The issue is the following. It is very difficult to define what is basic research in life sciences. For example, is the discovery of the first oncogene a discovery in basic research or a discovery in applied research?
It certainly was motivated by a basic problem trying to understand the genetics of human cancer. And its discovery will have direct implication and has had direct impact on the treatment and diagnosis of human cancer. So in that regard, it is applied in a sense.
But similarly, the discovery of the first oncogene, which is a mutant form of a natural gene that causes tumor cells to grow, was a fundamental advance in our understanding of the regulation of cell growth and division. So in a biological sense, it certainly is a basic discovery and would be defined as basic research.
Now, to my dean who's a physicist, all of this is applied research. Because we're not advancing the understanding of the physical laws of nature. And to a clinician who is interested in treatment and care of patients, clearly the isolation of an oncogene in a laboratory related to cell growth is basic research.
Because tomorrow, it does not help that clinician cure the sick or treat the patient. Thus, when we talk about life sciences and we discuss this issue, a basic and applied research, I prefer to call it research. Because it is terribly difficult to know when one in this science is moving from one agenda to another.
Now, the case for supporting research is terribly strong, particularly in life sciences. And Leon has commented on that before me. But I'll broaden what Leon's comments are, beyond that of medical science. If you look at the great challenges facing this country, one is providing a good health for its citizens at a reasonable cost.
Another is in generation of a growing economy. And the third is maintaining an attractive and clean environment. And if you think about each of these three, all three are directly related to issues of life sciences. Thus, building the science and technology base in life sciences is critical for the future of this country in terms of both public health, economy, and the environment.
Now, I'll come back to some opportunities in life science and applications of basic research. But before I do that-- or research-- I want to comment on just how little we know now in terms of life sciences. We have seriously studied the fundamental nature of biological systems for a little more than 40 years.
I date the beginning at the discovery of the structure of DNA, which was done in 1953. It wasn't until the recombinant DNA era, which is only about 20 years ago, that we were able to isolate and manipulate genes, particularly those genes from human cells.
So until this resolution, the revolution, the recombinant DNA era, the biochemical realm of human cells was largely unexplored. Today, we're in the process of discovering new genes and new protein structures almost at a daily pace.
However, we know very little about how these cellular components interact and how they execute their chemical functions. This integration step of integrating our knowledge of the structure of genes and protein structure back into how a system works is something that we're still struggling with in terms of advances in research.
There are many fundamental questions that one can phrase that illustrates our lack of knowledge now. For example, we don't know how a protein folds into its native structure, even though we've known for decades now that that structure is critical for its activity.
We do not know how many genes are important for the formation and maintenance of man. Do we have 30,000 genes in our chromosome? 50,000 genes? 100,000 genes? We don't know yet. Finally, we know very little about how the brain works and how memory is recorded in our brain.
These are questions that lie before us as well as many others. So we can easily get overwhelmed with the flow of information and advances in life science research, but we have many things to learn yet. The development of a recombinant DNA actually brought forward the biotechnology industry.
And many novel therapies and treatments have emerged from that industry. And at this moment, about 100,000 people across the country in some way or another or employed in this sector. It is still a growing area of technology, which is becoming more diverse with new discoveries and new societal needs almost daily.
In fact, the rate of change of technology and science in the realm of biotechnology is staggering. Who would have suspected a few decades ago that the nation would be focused on a DNA fingerprinting evidence in a murder case, or struggling with the issue of an epidemic of a new virus, the AIDS virus in this society, or planning the sequencing of the total human genome?
Biotechnology is an activity that depends on advances in research and is one that is still rich with discoveries and new applications. Now, one of the societal concerns I mentioned before was the improvement of public health as well as the containment of cost of providing that public health.
There are two opportunities to simultaneously improve health as well as contain cost. The first of these are prevention. In most cases, preventing a disease is clearly cheaper than carrying a disease. However, many diseases-- and at this stage, the largest number of them cannot probably be cured.
They're probably related to natural processes that are inherent to us as biological organisms. A second opportunity is the development of more effective treatments. And this depends critically on advances in research. Better treatments primarily arise from the development of better drugs that reduce damage due to disease or drugs that cure disease.
More importantly, better treatments and diagnostic methods is the only hope of improving the quality of health care in the future as well as maintaining the cost of health care. The quality of health in an environment of constrained costs, then, depends critically upon research and innovation.
For example, the most successful biotechnology product produced today is Epogen, a drug developed by Amgen, which stimulates the production of red blood cells. This protein and similar blood factors are now making it possible to do bone marrow transplants on an outpatient basis without having patients hospitalized for weeks to months in terms of a regeneration of the bone marrow system.
This will allow more effective and intense treatment of cancer patients, which will be more effective in curing them, as well as reducing the total cost of such clinical protocols by reducing hospitalization costs. Now, at the moment, that's terribly expensive.
But 10 years from now and 20 years from now, it'll be routine. And it'll be something that's within the realm of all medical centers. The process of developing new pharmaceutical agents is beginning to rapidly change. And I think Leon referred to this.
At one time, large pharmaceutical companies almost exclusively developed new drugs by screening natural products. And I would comment that Taxol is an example of this. This worked reasonably well, but it did not provide a learning curve to make the development of future drugs easier.
The era of modern biological scientists is beginning to change this tradition in terms of drug discovery. The inner workings of a human cell, the nature of mutations that cause disease, the ability to design animal models that emulate specific human diseases is revolutionarily changing how the pharmaceutical companies and discoveries work.
We are only at the beginning of this revolution. These advances in cell biology are complementary to advances, both in chemistry in terms of combinatorial libraries that allow the sampling of large numbers, billions of different chemical structures and in structural biology, where one can determine the atomic structure of very complicated macromolecules with ease and with some rapidity.
All of these technologies will merge in the production of a new pharmaceutical industry producing new drugs with better treatment and better quality health care. There is no area of life science that offers more opportunity for research to advance public health than that of neuroscience in my opinion.
It is commonly agreed that about half of all health care costs are related to mental disease in some form. With a large, aging population, these costs will only increase. Dementias such as Alzheimer disease are not yet understood.
And therefore, there is no effective treatment to either delay the onset or effectively treat the symptoms. This can be said for almost all medical health problems. The science of the brain and the processes can now be approached with vast new arrays of technologies.
These include powerful methods to identify human genetic diseases and the genes that are responsible for them related to mental diseases. It includes the ability to manipulate mice as a model system to understand the workings of the brain. And it also includes advances in our understanding of the cell biology of neuronal cells, which the brain is composed of.
And all of these come together now to offer rich opportunities in the area of neuroscience for advances that will impact on public health. In my comments, I have tried to stress the riches of opportunities that lie before us in research in life sciences.
This research is in a national interest as it directly provides a basis for improving the quality of health care and for future uses of biological systems to produce materials, and agricultural products, and other useful applications. In addition to these opportunities in new products, the environment impacted by our technology, as I mentioned before, is largely a biological driven issue.
Thus, biological scientists have a lot to offer. And the relationship between industry and MIT, or industry and the basic science community is one that has enriched the transfer and development of the science that I have discussed into more practical and more worldly uses. Thank you.
KNOWLES: If Dr. Rosenberg could come and sit down, our two life scientists are available for your questions or comments. Please. Yes?
AUDIENCE: I wanted to ask both Phil and Lee a question. You talk about the importance. And obviously, I think that this particular audience-- no one would dispute the importance of the basic life science base for the future. In the case of medical sciences, we've got a corollary professional school structure-- the medical school structure and the medical profession.
When you talk about the need to have more life sciences entering into things like environmentally [INAUDIBLE] et cetera, would you-- is there anything you'd like to say about how one might think about providing the incentives in undergraduate education that would lay the groundwork for drawing students into those areas?
What I'm trying to say here is that the match between the disciplinary base, and the way we train graduate students, and the need that we may have in that area is not good. It's easy in the medical case, because we've got the medical schools and medical problems.
But certainly, that's been a great concern in the environmental technologies area and in the [INAUDIBLE] environment and natural resources to talk. This has been a stumbling block. Where do these people come from, who have the expertise to integrate that material and provide them in the basis for the future?
SHARP: In this, I think MIT has been innovative. And the reason that I say that is the following. MIT-- about five years ago-- and then affected about two years ago-- made the decision that all students who obtain a degree at an undergraduate level at this institution must take a course in biological science.
And the impact of that primarily is going to be in broadening the technology and science base of those students who go into civil engineering, and mechanical engineering, and even into physics and economics in terms of how biological systems work and what we know and don't know about them.
And what we're seeing across the campus now is something I call the greening of MIT in that we're seeing increased interest in relationships between biological sciences and engineering, both in chemical, and mechanical, and in civil, and in other areas of technology.
So I think at an undergraduate level, a breadth of education, particularly in terms of science and technology, provides a student for the opportunity, then, to-- or the background to move into environmental and more interdisciplinary activities.
KNOWLES: Frank Rhodes raised this question of scientific literacy this morning. Do you want to say anything from your Yale days, Leon, about this problem?
ROSENBERG: Well, I was going to just pick up on Marcy's question in that regard. Because I think would say, Marcy, that if we are going to find exciting people examining questions about environmental science, they are going to have to understand that environmental science is, after all, science.
And therefore, it comes back to the issue of whether our people do, in fact, have an appreciation of the principles of science. And I think that's an enormous challenge for everyone who cares about science in this country. We have not done a very good job of extending scientific literacy to the population.
All you have to do is be a person who, like I do from time to time, go around and try to talk about the importance of medical research to lay audiences. And you will find out just exactly how few people understand even the words that you use.
I think it has been estimated that only 5% of people in the United States know what the initials NIH stand for. In my company, they stand for not invented here. But somewhere else, they ought to stand for the National Institutes of Health.
And I think it is amazing how far we have to go in education in the elementary schools. I used to say to doctors that I used to lecture to, about genetics that my 13-year-old son at that time knew more genetics than I was sure all of the members of my audience did.
And that was not a put-down of the practicing physicians. It was simply a statement of what Phil said earlier, that the world of genetics began after most of the people who were practicing medicine up until very recently had completed their education.
That's something dramatic that we all have to understand, the evolution of the life sciences and the rapidity. When I was a first-year medical student, DNA was dismissed in my biochemistry lecture in 1953, just months before Watson and Crick's paper as a macromolecule of undefined structure and function.
SHARP: Two sentences.
KNOWLES: We must be careful not with these questions to take from the next session on which, after all, is education and suddenly becoming sensitive. More questions, please.
AUDIENCE: [INAUDIBLE] the economics of bioscience and technology with respect to ownership of intellectual property, and the cost of research, and the benefits from government support and so forth.
SHARP: Repeat the question. This is a good one for you.
KNOWLES: It's your kind of question.
ROSENBERG: --it sounded like one that might have been directed either to somebody from Biogen or somebody from Bristol-Meyers Squibb. But I take the question to mean, what is this biomedical enterprise as it relates to proprietary rights, as it relates to cost of research?
I can only say this much. The quest for new treatments or new diagnostic agents simply demands an industry a possibility of obtaining a return on investment. I carry around with me some of these awful numbers, like the average cost of developing a market and pharmaceutical is $250 million.
For every compound that comes to market, there are 5,000 chemicals that begin study in a discovery laboratory in academia or industry, that even those compounds that come into development in a company like ours, one in seven that find their way into animal studies make it to the marketplace.
Even those that make it into clinical study-- one in five in early clinical studies make it all the way through. And even in the latest clinical phases, that is those in which you are doing randomized double blind studies, so-called phase 3, only 50% of those things that make it to phase 3 make it to the marketplace. This is a high stakes gamble. There has to be a return on those things that make it.
Because if there isn't, then companies simply will not put the billion dollar investments, which characterize a company like mine, or Merck, or Pfizer, or a lot of the large pharmaceutical companies, or the Amgens, and Kyrons, and Biogens of the world. Because there simply will not be any way to justify it to shareholders-- the amount of money that you put in.
SHARP: I want to add just one little touch to that, and that's the following. One of the things that I have learned in my years of interacting with Bristol-Myers Squibb, or Biogen, or other companies is that this type of drug development cannot be done in the academic sector. It's just a scale of focus and commitment that you cannot mount in a public sector.
It has to be private sector activity. So if the country wants new drugs to improve the quality of life, then in some mechanism, they're going to have to fund it. And at this stage-- and I agree it's the best way-- is funded through proprietary patents and proprietary positions, which then allow the development of new drugs.
And it's almost a method of generating resources for further investment in society for further development. And it has historically, over the last 50 years, given us a vastly improved quality of health and one that I think we all appreciate.
KNOWLES: If the question was more deeply uneasy about intellectual property, which I thought it might be, I'm sure that should go on interestingly in the break. Other questions? There was no unease. Well, thank you both very much.
We are being reasonably obedient and good about time, so that there will be a break. And there will be time, I trust, to discuss some of these more interesting-- still interesting items. Let me move on now from the life sciences to the physical sciences. And our first speaker in the physical sciences is a PhD chemist from Brown. John McTague got his PhD in 1965.
And first, on a slight detour-- on an interesting first job, not a detour at all, of course, went to the North American Rockwell Science Center in Los Angeles. In 1970, he came-- he joined the chemistry faculty of UCLA. But by 1983, he'd become deputy director of the Office of Science and Technology Policy.
Of course, in Washington, he was acting science advisor to President Reagan in the mid-'80s and a member of President Bush's Advisory Council on Science and Technology. Now, we come to his third unambiguous third leg. Because in 1990 and since 1990, he has been the vice president for technical affairs for the Ford Motor Company, a splendidly eclectic set of activities in his career. Dr. McTague?
MCTAGUE: Thank you. I got a letter a little while ago from Chuck Vest with an assignment for today. And the assignment was to discuss, quote, "the evolving nature of physical science in industry and concomitant changes in industry-university relationships.
He didn't ask for this in 25 words or less. He asked for it in 20 minutes or less. However, that gives you two options. One of them is to have a extremely shallow overview or talk very fast. Maybe three options then. Or as the mathematician responded when he was asked to describe a cow, consider a one-dimensional cow.
That's the tact I'm going to take. I'm going to give a simplified version of a limited set of aspects in a subset of industries. In other words, I choose a lower dimensional cow. It's pretty clear from what we've heard already that science-based industries have a very long history. We heard earlier from Joe Miller about the fact that DuPont is almost 200 years old.
And certainly for at least 3/4 of that time, the chemical industry has had a very strong science base going back to the days of synthetic dyes. And then in this century, the science base that has strongly driven the development of polymers and very close relationships with the universities over that whole time.
As Jeremy mentioned to you, I was a graduate student slightly south of here at Brown University in the early '60s. And I can remember once wandering through the stacks of the library there and discovering a catalog from the early 1920s in which the name of the chemistry department, in fact, was the department of wool chemistry.
So early on, well before the days of government funding, there was a very close relationship between the academic establishment and the chemical industries, both in the United States and in Europe. We've seen developments of science-based industries in other fields, too.
An obvious example was the evolution from-- let me call it truly Edisonian days in the electrical industry to the more science-base days of the early days of the research labs at General Electric with Langmuir, among others, and of Westinghouse to today's microelectronics industry, which is extremely, strongly science based and has had very close ties to particularly materials developments that have occurred in universities.
Now, I want to get to what I really want to talk about, which is my own industry, which in contrast, perhaps, to the biotechnology life sciences industries, is in most people's minds, most accurately described as a Neanderthal industry.
Or if you want to be charitable, a Rust Belt industry. Off the top of your head, you wouldn't tend to think of an industry like this as having much science base. And over much of its history, that's probably true. The automotive industry for practical purposes is about half the age of DuPont.
It started off around the turn of the century, really, with a bunch of tinkerers, truly Edisonian tinkers, mostly people with no or little formal education, but with a great enthusiasm to experiment and to make vehicles literally in their garages. The big transformation of the industry occurred in the early part of this century with the introduction of mass manufacturing.
And indeed, ever since then, the great strength of the automotive industry and transforming life around the globe has been the genius of being able to make extremely complex consumer products extremely rapidly and at an extremely low cost.
But there was very little true technical innovation that occurred in this industry between roughly 1915 or so, after carburetors had been invented. And actually, even the first aluminum engine was built before that, until the period following World War II.
Even after World War II, the industry was sort of in a drift, I would say. So then for almost 50 years, essentially nothing happened in this industry. During that time period, the involvement of scientists or indeed of formally educated engineers in the industry was minimal.
Up until the period after World War II, there were essentially no formally educated engineers in the automotive industry. Product engineers rose from being draftsman. And manufacturing engineers basically served apprenticeships on the floor and just took over larger and larger components of manufacturing.
Right after World War II, though, formally trained product engineers started moving into the business. And about 20 years ago, manufacturing engineers started having formal educations. But it's only been probably in the last 10 years that more than 75% of the manufacturing engineers even had a degree.
All this occurred during a period of enormous prosperity for this industry, just going along, going along, raking in the money, not having to do anything else. It must have been a very pleasant life. I wish I had been there. But then in the late '60s and the early '70s, an absolute revolution occurred driven by external forces.
The first force probably was environmental awareness when people found out the causes of smog in the Los Angeles base-- and scientists did. And the general populace understanding the fragility of the environment basically going back to Rachel Carson.
And then in the early '70s, the oil shock, which made people aware of the fact that their vehicles we're actually burning gasoline, and that gasoline was not free anymore, and not necessarily freely available. These two shocks completely threw the industry off its pedestal.
And it had to respond extremely rapidly. How did it respond extremely rapidly? Well, fortunately, what had occurred about 10 years before that was something that occurred in many industries in the United States not unconnected with Vannevar Bush's report.
And that was setting up research laboratories. In those days, just about every respectable or even not respectable industry, so long as it was large, set up a, quote, "fundamental research laboratory." Just sort of had it there, just like most large industries supported the symphony orchestra.
Fortunately, when the shock of the oil problems and of the environment came about in the automotive industry, there were these people who understood science. And all of a sudden, they became relevant to their corporations, understanding combustion and controlling it, understanding emissions.
Being able to make engines that had very high fuel economy became extremely important. Bases behind these were sciences related to catalysis, for example. Surface science. A lot of work necessary on lightweight materials to make vehicles more efficient.
The necessity for computer controls for the first time in order, both to control emissions and fuel economy. The need to have science-based sensors to be able to measure on the fly the amount of oxygen going into cylinders, the amount coming out-- carbon dioxide.
The ability, then, as safety became important to be able to sense for torque, for example, for traction control. Accelerometers to control airbags, et cetera. So various aspects of science started becoming important to this industry, but not in the same way as it did, let's say, to the chemical industry.
Because the chemical industry had or developed along with it an academic discipline. There is no real academic discipline of the automobile, because like the airplane, it involves basically the integration into a system of a large number of sciences and technologies, yet it is becoming more and more a science-based industry.
Two things have happened as a result of that. One of them is that we realized that we can't do it all. Over the past several decades as both the demands-- the environmental, the fuel economy, and the competitive demands have increased, we have very substantially increased our R&D.
It's growing at roughly twice the rate of industrial R&D in the United States. But even with that growth-- I mean, for example, at Ford now, the R&D budget last year was a bit over $5 billion, which is a substantial amount of money.
That's roughly $1 in every $16 that's spent by industry on research in just one corporation. We obviously can't do it all, since we are basically systems integrators of just about any kind of technology that you can think of, be it electronics, or materials, or chemical.
So there've been a lot of experiments going on over the past decades, trying to figure out how we can work best with suppliers, how we can work best with universities, how we can work best with national laboratories, and even how we can work best with our toughest competitors.
As a result of that, the three US-based automobile companies-- Ford, General Motors, and Chrysler in 1988 started doing formal joint research together. And that result of that was the creation of what's now known as USCAR, which has a several hundred million dollar joint research sharing effort.
It has resulted in-- as Jack Gibbons pointed out this morning, in collaborative work with the federal government in an area where we have parallel interests. And that is in the area of fuel economy, energy efficiency, and manufacturing capability.
That's in the partnership for a new generation of vehicles. We're still evolving how to work together in that area. But that is one area where there is a substantive joint agreement on how the federal government and industry should collaborate. And it's a relatively large-scale collaboration.
Our relationships with industries, with universities have been evolving, but not in as easily categorized way as our relationship with our competitors or with the federal government. It's a very delicate balance that we have to strike.
We have a lot of interest in what you might call fairly applied kinds of research. Yet, we know that what we most want from universities is, in fact, fairly unfettered exploration into areas where the results are unpredictable. And we also want very, very good students.
So we've been fairly cautious about trying to work with universities in a more directed manner, yet we do want to maintain ties and make sure that the whole somehow is greater than the sum of the parts. One of the approaches to this was a recent meeting down in New Orleans that Peter Eisenberger, who's here, championed.
And that was sponsored by the Department of Energy and the National Science Foundation where people from the national labs, and from industry, and from universities got together to discuss general research priority areas relevant to where the automotive industry is going. The automotive industry, by the way, is a huge industry. And it affects about one job out of seven in the United States.
Something else has happened, though, that I find extremely remarkable and intriguing. As the industry has become more science based because of external exigencies, we find although it is not a discipline-based industry, the importance of researchers and of scientists has become considerably expanded.
When I came to Ford in 1986, except for the person heading the research lab, which was me then, there was one other individual in the company who was a corporate officer, who had a doctorate out of 40 corporate officers.
Nowadays, out of the 40 corporate officers, 5 of them have doctorates. Out of 40 corporate officers, 5 of them have come out of the research laboratory. Now, this didn't happen because all of a sudden, the corporation decided, gee, isn't science wonderful?
It happened as a result of the crises of the early '70s when the company was confronted with a series of very complex, technical demands that it didn't know how to handle. And the fastest, most nimble people, the ones who were able to grasp problems at the fundamental level most rapidly, indeed, had come from the research laboratory.
So the payoff from having the cultural equivalent of the symphony orchestra within the corporation was something which was extremely practical. And that was truly exceptional individuals. The net result of that today is that although the research laboratory represents only 1% of the professional workforce at Ford, it represents one person out of every eight of the corporate officers.
And they're not distributed in the way that you might expect. The most obvious one, I suppose, is myself who's in charge of what's called technical affairs, which is research, and environment, and safety, and other things worldwide, professional technical development for all the scientists and engineers in the company. All 17,000.
That's a pretty obvious kind of thing you'd expect for someone coming out of a technical background. And also, the vice president for what's called advanced vehicle technology, where all the vehicle programs in the company arise, happens to be a PhD physicist who came out of the research lab. But that's a technical area.
There are some other areas that are not so obvious. The vice president for customer service is a PhD chemist. The vice president for new business strategy is a PhD electrical engineer who did not start his career in the research lab, but spent much of his early career there.
But even farther upstream in the automotive business at Ford, which last year was 1.07 times 10 to the 11th dollars. Since it's a science-based business now, I have to put it in scientific notation. That's big money, by the way.
The head of all manufacturing for the world, Bob [INAUDIBLE], the group vice president for manufacturing, started off his career in the research laboratory. The head of all of the automotive activity for the whole company, that whole 1.07 times 10 to the 11th dollars is-- I'm sorry to say this in front of Alan Bromley-- a PhD nuclear physicist.
Pretty obvious, isn't it? Should be a PhD nuclear physicist who started his career in the research laboratory actually doing non-linear optics. And one step higher up, even-- there are two people who are higher up than the head of all of Ford automotive. That's the chairman and the vice chairman. The chairman is not a nuclear physicist.
However, the vice chairman who is the only other member of the company, who was on the board of directors is Lou Ross, who started off his career in the research laboratory. So the moral of the story that I wanted to draw from that is that one of the most important things about science-- and this will lead up to discussions later on this afternoon-- isn't necessarily the actual technical content of the science.
But it's the training and the education that researchers and research scientists receive, which enable them to handle on a very fundamental basis questions that are not understood and to do it in a timely manner. And the remarkable transformation that has occurred as the automotive industry has become more science based is nowhere more obvious than in the science base of its management.
And I think this has implications probably for other industries as they evolve, and that it also has implications for the importance of very solid research experiences in universities. But it also suggests opportunities for researchers and students in particular in the basic sciences at universities to start thinking a little bit more broadly about potential career opportunities. Thank you.
KNOWLES: Thank you very much, John. There is only one person more pleased than any of us that George Whitesides is here tonight-- to hear this afternoon. That's Mark Righton, who as you all know, was otherwise going to have had to try the impossible, which is to pretend to be George. Professor Whitesides started his career at Harvard.
And he's come back there after some detours, first to Caltech to get his PhD in physical organic chemistry. He came back from there to the faculty here at MIT and moved slightly upriver in 1982, where he is now [INAUDIBLE] professor at Harvard. George has served on more-- serves on more scientific advisory boards than most of us, as they used to say, have hot breakfast.
He goes from cars to perfume and from magnetism to biotechnology in his scientific contacts with industry. As far as the government is concerned, he sits on committees of the National Research Council, the National Science Foundation, the Department of Defense, and the Department of Commerce. With the greatest of pleasure, George.
WHITESIDES: We'll see how this works. My various colleagues who have come here, I think, to tell you useful things. I'm coming here, as you can tell, as an example of an area in which there's excellent basic research to be done. And that's virology.
Pathophysiology is a place where I think there's some good applied research to be done. What I would like to do is to give you my message as far as I can up front, because I'm not sure there is a later in this discussion. The message is that in fundamental research, there are some truly wonderful opportunities.
And I think from the academic point of view, many of these opportunities fall between the conventional disciplines. I think that in these areas, you can find splendid applications that will create jobs, and goods, and services. And given these facts, what's the problem? Why are we here?
And the answer to that is, I'm not quite sure. I think part of the problem is that we've fallen into a semantic trap in which we are spending quite a lot of time worrying about the difference between the fundamental and applied and probably not enough time going about the business of trying to make the whole system work better.
Let me spend a few minutes discussing some of the examples, which I would pick out as being good examples of basic research, areas in which there's good basic research, in which there also a splendid applications to go along with it. A way of analyzing research is to think about extremes. That is to think about what the frontiers are in science at any given point.
And what I'd like to do is pick out an example or two of things that are very small, things that have new properties, things that are very large, something to do with the Earth and the environment and just point out how the applied and pure kinds of distinctions probably don't, at least to me, make a great deal of sense.
In the area of small, for example, there are now wonderful, wonderful tools which enable us to look at distances that are in the order of atoms or fractions of atoms. This opens from the point of view of fundamental, condensed matter of physics the opportunities to examine things like quantum devices and quantum phenomena at the scales of movements of electrons.
Even if these devices themselves never make it into the real world, they will certainly enable new types of device fabrication, which in turn will make it possible to make portable telecommunication systems with very low power and offer the opportunity to change our communication structure.
So it's a kind of investment from a national point of view in which one almost can't lose as far as I can see. In terms of properties, throughout material science now, one is finding new kinds of materials being developed. One example is things which have the general composition of carbon nitride, which may have the property that they are harder than carbon, now harder than diamond.
The question of how you measure something that's harder than diamond is a little problematic, but we'll figure out how to do that. What does one do with that kind of technology? And the answer is you try to solve some of John McTague's problems.
That is, you apply it to extremely tough materials with extremely lightweight that will be the sorts of materials that are required to meet the performance and weight requirements for the ultra-low emission vehicles that the automobile industry is now facing.
Large. One of my favorite experiments now is one that's going on at MIT and also at Caltech. And that's the LIGO experiment, the gravity interferometer experiment to measure gravity waves. From the point of view of fundamental science, I can't think of anything more exciting than listening to the dying scream of a neutron star as it falls into a black hole a million years ago.
I mean, I think this is terrific stuff, far better than OJ. At the same time, this is producing technology which I believe will ultimately revolutionize metrics. What these people are doing in terms of their measurements simply takes one's breath away. And it's going to be useful for many things.
In terms of the Earth, we are using in a breathtakingly clever way the combination of ground-based sensors and space-based sensors to watch the whole planet shift and move, a wonderful combination of computation and sensing at a very fundamental level.
The application, of course, is to try to put off happening in San Francisco or at least warn San Francisco before it becomes Kobe. And this is no small stuff if you live in San Francisco or Kobe. And then finally, one of my favorite areas is this issue of the environment. And we all are very interested in the environment.
I fall a little bit in that category of people who believe the sort of Gaia hypothesis that there may be some-- at least self-regulating quality to the environment. And I thought for years that the way that Earth-- I was thinking of it as an organism might solve the problem of having too many pestilential human beings around was simply to turn up the temperature and cook us.
But the nice thing that's happened now is we have a second hypothesis, which is this interesting issue of estrogen equivalence, which is that many, many chemicals may interact with the estrogen receptor with the result that all of us males are sterilized.
And there is a very intriguing discussion going on now, which I think is good public policy of global male sperm count. What is the quality of the global male sperm count? That's a terrific, terrific discussion to have. So what does one do with this whole story at the moment?
It seems to me that I look at this combination of basic research and applied research. And I'm deeply convinced that one can have-- we can have our cake and eat it at the same time. We can do both basic research, which is very, very good applied research. So how do we go about it?
I think for this audience, I would make just three suggestions. And they're in the order of three, I suppose, criticisms. There are three parties in this story. There's the university. There's government. And there's industry. And each of us has done some things very well and has failed in some things.
From the universities' point of view, I think what we have done is to fail to explain what it is we do in terms that connects the fundamental to the applied. And we have tried to fuss too much about this semantic issue of the difference between fundamental and applied.
We should just forget it. It is not that big a deal. Just do good research, as Phil Sharp said, and make the point that we'll try to be helpful where we can. And I think in practice, we can also do probably a much better job of moving fundamental research to the point where it's prototyped or prototypable and can be understood by industry with its present product focus.
So we can help in both of those areas. From the point of view of the government, I would-- I think make two suggestions. I know trying to make suggestions-- the government is not a very profitable thing to do, but I'll make them anyway. I would criticize the government for inconsistency in the way it thinks about this problem.
The university responds to changes very, very sluggishly. After all, the lifetime of a student from entering undergraduate school to leaving graduate school is maybe 9 or 10 years. And if the problems on which we're working are changing every two years, we have a hard time keeping up.
I think if there were a clear national consensus led by the government as there were-- it was for many years as to what the important problems were. We could be very helpful. For years, we understood it was high-tech medical care, high-tech health care, and national security. And the university system was extraordinarily good at working on problems in those areas.
Right now, there's a bit of switching around. And the switching is going on a little bit more rapidly than we can keep up with. So for the government, I would say try to be consistent in the objectives and try not to use the distinction between fundamental and applied as an ideological distinguishing feature between the parties.
That doesn't help any of us. And then, actually, my most serious criticism as say for the last party, which is industry. Because I think there is something that industry can do that would be very, very helpful and which will cost almost nothing.
So it's serious in the sense that I don't understand why it isn't done. What industry can do is to encourage the senior management, the folks that John McTague was talking about, just to explain exactly what John said in public in Washington.
I mean, it's remarkable to me that industry, as it moves away from basic research internally, will say that we rely more and more on basic research from somewhere else. But the senior management appears and talks about tax policy, or about GATT, or about things of that sort. These are very important topics. I don't disagree with that at all.
But the notion of having industry be the bridge between the university system and the users in industry with the encouragement to government to provide support in the right fashion is something that I don't think would take a great deal of effort on the part of the senior management and would have enormous impact. So let me summarize by finishing and finishing by saying, as I said at the beginning, that the opportunities are there. The societal needs are there.
The distinction between basic and applied is purely artificial. It's very, very hard to find most good research being done in universities not traceable to some useful and interesting thing that appears in the real world in some fashion. And I think that we would all benefit by working hard to obliterate the distinction between applied and fundamental, rather than encouraging a debate as to the relative merits of applied and fundamental. Thank you.
KNOWLES: George, thank you for that heroism. And we'll see if your voice runs out, I'll ask Mark to come and sit at the table to give the answer you would have given had you been able.
Questions? Comments? Please.
AUDIENCE: I didn't quite get the sign of your-- I believe November to be consistent. The time scale of changing problems [INAUDIBLE]. The time scale of the excluded passing through is seven years or nine years. It's 12 years for certain [INAUDIBLE]. Were you advising that the government relax and adopt the nine-year time scale or that universities get nervous and adopt the two-year time scale?
WHITESIDES: What I was suggesting is that the government try to focus on problems which are sufficiently important that they're going to be around for a long while. Those are the kinds of problems that we can help with. That's right.
AUDIENCE: I was afraid that you were going try to [INAUDIBLE] on the feedback.
WHITESIDES: That's like trying to teach a hippopotamus to tap dance.
KNOWLES: And he speaks with experience.
WHITESIDES: I know about that.
AUDIENCE: I guess we've talked about the wonders of basic research-- or research is-- I agree with George. We should talk about-- but it seems to me the reason we're here is because the society has a finite amount of resources. And it doesn't do good enough to talk about the good things one can do.
One has to talk about how we can do them more effectively than what is in the past. Because that's what all sectors of the society are being asked to do. So I'd be interested to speak his thoughts on how we could accomplish our good tasks more effectively than we have done in the past.
MCTAGUE: Well, part of it, I think, was the issue I was trying to get at earlier. And that is this country wastes an enormous amount of resource on litigiousness. It does that in the obvious area of civil trials, et cetera. But it also does it in the area that we're talking about here.
The adversarial relationship that exists between the various elements of our society and in the technical arena really hurts us. Environmental regulation, for example. Regulations are arrived at by an adversarial proceeding, believe it or not. So-called reg negs. People get in the room and argue with each other. And finally, somebody gets tired.
They're not done on a science base, or on a risk assessment base, or an economic assessment base. They're done on the basis of how two lawyers would argue a case. What has happened in Superfund, for example, is a classic example of what's going wrong. I just noticed Nick Samios shaking his head, actually shaking his head yes. He always shakes his head no to me.
The problem of how we run our national labs is a disgrace-- absolute disgrace wasting an enormous amount of resource that could be aimed at a technical purpose, instead of these silly response to my request or my demand for this report by October 15. If we could find a way to agree that we're all in this boat together at some level, there is an enormous resource that we can free up that could be applied toward the tasks at hand.
KNOWLES: Certainly, George's three points-- were each of them, as it were, educational in some sense? That people-- that neither government, industry, nor academia were clarifying their objectives and certainly not transmitting them. Is that fair, George?
WHITESIDES: That's a point. And then the other point-- I would precisely support what John just said. If we, in fact, agree that we are all in it together, then I think the resource is there, the manpower is there. The question is, how does one most effectively find out that we are all in it together?
And if I had probably a single suggestion to make, it would be to work much harder to find ways of moving people back and forth between the sectors. Because I don't know how you really find out what's going on somewhere else, other than to go there. And then among these three sectors-- the university, and government, and industry-- one might as well be talking in many cases about China, Nigeria, and Madison Avenue. It's just the cultures are very, very far apart.
AUDIENCE: Look, It seems to me, we talk about communications, public relations, working better together, but are not really digging in deeper and asking serious questions, which I believe industry has done when they look at themselves of, how we can do this research enterprise better? Really make changes that may have some bite into them, some change. I get a feeling somehow if we-- the impression we're giving is if we talk better, if we're all friends with each other, that we'll solve all the problems that exist.
And I just don't really believe that that's the case. I think there are really resource questions, lots more to be done, ways of making choices, ways of doing things in the education system better than we have done in the past that require real change, not just relabeling and not just better advertisement or communication.
MCTAGUE: I think part of that is true. And one of the things I was trying to discuss earlier was the necessity for actually working together, and that you can get twofers. I know, certainly, when Ford, General Motors, and Chrysler work together, it's not because we love each other. We will kill for one point of market share. And probably that has happened in the past.
Yet, we do find it in each in our self-interest to work together in a circumscribed set of circumstances which are not transparent and which are not obvious to the customer that are not customer differentiators, where it gives us the capability of doing things that we couldn't otherwise do with our limited resources. We're doing the same thing with the government.
The government is doing it with us, not because they love us, either. So there are efficiencies that can be gained by trying to find ways to work together. But I truly believe that there are much bigger efficiencies. And the bigger efficiencies are in these terribly dissipative, adversarial processes that our society-- our autoimmune response of our society has been producing.
KNOWLES: Good. I think probably we better move on without wishing to move on. While we're still our kinder and gentler body, I hope that our questioner can sharpen up our two recipients during the break. But thank you both.
We now have, maybe appropriately, the final-- I mean, the final section on information and on information transmission, after all. The first speaker is William Brinkman. We leave the chemists now. He did a PhD at the University of Missouri in physics and joined Bell Labs almost immediately in 1966, having done-- and this is the sort of quasi-academic activity, a post-doctoral year in Oxford. I call that quasi-academic.
Then by 1981, he was director of the physical research lab at Bell. And he then took what one might call a quasi-governmental stint by three years at Sandia before returning to Bell in 1987, where he is now physical sciences research vice president. He's responsible for all the physical sciences, optical electronic devices, optical fibers, and almost everything else. Dr. Brinkman. Where is he? There he is.
BRINKMAN: Well, I have to tell you that I about panicked when I got the program and discovered I was supposed to talk about information and communication sciences. I am a physicist. I run the physical sciences part of Bell Laboratories and know nothing really about information communication sciences.
But I decided that that was okay, that what I would do today was talk about AT&T. After all, it is your communications company, right? And if it isn't, your communication company is truly your true voice. It is amazing. When I go around the country today, everyone tells me how much they regret the breakup of AT&T.
The only people I know of who don't regret the breakup of AT&T are the people who own stock in AT&T just before divestiture. Because if you actually add up today the value of the stock that you would have had, that was the old AT&T stock. And look at the new AT&T.
Look what it would be today. It's roughly three times what it was in 10 years-- a little over 10 years. So it's done very well as a stock. AT&T before it was broken up was about a $60 billion company. Today, AT&T itself is a 70-some billion company. And its seven operating companies are roughly 10 each.
So it went from a $60 billion company to $140 billion if you add them all together. So I think we should think of this as the new communications era. It is totally different than what it was before. The people that-- the things that are happening are dramatic.
If I showed you a view graph of what we thought of, this being the AT&T network in 1984, it would be a network with telephones hanging off of it. Today, our network is-- the last thing we would show you is a picture of the telephone hanging off at the end of our network. It would have computers. It would have video terminals.
It'd have cellular phones, et cetera. Like that. So the industry is changing enormously. Cellular business is growing at about 30% per year. Internet, as I'm sure we'll hear in the next talk, is growing at 40% per year. The long distance network-- our long distance network-- overseas network will probably change from a voice dominated network to a data video, et cetera.
A dominated network this year if it isn't already. Our overseas communications certainly is already dominated by fax. And so what has happened is this thing that for 110 years was a voice dominated telephone network is turned into a totally different business.
And it's so rapidly growing that we today think that if we want to maintain our market share in the next 10 years, our company has to grow from $70 billion to $200 billion. Because we think that communications is going to grow at that kind of rate.
So needless to say, there is enormous change occurring in telecommunications. It's probably one of the most rapidly changing businesses today. It's exciting. Why? What has brought this about? Why now? Well, the answer is that everyone-- all the scientists and technologists in this room are largely responsible for, why now?
It's very interesting to go, and look, and try to understand that and in some detail. After all, the transistor was invented in 1948. But that isn't really the story. The story is that the transistor was invented in 1948. The FET was actually patented in 1954. But the FET first. FET was made in 1960.
That FET was such a lousy device that nobody used it until the early '70s when places-- while it was made at AT&T, places like IBM cleaned up the FET, until it was a good and sufficiently good device-- took the sodium out of it, et cetera, until it was a sufficiently good device, so that we had-- we could use-- we could start doing CMOS in the late '70s.
The late '70s started CMOS. And that's what started to generate-- to integrate a circuit business in a big way. And so that throughout the '80s, the integrated circuit business went from sort of a kilobit storage to megabits we know today. And we'll go on to gigabits in the next century.
But what that-- going for that last 10 years meant-- is we've undergone enormous change. As everyone knows about the computer industry, that we now have computers that are as powerful as these things-- the big mainframes of the '60s and '70s, sitting at home in our desktops.
But not only have we got that, we now have cellular phones that you stick in your shirt pocket that are equivalent-- that are put on PCMCIA cards which slip into your computer. Those cellular phones in the late '70s were actually a suitcase sized box in the trunk of your car with wires running up to a phone in front of your car.
That's the size they were only 12 years ago. So the integrated circuit business-- while it sounds like it was here a long time ago, actually arrived about 1980. And that's what we really have to recognize. The second thing that arrived almost simultaneously with that is a new means of communication. And that is, of course, optical fiber. Optical fiber was worked on throughout the '70s.
But the first system was actually introduced in 1982. We are only 13 years later. That system is an interesting system. It's 45 megabits on a single fiber, first 45 megabits per second on a single fiber. 1982-- we had vice presidents at AT&T saying, gee, we don't need 45 megabit transmission.
We will never use fiber. We will always use coaxial cable. We don't need 45 megabits. Today, we are installing at the end of this year across the Atlantic like five-fiber cable that will have five gigabits per second on each fiber. We do experiments in the lab today at a third of a terabyte per second.
So fiber has dramatically changed what you can do in communications. Yet, think about it-- a third of a terabyte per second. Let's see, 5 gigabits is 250,000 phone voice conversations, some crazy number like that. It has reduced the cost of a transatlantic cable on a per voice basis to $400. $400.
That's less money than it costs to put a drop from a telephone pole into your house, right? But you could get a voice channel across the Atlantic for $400. So what has happened is there's just a dramatic increase in what you could do in communications brought on by the technology that has brought about.
What we find ourselves today is in an industry in which this stuff has all come upon us. And now, we're not sure how to use it. How the world will communicate in 20 years from now is clearly going to be entirely different than it is today.
And you're starting to see this coming very rapidly now with CD ROMs on your computers. We just introduced a telephone-- a set-top box that you can plug your telephone into, so that, in fact, you can dial up and put things right onto your television without a terminal, without a PC.
There is enormous change. In fact, I maintain there is a major battle going to occur in the next 10 years to figure out what the home will look like from an electronic point of view, okay? Now, there's another major thing that we have to resolve. And that is that all of this stuff has to communicate with itself, with other pieces of this network that we're going to create.
And there is an enormous challenge in trying to understand how to see to it that everything communicates. And what we're going to go through is a period of time in which that is all in a very confused state of affairs. Because we have lived for 110 years with a very straightforward way of communicating.
And that is when you dial up your telephone, with your telephone, you get a connection. Now, it may be a time slot on some very high bit rate stream. But you get that time slot every time you reserved it, okay? That time slot, therefore, is like a wire is for all practical purposes between the two locations.
That is not what you're going to get in the future. And the reason we can't give you that in the future is that sometimes you're going to want voice. Sometimes you're going to want video. And sometimes you're going to want to hit us with an enormous amount of data and in one big shot.
And so if we want to do all those things, it isn't going to be possible to just decide to give you time slots. Because how are we going to decide what time slots to give you? And how much to give you? We can't give you video rate time slots all the time when you're only doing voice.
So whole new algorithms are being established, protocols, et cetera, to try to understand these things. These things are enormous challenges to us as a research and development arm of AT&T to understand these things and to work these things out.
So when I look at our industry today, we are confronted with something that's very, very different than it's ever been in the history of AT&T and that will clearly takes us in a very different fashion. And if you think that this can-- that somehow or other a research organization is going to be immune to this kind of dynamic, forget it.
That's not in the cards at all today. And in particular, we have to get back out. And we have to be in front of this. After all, this is a revolution we created. It's an exciting-- one of the most exciting thing that's happened in a long time. And so we have been moving very rapidly and into creating a much larger organization that addresses the kinds of issues I've discussed.
These are important, fundamental issues to communication theory. And it's going to change the dynamic of the world. Therefore, what's happened in AT&T in the research is we've had to de-emphasize, to a large extent, a lot of the physical sciences, so that we can attack these kinds of system problems.
Now, I asked you. I just described the industry to you. It's very hard to figure out what else you would do in that circumstance, right? And so you've got-- we are in a mode in which we must address these issues. We must understand as much as we can about what people will want in future telecommunications.
That is a critical issue. Now, personally, I think they'll want everything. What I believe we're going to find is that the world has an enormous diversity of people. And how people use communications will vary in just about as enormously as the diversity of people. And so there will be no answer to who-- to what exact set of tools we will use in communication.
Some people will want wireless. Some people want videos, et cetera, et cetera. There's going to be a lot of different things. And so we have to do-- to address these kinds of issues. Now, that doesn't mean at the same time that there aren't enormous issues on the physical sciences on the side of the ledger.
For example, right now in the industry, essentially what is going on is there is a battle for the access to the home. Right now, if you look at it, the regional operating companies and the CATV companies have the communication access to your house, all right?
And the question becomes, how is that going to evolve? Well, you know for sure that in the near distant future, wireless access will be available at your house. You won't have to just use the cellular phone and outside. Cellular phones will come down in price.
There will be more bandwidth given out. The SEC is doing it today. And you will be able to just simply use wireless in your home. Now, whether that'll come-- so that's one aspect. The other aspect is, who is going to supply you video? How many? Will just the telephone companies? Will just the CATV companies supply you video?
Would you get video from both the CATV and your local operating company? Will a local operating company essentially be a conduit which allows other people to rent the line to your house? So there's a brewing in Congress. In this country is an enormous battle over this access to your house.
That's a very important battle. And how that goes is going to be a lot determined by technology, whether we can make-- continue to drive the price of wireless down, whether we can figure out how to make microcells instead of the base station cells you have today, which are typically five miles apart.
When you can make much closer-- more power directional, et cetera, so that you can increase the bandwidth of wireless. After all, if you really make the cell small enough, you could put video over wireless, just as well as you could put it over wired things.
And so exactly how the technology-- all these technologies evolve, is going to redefine the communications industry in the United States in the next 20 years. A very exciting time. I think when I think about basic research and ask myself, how does it interact with this industry?
I think the first thing you have to understand is that paradigm that I just described-- enormous change. We have to go with it. The people that are a premier-- premier people today are people who know about communications, who know about protocols, know about software.
Those are the kinds of people that are at a premium-- are being eaten alive today. In addition, the wireless business is growing so rapidly, it has left us in a very peculiar place in this country. We cannot find enough RF engineers. Isn't that strange? You would never think you would run short of RF engineers.
But today, I can tell you. If you're an RF engineer, you can get a job all over our industry, because there just-- the university system has essentially decided 20 years ago, in fact, as we and many other companies decided 20 years ago, wireless wasn't-- RF was not a very important subject anymore. And so most people got out and left it behind.
And it's no longer-- the people aren't trained in RF design. And so we're very much missing these people. And we find everybody paying a high premium for that right now. But in our world today, because of this [INAUDIBLE] change, when we look at the physical sciences, we have to take much harder looks than we've ever had before.
And we have to rationalize our set of activities in some kind of prioritized fashion. We cannot sit back and say, gee, let's support them all and let the flowers bloom. And then we'll go pick them. That is not the mode we're in. We don't have the money or resources to do that.
Therefore in physical sciences, we've had to look at these things very carefully. And it is interesting that the community as a whole, in my opinion, has not stepped up to some rationalization of priorities on the basic research side that helps you walk through this kind of change.
And I mean that sincerely. if I leave it-- if we leave it to corporate America, corporate America will make that judgment strictly on the basis of applications. And I believe the issue is deeper than that. We talked about it here today. But the fact of the matter is it is being made on that basis today.
If there is a better basis, we have not found it yet. And I personally believe that there is-- I believe that we are glossing over the problem when we say, hey, we're going to let all these things-- and we're going to keep everything going. And we don't know where the big, great breakthroughs are going to come.
So we have to do that. I just-- I believe that we have to come up with a better rationalization of that than that. And I think the way to do that is you do have to look historically at fields and ask where they are in their maturity curve. I, for once, have been a condensed matter physicist. And I don't care what anyone says. Condensed matter physics today is a mature field.
It is not having the impact it did in the 1940s, '50s, and '60s. In the 1940s, for God sakes, the concept of a hole in a semiconductor, which was a basic piece of fine science, led to the beginning of this whole revolution. I don't know what concept-- and that's matter physics today that could-- that anyone would like to argue-- would lead to that kind of revolution.
And so we can rationalize these things. It is not impossible to decide what fields are matured and what fields aren't. And we just have to face up to that issue. And I think we haven't done it. And I think you were getting nervous in your chair, so I better sit down. Thank you very much.
KNOWLES: I was only getting nervous about being attacked by condensed matter physicists. Our last speaker is Dr. Anita Jones who, in case you haven't read quite carefully enough, her AB was from Rice in math. She then took a little detour and did an MA, a master's at the University of Texas, Austin in literature. So we are going to have a properly cultivated last talk-- followed that by a PhD at the Carnegie Mellon in computer science.
Dr. Jones has been chair of computer science at the University of Virginia and then entered government service on the Defense Science Board and the Air Force Scientific Advisory Board. Since 1993, she has been director of Defense Research and Engineering that is responsible for all science and engineering at the Department of Defense and including all kinds of excitements like ARPA. With pleasure, Dr. Jones.
JONES: Thank you. I'm the hybrid that represents the government on this panel. And I thought I could best play the part by making a handful of observations of how, why, and what's changing about the way government funds information science. So I want to start with the fundamentals.
There are a number of reasons, places where it's appropriate in the United States for government to fund science and general information, science in particular. This is a time of change. Certainly speaking for the Defense Department, as you are often told, are a threat of imminent, large-scale warfare-- is tremendously diminished.
From my point of view, looking over the science and technology program, I will tell you that the competition just heated up. We had leisurely procurement cycles that happen to be precisely the same length on the average of 15.5 years or that of the former Soviet Union.
But the next kind of battle, albeit are smaller-- our forces will be fighting against any kind of arms that you can find in the international arms market. And that could be based on technology developed in any advanced countries in the world. And so on item two, our mission remains. And the competition is heavier.
Let me point out that in this debate about how and what governments should fund in particularly in the universities, particularly in fundamental science, whether in industry or the universities, a objective and obligation that the government has had over the last several decades is paying a substantial portion of the graduate education in a number of fields.
And I can tell you in DoD, we put a substantial amount of resources into funding graduate students, whereas the National Science Foundation funds about 12,000 students a year. DoD funds 8,000, so about 75% of that amount. And as questions are asked by the military inside the DoD and on the Hill, what should Defense fund?
There is a real issue, I think, that will be-- come to the fore again this year about what defense should do in terms of funding students. I believe it's important for the government agencies across the board to nurture the future scientists and engineers.
And picking up a point that George Whiteside made earlier, I think industry needs to make a statement in this discussion of what kind of investment the government should make with their tax money in nurturing the infrastructure of the country.
Let me turn to funding of information science. And let me start with something that Jack alluded to. He and the Clinton administration created the National Science and Technology Council. It's got nine subcommittees. I chair the Committee on Information and Communications. It's one of nine.
And I want to tell you a few of the things that's happening under that aegis. Budget numbers are hard to pin down. They're very slippery devils. We took a crosscut voluntarily-- not all the numbers approved-- looking at R&D in information sciences.
And this is the kind of results we got. You see across the horizontal axis some areas of investment. The units are millions of dollars-- or hundreds of millions of dollars. Excuse me. And the two codes-- the upper portion of the bars are those things that are included under the major initiative that's under the NSTC Committee on Information Communications.
And that's the High Performance Computing and Communications Initiative. This roughly adds up to $2 billion. Again, it's at the back of the envelope accuracy, no more. And roughly half of it is considered part of the high performance. It's an investment in the high performance and of these areas.
This is a major-- a flagship initiative for the Clinton administration. The major components today of this initiative are listed on this particular slide. One of the basic premises are that the government should invest in the very high-end problems, the very high-end technologies.
The government invests in the long term. The government ought to solve the problems we're going to see a couple of generations out. Because solving some of those problems at the research level in the laboratories will set the scene for new products for industry to derive at later dates.
Frank Rhodes said this is a particular area. And he called out the High Performance Computer and Communications Initiative and said it's in need. It needs a synthesis that the National Science and Technology Council provides. And I'd like to say that it's getting that.
This is a premier example of the virtual agency concept. There are a number of agencies. The large dollar investors in this initiative are DoD through ARPA, Energy, NASA, and NSF. This is not new money. This is money that those agencies were-- probably would have put in computing and communications.
The initiative is the forum, though, for the agencies to come together. The program is bottom up planned. The money remains in the agencies. And those agencies put their mission first. I happen to believe that's appropriate. But in fact, information technology and the long-term research problems are so ubiquitously applicable that many of the agencies see the same kind of problems-- need to be solved to achieve their mission.
And so you see a lot of joint investment. For example, under this initiative, we are fielding the so-called handful of gigabit test beds. A more recent program under this initiative is joint between ARPA, NASA, and the National Science Foundation in digital libraries, which is an interesting arena.
And from my experience with this virtual agency notion, which I was, Jack, quite skeptical of when I first joined the government a couple years ago, really is working. It does eliminate much of what is unnecessary duplication. There's a lot of partnering as exampled by the gigabit test beds, which are predominantly funded through ARPA, and the National Science Foundation, and the digital libraries.
This is a particular initiative. And a second point I want to make is that it's evolving rapidly. And the interagency process is helping that. When this initiative started in the early '90s and it was difficult to get the agencies around the table in a cooperative way and quite a heroic thing, I think, the focus was on the grand challenges of science, real-time weather, modeling fundamental physical reactions, computational fluid dynamics.
The focus was on teraflop computing and on gigabit networks. In fact, within a couple of years later, the initiative evolved. Because I think the research problems are changing as rapidly as the information industry's product cycles. And so new things come to the fore.
And the major event, I think, and the evolution of this particular initiative was to add a new portion of the High Performance Computing and Communications Initiative. And that was the society-based national challenges, health care applications, energy, global climate applications.
And in fact, right now, I can tell you from my vantage point that this is initiative that is changing. The next step is to rethink the set of problems and the fundamental basis for the initiative. We have done a strategic planning exercise under the NSTC
Aegis have picked these six focus areas as being important. And we are going to re-evaluate-- make a slight change to the strategic focus of the initiative. It has gone from having a very hardware focus to a much heavier software and application focus.
I think in this round of evolution, you will see some of the more holistic problems addressed. One of the comments John Hopcroft from Cornell made over lunch was High Performance Computing Initiative focused on networks. That's just getting bits from here to there.
The real issue is communications, whether it's-- as Dr. Brinkman said, the RF communications, the scale of network speeds that one has to cope with, all the way from very slow speed, which is something that we see, partly by choice to have a low profile in DOJ battlefield applications, up to very high speed.
And these things have to coexist and be communications systems. We do have a fairly chaotic way of having the government talk to the nation, talk to the universities, more particularly talk to industry. It's usually through a set of forums, a set of advisory groups, a set of workshops.
And so on one next action that will happen in this context-- and I'm only looking at a narrow context in the larger science and technology program, is to have some workshops and dialogue with industry, the industry associations about what should be the strategic foci in the information science area.
And so the second point that I wanted to make is that this virtual agency notion has become real under this administration. And I think it's returning value. And I've tried to show you one example and some of the changes that are happening in that example.
Next observation [INAUDIBLE] is I want to talk about one particular relatively small program that I think is very exciting that is funded in defense. Because I want to tell you about how we make that investment and how that is quite different than typically what you'd find a university doing or what industry doing.
The area of technology development is microelectrical, mechanical systems. For those of you who haven't met these little beasties, it's very exciting. And then instead of just having a chip that kind of lays there and computes on information, you give it a sensor.
So it can sense the environment. And you're given an actuator. So it can do something about the sensed information that it's computed on, and chewed on, and take an action that is felt in that environment. And there are lots of applications. And fluid sensing, accelerometers, optical devices, sound, and even liquid movement kind of applications.
One of the-- and again, I said this was going to be a small example. One of the programs that we've put together is a successor to MOSIS, the metal oxide semiconductor service that we had for semiconductors, is called the Multi-User MEMS Projects.
It's through the microelectronics center in North Carolina. And I think this is a good example of how the government, how DoD in particular ought to invest in a particular technology. This is a fabrication facility. If you have a design of a combination mechanical and an electrical MEMS device, send us electronically, of course, that design, plus a token amount of money.
And some number of weeks later, you get half a dozen [INAUDIBLE] having fabricated your design. And so this is a service. The users are depicted here in color. Federal use is the top slivers in blue. University use is in black and is slightly dominant.
Industrial use is in red. And about half of that industrial use is small business. And so what this service permits is for both university projects to operate in startup mode and get some experience with this technology. And for small business, also, to have an entree to fabrication lines that they could not otherwise afford or get contractual access to.
This is the kind of thing that government ought to do. This is long term. There are some short-term applications that are coming to the fore. One of the premier applications that exist today is your airbag trigger device. That's a MEMS device. I think it's important that the government-- and this is my observation for this segment is the government ought to look for enabling ways of making investment.
And I proffer this as a reasonable example of that. I'm going to now to investment and information science based on the NSF Science Indicators Report. This is what the government is investing in-- excuse me-- both basic and applied. I won't use those terms again.
Research in information science. That report breaks it out into math, and computer science, and electrical engineering. This is part of a consistent, sustained long-term, many decades investment by the government in information sciences. The government has been remarkably consistent.
I want to put my DoD hat on. The smaller bars are the DoD component of that investment. The bars in green are the total government investment. DoD is roughly, I think, 16% of the total set of research dollars that are being considered here.
As I said on the very first slide, agencies invest in their mission. Two of the most important areas are a most important set of technologies for defense. Certainly during my tenure and in overseeing the science and technology program is information sciences.
And so we are inordinate percentage of that investment. There are chilly winds blowing, particularly on the Hill in terms of asking questions of exactly what or even whether DoD should be investing in university research. And that gives me great concern.
I've already highlighted the student investment we make. In fact, I think to a first approximation-- well, let me say differently. DoD has funded the majority of microelectronics advanced graduate students who have been funded by the government.
Because we have a substantial portion of not just these two investments, but later stage electronics investments as well. It's absolutely crucial, as every one of you know, to have a sustained investment in a research area over years. It doesn't matter whether it's coming from DoD, or coming from DOE, or coming from NIH.
One needs to sustain it to sustain the infrastructure in the country. And so let me also, again, go back and take further the comment that George Whiteside made. I think an important thing that industry can do is to say to the nation at large what it feels or the degree to which it feels it's important for the government to sustain these long-term investments.
The government and long-term investment in information science is, as best I can figure it, several times larger than the long-term investment, not the short term. But the long-term investment that industry is able to make. Because typically, it must have something like several year horizons.
Let me close with this slide. This is a derivative slide from a really interesting piece of analysis that is in a report that's just coming out of the academy chaired by Ivan Sutherland and Fred Brooks. Red-- the horizontal axis is time. We're talking information science now and technology, really.
Because you're going to see market on here. The vertical axis is just-- if you'll give me-- cut me a little slack-- breakthrough technology areas within information technology. The red is when the government invested. Not how much, but when the government invested.
The blue is when there was a serious industry investment. And the black is when you had more than a billion dollars of business based on that aspect of information technology. And you see it goes all over the map. In some cases, government made the early investment followed by industry.
If you look at the risk, that's the reduced instruction set computers. The concept of make instruction set simple rather than complex. In fact, IBM made the early investment where the IBM made a 1. I've got some of the notes on the side flipped, so ignore that. But apparently decided that was not a good perturbation in their market.
And so they did not follow it. Government then invested in risk. And you will see on the front page of, I think, yesterday's Wall Street Journal-- Intel making announcement that its next generation processor would be risk based. This partnership between the government and industry. But by the way, university.
Because the bulk of the-- or the majority of the long-term research funded by government, at least for defense is out in the universities, is absolutely crucial in information sciences. And I think this shows you some of the richness of that fabric. It is important, I believe, to sustain this investment. And I believe it's absolutely crucial that the nation understand this notion of long-term government investment.
And I think it's likewise, absolutely crucial that the one party who is most objective and that being industry as opposed to the self-serving government people like me are the self-serving university people who get the bulk of the government money into long-term research stand up and talk about what's important and maybe what's dumb about the way we're making this investment. But I think it's crucial to the country that we continue. Thank you.
KNOWLES: I need first to ask the dean's permission. Where is he? For whether we may have a couple of questions. We are now 18 minutes. We started 15 behind schedule. And we're now 18. Not bad. But are we-- sorry? All right. We are. Please. Dr. Brinkman, where are you? Come. Questions. We have been given a dispensation by [INAUDIBLE]. Yes, Phil? Oh, sorry.
AUDIENCE: No. I think I would make an observation. I think that if you're not careful with your charge to the audience, that now's the time to pare down. You're going to end up with no radio engineers or the equivalent in the technical field. I tend to be more of a free market person.
Let the market forces dictate the areas that we invest in basic research. So I tend to support a much broader base of research base. I don't think because times are hard that now's the time to retrench and start making selections, so it's an observation.
BRINKMAN: I guess I don't understand, because we all have to make selections--
BRINKMAN: --every day. And I mean, I make a selection once a week. And we have to do that on some basis. And--
AUDIENCE: There's a difference.
BRINKMAN: --it determines what-- the funding of science. And if you look, it's certainly clear that biological sciences have become very large compared to physical science in the last 10 or 15 years. It's also clear that computer sciences has grown enormously. So we have done some things. The question is whether we've done enough.
AUDIENCE: I was just curious at the same thing being from the field. What is the investment by AT&T in terms of R&D vis-a-vis and then broader from the industry. I'm just interested in the total scale of research.
BRINKMAN: The scale? I don't know if I can actually speak for the industry. The industry-- it's a little hard to figure out what it is these days. But AT&T certainly invests-- let's see. It's about 30,000 employees in R&D. So that's probably several billion dollars.
Our research organization is about a $350 million operation. And it's like any other central research organization in a large corporation-- has had a lot of pressure on it to survive. To help the corporation from 1984 to today, I would say AT&T is a new company.
And it is broken up into business units. Many business units went through enormous change to survive, basically, from a non-competitive world to a really tough competitive world. And so those changes have caused-- there were a lot of people screaming at our doors, saying, hey, we need help and now. Not next week. Now. And so we've had to adjust to that very strongly.
KNOWLES: But underneath that question was presumably a desire to know the comparative or proportionate investment into R&D of, say, biotechnology versus communication. Was that underlying your question?
AUDIENCE: Just I was interested in the scale in which research was ongoing in terms of investment in research and R&D.
JONES: Let me try that from the-- I've tried to get the same answer generally. There are several-- I've looked at several sources to make estimates of, how large is the information technology industry? And the best I can figure out is it's $500 to $600 billion a year.
So let's say this is a very aggressive, fast moving industry. Let's say they put 10% in R&D, okay? So that's $60 billion per year. How much of that is long term? Very small piece. Should I pick 10 or 20%, John?
AUDIENCE: That's low.
JONES: Of that?
AUDIENCE: 5? You're being generous with 5. Well, 5 times 60 is--
JONES: No, no. 5% of $60 billion is $3 billion a year.
AUDIENCE: All over?
AUDIENCE: All over. That's [INAUDIBLE].
KNOWLES: Jim Vincent, you have a question?
AUDIENCE: Well, when you say that, what's the comparison? If it's $30,000 a day and say round numbers. That's $3 billion or something like that at AT&T R&D. How would that compare with the time you split up?
BRINKMAN: It's almost exactly same size, but very different. You have to remember a couple things. AT&T acquired NCR-- $9 billion dollar company. It acquired McCaw, another $12 billion company. So if you look at the core of AT&T, the total R&D has gone down. But with the acquisitions, it's kept almost constant.
KNOWLES: More concerns? The primary concern maybe for tea, in which case, let me thank all the six contributors very much, indeed, for this afternoon.