Future R&D Environments: A Report for the National Institute of Standards and Technology (2002)

The workshop held by the Committee on Future Environments for NIST on July 20-22, 2001, in St. Paul, Minnesota, was attended by committee members, invited experts from various sectors of industry and academia (see Appendix C for the list of participants), and staff. In addition, the acting director and a dozen other leaders of NIST were present as observers. All participants were provided with copies of the commissioned papers (Appendixes E through I). Three of the commissioned papers (those by Young, Bean, and Finneran) were presented during the workshop. A large part of the workshop was devoted to three half-day sessions of breakout groups. These sessions were organized along the three themes selected by the committee: “push,” “contextual,” and “pull” factors. The push factors session gathered together judgments on possible futures for a set of scientific and technical fields. The pull factors session focused on societal demand factors and the economic, social, environmental, and political needs and sensitivities that would promote or inhibit research and development in certain areas of science and technology as well as innovations based on that research. The contextual factors session considered issues such as changes in the organization and support of R&D in both the public and private sectors, educational goals of students and methods of delivering education, and patterns of investment by the private sector, all of which might be expected to change the process by which ideas move from research to product. While obviously this organizational principle is somewhat arbitrary, the committee nevertheless found the principle a powerful one, given the task before it.

The workshop proceedings are organized around the topics considered by

the breakout groups at the three sessions (see Appendix B for the workshop agenda). In the first session, each breakout group was asked to consider push factors parsed into science and technology trends in three fields—one group considered the biological sciences; another considered materials science and technology; and the third considered computers, communication, and information technology. At the second and third sessions, every group dealt with the same topics and questions. In the second session, which focused on contextual factors, the three breakout group topics were the organization of research, investment patterns, and public policy issues, respectively. In the third session, which focused on push factors, the groups discussed social and cultural trends and attitudes; economic considerations; and the international context.

The chair of each of the three groups was given a template—a form on which to summarize the results of their three sessions—for cross-comparison in a plenary discussion that was held on the last day of the workshop. In addition, on that last day, the template-ordered results were displayed and distributed, and the entire morning was devoted to brief reports from the breakout groups, followed by plenary discussion.

The committee is the first to recognize that analyzing trends in science and technology by dividing them into push, pull, and contextual factors is arbitrary when discussing any particular technological development. It is attempted here, at the risk of overlap and redundancy, to avoid the trap of technological determinism, that is, assuming that what is technologically possible will be used as soon as it can be developed. The rate of innovation—that is, the speed at which scientific advances are made and turn into economically viable technologies—also depends on the infrastructure for innovation and, perhaps more important, on social and economic factors. The innovation infrastructure includes the supply of human capital afforded by a nation’s education and training systems and the funding of research and technology development by industry, venture capitalists, nonprofits and foundations, and government. Social and economic forces accelerating or impeding technology are legion. The committee’s approach recognizes that social and economic factors play a large role in the fate of any technology, and it underlines the importance of the innovation infrastructure in enabling economic growth and increased productivity.

The purpose of this session was to identify and discuss the scientific advances and new technological developments that will create the potential for new applications in the next decade. Each subgroup was asked to consider a different set of questions, focused in a particular area of science, although they were all encouraged to address new developments arising at the intersections of the areas.

1. It is possible that for each of the major diseases a relatively small number of genes—probably hundreds rather than thousands—will be identified over the next few years, selected, and studied extensively. Large-scale screening of genes will be less practiced in the future.

2. The large-scale screening of proteins is not likely to be an effective approach, because analyzing amino acids (proteomics) is more difficult than analyzing nucleic acids (genomics). Some of the most important protein structures are still unknown.

3. The technology for protein analysis is improving (mass spectrometry and other technologies are being used). The advance in protein analysis is hindered by the difficulty of crystallizing materials for analysis, but efforts will be made to improve that process.

4. There is a sequence of tasks involved in understanding the role of genetics in human health and disease—identifying the gene, then the protein, the protein structure, reaction pathways, cell biology, organ function, and so on. In next 10 years, work will concentrate on the gene, the protein, and the protein structure; the remaining steps pose more difficult challenges.

5. Genetic, immune, and small-molecule therapies are possible applications. Custom drug design is possible, but companies are concerned about minimizing drug side effects as well as maximizing efficacy.

6. Multigene traits are more important in agriculture than in human health, and they pose a much more complex computational problem.

1. There will be important advances in tissue engineering, based on better understanding of synthetic surfaces and how to characterize and modify them, and on improved understanding of cell membrane structure and function. Interfacial science—that is, the study of the interfaces between biological and man-made materials—will drive and be driven by tissue engineering.

2. It is now possible to generate certain cell populations—beta cells, liver cells—that are two-dimensional tissues (epithelium, endothelium, vasculature). There has been progress in regenerating nerve cells. Synthetic materials can be used to guide the morphology of cell growth and proliferation, develop hybrid organs, and protect cells against rejection. But the ability to grow whole natural organs is unlikely to be achieved in the next decade.

3. The use of MEMS for in situ intelligent sensors is growing in importance.

Microsurgical instruments can be operated remotely, which allows less-invasive surgery. Robotic instruments are already used in neurosurgery and eye surgery.

4. More and more site-specific imaging markers are becoming available, which is leading to advances in imaging.

5. There are new imaging techniques by which to study brain metabolism, which is enabling better interpretation of functional MRI images.

1. As classical functional taxonomy shifts to the genetic classification of organisms, it will be important to identify the traits to focus on, so as to reduce to a manageable amount the information that must be sifted through.

2. Advances in genomic data on plants will improve the analysis of inadvertent sequence changes during the genetic modification of organisms (and also during traditional plant breeding).

3. Biological computers based on DNA will be possible, but they will be curiosities rather than practical devices because of the limited range of problems they can solve.

4. There will be greater use of the microorganisms known as extremophiles for environmental cleanup (oil spills, etc.). Such organisms can function in conditions we have not been able to manage before, such as high heat or pressure or in the absence of oxygen, and they are beginning to be used as catalysts in fine chemical production.

5. It may become possible to use biotechnology for high-efficiency energy applications.

1. Atom-by-atom customization of commercial products will come about in the next decade, enabling the surface properties of materials to be manipulated. This in turn will impart functionality to the materials, allowing them to be used in many different applications.

2. There will be further development of photonic crystals and their applications in communications.

3. There will be new materials requirements for electronics: low-K dielectrics with acceptable mechanical properties and high-K dielectrics with low trap and interface states.

4. Nanotubes are suitable for a number of specific applications. However, the ability to process and prepare them in commercially useful forms and

quantities, and at acceptable cost, has been a substantial barrier but will probably be overcome during the next decade.

5. Because of the small size and optical properties of quantum dots, the technology for producing and using them will have potential application in computing, optical communications, and biological and chemical sensing.

6. Instrumentation and techniques to identify, measure and analyze, and standardize exist for bulk materials but not for nanomaterials. It is expected that they will be developed during the next 10 years.

1. The efficiency of fuel cells has increased greatly over the past decade. They may one day replace batteries for consumer products, but the fuel, hydrogen, is difficult to carry in an electric automobile. Either an alternative fuel must be found or an infrastructure developed to supply hydrogen. The first important application for fuel cells and other technologies in the next 10 years will probably be in local power generation for the power grid.

2. Light-emitting diodes (LEDs) are a promising low-energy source of white light and energy and electric power, if the problems of producing low-cost blue LEDs or organic LEDs (thin films) can be solved.

3. There is a need for catalysts to produce lower-cost, more-efficient materials. Nanostructures might be used as catalyst carriers.

4. There will be advances in the development and use of bioremediation materials, whether biological or other materials.

1. Thermionic cooling will be developed as an alternative to refrigerants, which harm the ozone layer.

2. MEMS will be applied for new methods of optical switching and self-replicating/self-organizing/self-assembling.

3. The technology for organizing quantum dots into useful systems, which may be developed during the next 10 or 20 years, will have important applications in archival data storage devices.

4. MEMS devices could be developed for microfactories that can generate and deliver power on a local basis without needing more transmission lines.

5. MEMS will be used for cell sorting.

6. Integrated systems of sensors and communications will be developed for medical care and other applications. MEMS, with their sensor capacities

and delivery capabilities, may succeed in making artificial organs within the next 10 years.

7. New materials will be developed for the separation sciences.

1. Component-level advances will continue through the decade.

2. Storage will become sufficiently cheap to permit continuous recording of large amounts of visual and physiological parameters, and reductions in the size of storage devices will enable complete personal information to always be in each person’s physical possession.

3. Advances in computation and other components will allow very smart, very small, and very cheap sensors that can be widely distributed and communicate easily with networks.

4. Simulation and modeling will continue to displace physical experimentation.

1. The complexities of large systems will limit the rate at which the substantial increases in computational and networking power can be applied.

2. There is a long history of innovative applications of artificial intelligence, and advances in sensors that are intelligent and able to integrate information will greatly improve a system’s ability to interpret data in context. Speech recognition has been improving tremendously, but speech understanding under natural conditions requires a degree of context that is still not possible. Visual recognition will still be difficult, even with 100-fold CPU speed advances expected during the next decade, although existing capabilities are used routinely by the military—for example, for target recognition by moving aircraft. Advances in personalized interfaces discussed above would help, because each person could provide the computer with personal data that would ease the recognition and some of the understanding problems.

3. Systems are gradually improving at performing both self-diagnosis and self-correction, which will help manage the next level of system complexity.

1. Interface issues will dominate, and limit, the adoption of most new applications. Interfaces have to be natural and easy to learn. An example of a

successful interface is the cellular telephone, which the consumer can easily use despite the sophisticated computation and networking involved. Unless there are major advances in interfaces, there will be limited commercial demand for more increased complexity or further performance breakthroughs in computers and networks.

2. Primitive telepresence will be available within 10 years, but it will take a two-orders-of-magnitude improvement in computation and communication for videoconferencing to begin to substitute seriously for face-to-face communications.

3. Personalized interfaces that recognize individuals and their habits will enable less sophisticated users to tap into high-powered computers and networks.

4. Better technology for secure data transmission, storage, and authentication will be developed, but privacy issues cannot be solved just by better technologies. They involve trade-offs and choices. For example, a person carrying his or her complete medical record may want total control over access to the record, but what if that person arrives in an emergency room unconscious?

The first observation was that separating technology push issues from societal pull issues conceptually is very difficult to achieve in practice. Those entities developing technologies are well aware of the economic, social, political, and legal opportunities and barriers. Trends in industrial research, especially the development of closer linkages between commercialization and research, are imposing an even greater consideration of pull factors in decisions on research directions and technology development. The subgroup on push (supply) factors in computational science and information technology devoted most of its time to the need for improvements in machine—human interfaces in order for computer and information technology to develop further. If technologists do not improve those interfaces, demand for more advanced technologies will be dampened by the reluctance of users to buy more complex products.

There was a discussion of the potential for advances in artificial intelligence (AI). One participant argued that there has been a long history of innovative applications of AI, and many Fortune 500 companies are using it routinely. The development of better sensor devices will enable more intelligent integration of information and improve the capacity of systems to interpret data in context. Systems for visual and speech recognition are in routine use, for example, by the military in target recognition. Another participant agreed that AI has achieved much but said there was a long way to go to achieve speech understanding in other than controlled contexts.

The position that only a few hundred genes at most will be involved in studies of health and longevity was questioned. Recent understanding from sequencing the human genome has moved past the one-gene, one-disease paradigm, and gene chip technology may show that one disease is associated with the expression of many genes. One response was that it looks as though only a few primary genes are involved in each major disease, although there may be many down-stream modifiers. Gene expression chips allow identification of many genes involved in a disease, but the chips show the effects of the disease, which can be very complex, rather than its causes, which can be very simple. It was observed that the issues addressed by the groups did not include trends in manufacturing, which is still an important although declining set of activities. This turned the discussion to the reports from the second breakout session, which focused on, among other things, shifts in the organization of research, development, and innovation.

1. Universities are important in fostering industrial innovation and speeding the commercialization of innovation, but needed changes in their organization and practices are taking place only slowly:

   • Traditional academic organization by disciplinary departments does not facilitate the interdisciplinary training that industry wants, although universities are trying to respond to the demand for such training.

   • The incentive system in academia is changing. Students are now more interested in patents than papers, and the best graduates are going into industry (especially small companies), not academia, which may make it difficult to maintain the quality of university faculties.

   • The review and reward system should be revised to promote interdisciplinary research and graduate training.

   • Successful communication across disciplinary boundaries must be ensured. Vocabulary differences between disciplines must be bridged.

   • Higher education should provide more effectively for the technical agility of the workforce, and education and experience need to pay more attention to entrepreneurial skills and teamwork.

   • Scientists and engineers need to be taught to be integrators across disciplines, including business, manufacturing, patents, etc. This will require strong continuing education programs.

   • Large global universities will emerge, and private universities will have to specialize to become global centers of excellence. MIT, for example, is moving its Media Labs into India and Europe.

2. The organization of research is changing and will continue to change in industry. The trend is toward decentralization within companies, greater outsourcing, and partnerships with other companies and research institutions.

   • Corporate central research organizations are in decline. Universities could become the suppliers of just-in-time knowledge. Venture-capital-funded enterprises and independent labs like SRI and Sarnoff may also replace internal corporate research operations.

   • There will be more cost sharing through jointly owned central research labs. For example, Rockwell Science Center is now owned jointly by two separate companies, Rockwell Collins and Rockwell Automation, and HRL Laboratories, formerly Hughes Research Laboratories, is owned by Boeing, General Motors, and Raytheon.

   • Roadmap activity in some research areas is not adequate and would help focus R&D on needs. This should be an industry activity, but government could catalyze roadmapping and make some changes to avoid inhibiting collaboration. This may be difficult for some industries, for example, the pharmaceutical industry.

   • Mobility of people and ideas is a competitive advantage for industry as a whole, although out of self-interest, an individual company will typically resist mobility at the margin.

   • Technology development will be significant in new companies and will be marketed most effectively by larger companies with the necessary infrastructure. This is an underpinning of a successful mergers and acquisitions activity.

   • Larger companies will need to organize project teams internally (and perhaps externally) to foster rapid innovation and accelerate the integration of technologies.

   • Alliances between companies can facilitate development and marketing and will be an important activity in the future. Alliances speed time to market and facilitate specialization and learning, but they also pose serious organizational interface challenges for and could cause conflict between longer-term strategic goals. Unsuccessful alliances are more common than successful alliances.

   • Geographically distributed research is already a reality for information sciences, but it is more difficult to achieve when physical laboratory facilities are involved.

   • Merger and acquisition activity among large companies generally decreases investment in R&D—companies merge, there is overlap, and the number of people involved in R&D is cut—and does not foster innovation.

• The benefits of globalization, including 24/7 project activities and greater diversity of thinking, are increasingly important.

3. Interactions and relationships between academia and industry are becoming closer, more frequent, and more complex, and the roles of faculty and students in industrial activities and their implications are creating issues that must be worked out.

   • The university-industry interface varies by field. It is more effective in the life sciences and is more likely to continue there than in the physical sciences. The difference stems in part from the looser connection between research and application in the physical science areas and the large capital investment that industry itself is making, which lessens the need for university relationships (universities are still important for supplying highly trained physical scientists and engineers).

   • The high quality of industrial research is increasingly attracting faculty to take their sabbaticals in industry or otherwise move between the private sector and universities. This trend involves trade-offs as faculty allocate time between their start-ups and their students.

   • University faculty have become involved in company start-ups as entrepreneurs. There is increased collaboration between universities and industry, including equity participation in start-ups based on university technology and resources. Conflicts of interest and/or the appearance of conflicts of interest must be addressed as collaborations and entrepreneurship increase. Public disclosure of all information will be a key element in addressing this issue.

4. The organization of federal research agencies around missions and diseases is not conducive to support of basic research. For example, the orientation of the NIH toward specific diseases and disease hypotheses does not encourage the kind of multidisciplinary collaboration on basic phenomena that leads to fundamental advances even though it is not aimed at curing a particular disease.

5. Some other trends in the organization of research are these:

   • Financial rewards for researchers vary greatly from sector to sector— government, academia, large corporations, and venture capital startups.

   • There is no silver bullet in organization structure. Whatever the structure, it must include business knowledge and not inhibit innovation while keeping that innovation focused on the commercial objectives.

   • The free market is a major asset in providing rewards and recognition and handles this well. Some are concerned, however, that the system

negatively affects the advancement of some employees—for example, women with families who cannot put in long hours at critical phases in their careers.

1. Innovation is capitalized quite well by both public and private capital under the patterns that exist today. However, environmental, political, and other issues may change this in the future, and new accounting rules may affect investment patterns.

2. The venture capital model is creating multiple points of technology innovation by funding many small companies, especially in the biotechnology and information technology areas, that are pursuing parallel research.

3. Funding models (e.g., small, short-term grants) conflict with the creation of centers of excellence, which require stable, long-term support.

4. Support of basic research is in question and needs to be addressed in both government and industry. Such research takes place largely in universities, and sources of funding need to be identified. But public funding of research may increase, owing to public anxiety about terrorism.

5. Short-range accountability for results is hindering long-term research as well as the development of infrastructure necessary for greater innovation in the longer run:

   • Innovation in industry is constrained by a need to have good quarterly results, especially among companies that go public early. Long-term research is less constrained by short-term financial expectations in private companies.

   • In some particular areas, metrology and standards require more investment. Deficiencies exist today in standards, materials properties, and metrology that would facilitate the insertion of innovation into the economy.

   • Infrastructure needs that are not adequately supported by the profit motive may require government resources.

1. Peer review has become more conservative as the number of applications increases relative to the amount of funding. Mechanisms to encourage risk taking need to be developed.

2. Despite globalization of R&D and innovation, regional clustering of innovation remains important because it concentrates resources and satisfies the need for critical mass. Close connection to customers and their needs also supports localization despite global ownership.

3. There are more and more opportunities for private and individual investment in education and continuing education:

• Information technology developments will cause major changes in education. Formal education and continuing education will become less important as it becomes possible for individuals to solve problems from information they can gather rather than from knowledge they possess.

• There is a rapidly growing for-profit education sector, including credentialing, e.g., M.B.A.’s.

• Effective techniques to invest in educational needs should be identified.

1. Workforce needs of the future will not be met by the educational system without some changes. The gap is widening between traditional K-12 education and a technology-driven society. The problem needs to be addressed at the primary and secondary school levels.

2. A large and growing percentage of graduate students and Ph.D.’s is foreign, which may become a problem if the supply drops off, but the trend does promote communication across national borders as some of these people return to their own countries.

3. The U.S. intellectual property regime—including patent and copyright laws—is currently a constraint, as innovation accelerates:

   • Although intellectual property issues may be on the path to resolution, there is still a problem in newer fields.

   • New waves of innovation usually result in patent wars. The costs of litigation must be minimized.

   • U.S. patenting policy on gene sequences needs revision if multiple gene applications are to be encouraged.

   • Patent law in new areas such as certain biotechnology areas may have several years of uncertainty that may inhibit investment in innovation.

   • U.S. copyright policy and law also currently inhibit progress.

   • Differences between U.S. patent and copyright policy and patent and copyright policy of our trading partners need to be resolved.

4. There is a growing need for standardization. For example, standards are needed in genomics/proteomics (judging sequence accuracy, standardizing software, and so on), which will not be met by the National Institutes of Health or the Food and Drug Administration. Also, there are data format issues (longevity and data migration) that require formal standards. However, while physical standards are always useful, they sometimes serve to inhibit trade and other times to promote it. Defining standards so that they promote it will be a major challenge.

5. The security and validity of online information are becoming an important issue.

6. There are some other impediments as well:

   • Provisions in laws and regulations that inhibit adoption of technology without reason (some of them are more than a hundred years old) should be reviewed and abolished or modified, as was done in Massachusetts.

   • Regulation and taxation of Internet activity may inhibit progress.

   • Foreign outsourcing of research raises transfer of technology issues.

Some participants argued that the assessment of the patent and copyright systems was too negative. They said that while not perfect, the systems are generally working, although there may be problems in the fast-emerging area of biotechnology. Others restated the case that the patent system is not keeping up with changes in science and technology and is unduly hindering positive developments. There were a number of examples of problems with patents, including patenting too early and too broadly, which hinders development; unrealistically high expectations on the part of universities that licenses will become a significant source of revenue; the use of defensive or blocking patents; and the reliance on litigation to make policy because initial patents are too easy to obtain. History shows that periods of accelerated innovation have usually led to patent wars, which imposes large financial costs and hinders innovation, and public policy should try to minimize such conflicts and litigation costs. There are also substantial differences in the patent and copyright policies of the United States and its international trading partners.

It was argued that differences between public and private research universities are small and narrowing. State funding is declining as a percentage of public university budgets, and public universities are increasingly dependent on the same sources of support (i.e., federal R&D agencies, foundations, and industry) as private schools. Also, research is globalizing as is the delivery of education, another reason to have a global university. State universities already offer programs abroad. Another participant agreed that public and private research universities are going in the same direction, but state schools will lag because of state legislative pressure to increase the emphasis on teaching relative to research. Also, the corporate status of state universities is not the same; some are more independent from state control than others.

One participant pointed out that the groups did not address trends in the status or role of the DOE national laboratories. In ensuing discussion, it was observed that the national laboratories are good at pursuing complex, long-term, interdisciplinary projects of the kind that industry values. The example given

was the extreme ultraviolet lithography project, which involves three national laboratories and a consortium of companies, including Intel. On the other hand, there are constraints on the national laboratories when it comes to joint activities with other research institutions. The national laboratories are relatively isolated from industrial involvement, more so than universities, whose professors are increasingly taking sabbaticals to work in industry and whose former students working in industry provide useful feedback. The national laboratories have appointed individuals and established advisory groups to pursue industrial commercialization and venture capital resources and have met with varying degrees of success, although it is perhaps too early to evaluate such initiatives. The national laboratories are also diverse in their missions and roles, which makes it hard to generalize.

There was an extended discussion of the trend toward outsourcing of R&D by industry and the implications for future innovation of reductions in central research units. Companies can fund R&D internally, outsource it, or acquire it. It is an expense that, like other expenses, is judged on its rate of return. Not all research is a core competency that a firm will want to maintain internally. Will innovation tend to take place in small companies and be acquired by large companies, and will that offset the downsizing of central research programs? Or will it be outsourced to small firms or overseas? Are time differences and geographical distance an important constraint on outsourcing overseas? Will increasing dependence on outsourcing lead to inadequate research bases in the United States when foreign research laboratories become competent enough not to need U.S. support anymore? Are start-up companies sources of innovation or are they spun off from big companies or a university with a great idea? Do the central laboratories that remain no longer see themselves as independent campuses but as an integral part of the business development process? Will universities or national laboratories be able to become a more direct source of industrial innovation? Will national laboratories that move into new mission areas compete unfairly with industry because they are federally funded, or are they too shackled by government rules and regulations? The consensus was that it is too early to tell how it will all turn out or to know the consequences of, say, decentralizing R&D or moving it to foreign locations.

1. Consumers worldwide increasingly determine products and technologies, but there are significant regional differences in acceptance of new technologies (e.g., nuclear power, genetically modified organisms and foods, and medical devices). U.S. firms need to be aware of these differences in conducting R&D as well as in marketing.

2. Environmental concerns and the need for cheap sources of energy will be a major driver of technical developments as a result of both economic and political pressures. The U.S. economy is built on the assumption of cheap energy. The price of energy cannot be sustained without significant new technical developments.

3. Dramatic and unforeseen events will also affect the adoption of technologies. For example, according to public opinion polls there, the energy crisis in California has increased public acceptance of additional power from conventional and even nuclear sources.

4. There will be growing concern about the impacts of emerging technologies on values important to citizens—for example, the impact of information technologies on privacy and biotechnology’s ability to genetically modify humans, animals, and plants—although the real risks are sometimes misunderstood:

   • There is often a difference between the public perception of a risk and the objective assessment of that risk. These differences can lead to overestimates of risk (as in the case of nuclear power) and underestimates (as in the case of large software programs). The public, including the media, has the unrealistic expectation that risk should be zero, and it must be educated to better understand risk.

   • In the United States, there is now general acceptance of genetically engineered drugs, such as t-PA and insulin, but growing concerns about genetically modified organisms and food.

   • Free markets tend to respond quickly to needs, and public attitudes in a democracy may be quickly modified in response to any perceived crisis. Transparency and real-time availability of information will have an increasing impact on the development of public perceptions and therefore political views.

5. Concerns about the impacts of technologies vary from group to group within the United States and between nations:

   • There are likely to be sensationalized negative portrayals of new technology, leading people to unreasonably fear that they will lose their privacy and control over important parts of their lives.

   • There is a growing disparity between groups that benefit from technological advances and those that do not, which may lead to backlash against some new technologies by those not benefiting. Increasing cultural diversity within the United States and in world markets for U.S. goods and services makes reactions to events unpredictable.

• Different generations of Internet users may have substantially different sensitivities to the privacy (and other) implications of new technologies.

• Globalization may cause the United States to alter its positions on research and technology developments in response to the attitudes of other governments and cultures.

• The degree to which new technologies are adopted will be impacted by the technical literacy of a society.

6. Popular views of technology may impede research in certain areas but promote it in other areas:

   • A substantial proportion of the public objects to the conduct of stem cell research and the creation of genetically engineered organisms (impede).

   • Social issues have caused some medicines to be removed from the market (impede).

   • Nuclear energy development in the United States has been inhibited, forcing the pebble-bed reactor to be built in South Africa (impedes some energy technologies while promoting others).

   • California regulations may delay the acceptance of improved diesel engines in the United States (impede).

   • Environmental concerns drive new market opportunities such as the development of recyclable polymers (promote).

   • Green manufacturing and sustainable development will be demanded (promote).

7. There is a need for honest brokers to gauge the risk for emerging technologies and help build public confidence:

   • Society needs mechanisms to instill trust and defuse the threat of the unknown. The National Academies have the potential to assist in this area by conducting reviews of technology impacts, but additional approaches to creating confidence in new technologies will be required.

   • The science and technology communities have done things to undermine public trust. They need to learn what they must do to generate and maintain trust. This will require the fair communication of diverse viewpoints.

8. Human—machine interfaces need to improve generally to increase the use of, and therefore the demand for, high-tech products and services.

9. Careers in science and technology are not as valued as they should be (and once were) in the United States:

   • Improving the social status of engineers and scientists will increase the number of people who pursue careers in these areas.

   • The number of trained scientists and engineers may not be adequate if foreign-born scientists begin to return home or stay home instead of seeking education and employment in the United States.

   • The stresses and demands of technical development jobs may lead to burnout and early retirement.

1. There will a dynamic interaction between technology developments and the structure of the economy, each driving the other at the same time as it is shaped by the other.

2. Opportunity (development of the Internet, for instance) and need (for, say, multidisciplinarity or life-long learning) will drive significant changes in education—for example, much improved online courses.

3. Environmental concerns should stimulate demand for technologies that increase efficiency, although this demand is more likely to come from industry than from consumers.

4. Opposition to nuclear power may abate as the economic and environmental risks of fossil fuels are understood.

5. There will be considerable demand for improved health technologies, and the direction of biomedical technology will influence the structure of health care itself:

   • Health-care cost pressures are accelerating, in part as a result of technological advances, and those pressures are compounded by trends in demographics (aging of the population in the United States and other advanced economies).

   • There will be demand for improved connectedness and communication in medical records and for the use of remote medical technologies.

   • Reimbursement policies will impact the decisions to support new biomedical technologies.

   • Advances in biomedical technology tend to increase costs, while advances in other technologies tend to decrease cost. Can this be changed?

   • Technology can improve the health-care delivery system in areas such as record keeping, diagnostics, and so on, but this improvement may not decrease total cost if patients come to take for granted each new technical development.

• There will be growing demand for technologies that increase health care autonomy (self-monitoring, for example).

• Product liability may inhibit the incorporation of new technology into invasive biomedical products.

6. Although manufacturing may be declining relative to services, it is still a sector in which R&D and innovation are important to the national economy.

   • Economic forces will push manufacturing closer to the source of natural resources and to end-use markets. This may reverse the flow of manufacturing from the United States in certain industries.

   • The trend toward outsourcing by major corporations is being extended to R&D.

7. The U.S. economy promotes capitalization of innovation, which in turn facilitates the adoption of new technologies. This advantage of the United States will be increasingly shared by other economies as they adopt structures that can capitalize innovation.

8. Economic rewards would come from innovation in other areas as well:

   • Greater attention to standardized ways of presenting/mining information and retrieving information online.

   • Improvements to the modeling and simulation of economic systems.

1. The globalization of science and technology will have a major impact on the economy and the development of technology.

   • Multinational corporations already conduct borderless R&D.

   • A greater proportion of fundamental research may be conducted offshore.

   • Dramatically lower costs can be realized by relying on highly educated populations in India and China instead of the United States.

2. Advances in other areas would have significant impacts, positive and negative:

   • Communications standards. International standards will encourage development because they will generate confidence that the development will be used.

• Environmental regulations. International agreements may encourage some development and inhibit other development, depending on the needs generated and the activities prohibited.

• Intellectual property policies. Lack of intellectual property protection will inhibit development.

3. More consistent intellectual property protection would encourage investment in R&D:

   • Country-to-country differences in patent policies are a problem, and increased harmonization is needed.

   • First-to-file patent policy followed by most other nations would create a bias in favor of larger organizations and might inhibit innovation. (Not all members of the subgroup agreed with this forecast.)

   • Inconsistent interpretation of fair use of databases impedes scientific research.

4. The European Union is becoming a second large, homogeneous market. It has more aggressive environmental regulation and is driving international standards that could become trade barriers.

5. International cultural differences are important. The United States excels in creative industries (e.g., software, entertainment), Japan in incremental improvements (e.g., consumer electronics).

6. A number of technological developments, including the following, will be either driven or impeded by national security concerns:

   • Encryption technology (impeded),

   • Commercial use of high-precision Global Positioning System technology (impeded and driven),

   • Electromagnetic wave absorbing coatings (impeded and driven),

   • High-performance computers (impeded),

   • Satellite technology (impeded and driven),

   • Explosives detection (driven), and

   • The Internet (driven).

7. There will heightened concern about dual-use technologies that can also be used by terrorists in weapons of mass destruction: biotechnology and nuclear technology, for example.

There was an extended discussion of the public response to technological change and research risks and what explains that response. Public opinion in

France, for example, is opposed to genetically modified organisms but is not very concerned about nuclear power, which provides 75 percent of France’s electricity. U.S. public opinion turned against nuclear power after the Three Mile Island incident, although no one was injured, but it is not very concerned about genetically modified organisms. The German public is much more accepting of new medical devices than the U.S. public. There was consensus that most people do not understand statistics, leading to a gap between perceived and objective risk. For example, airline safety is much more stringently regulated than automobile safety, and many more people die in car accidents each year. Some workshop participants argued that the public is more rational than most scientists and engineers give it credit for. In the case of automobile safety, people have much more choice than they do in the case of airplane safety. Although the idea that scientists and engineers are rational and nonscientists are not is obviously simplistic, there was consensus that public understanding of science and technology should be increased.

Another discussion ensued, this one on the difficulty of agreeing on the facts when negotiating international treaties, such as those on controlling global warming, biosafety, and patents, and on what the scientific bases for international standards are or should be—for example, should the Kyoto Protocol be based on the findings of the Intergovernmental Panel on Climate Change? The U.S. approach to setting product standards, working through the American National Standards Institute, is more private and voluntary than the European approach.

Inequities such as the digital divide here in the United States and the implications of technology transfer for developing countries were discussed in the light of new threats of terrorism. The consensus was that the pressure for technology transfer will increase and that modernizing trends in many countries will make them more capable of incorporating technology; however, governmental constraints dictated by foreign policy or national security considerations will also be an important factor.