Revolution in the U.S. Information Infrastructure (1995)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

The Evolution of Information Infrastructures: The Competitive Search for Solutions JOHN S. MAYO While reflecting on what might be said about the “search for solutions” noted in the title of this paper, I recalled the story of the young executive who was about to leave the office at 6:00 p.m., when he found the boss standing with a piece of paper in front of the shredder. “This is very important, and my secretary has left,” said the boss. “Can you make this thing work?” “Certainly, sir,” said the young executive. He turned on the machine, inserted the paper, and pressed the start button. “Excellent, excellent!” said the boss as his paper disappeared inside the machine, “I just need the one copy.” The message of this story is that a solution that has worked well in the past might be totally wrong for the problem at hand. This message also applies to the search for information infrastructure so- lutions. With this in mind, I plan first to examine the driving forces that are propelling the emerging multimedia revolution and the evolution of information infrastructures. Then, I will discuss where these forces are taking us: to the National Information Infrastructure (NII) and, ultimately, to the Global Information Infrastructure (GII). Finally, I will briefly examine the consequences of this powerful and pervasive technological change on those involved in research and development. 1 

2 JOHN S. MAYO FORCES DRIVING CHANGES IN THE INFORMATION INFRASTRUCTURE It is no secret that information technology is the driving force and the key enabler of the emerging multimedia revolution and the evolution of the NII, as well as a host of other advances that together are changing the way we live, work, play, travel, and communicate. We have seen technological capability double every year in certain fields, such as computing and photonics, and double every 18 months in microelectronics. Even software, once a “bottleneck” technology, is beginning to advance rapidly in major areas like telecommunica- tions, thanks to object-oriented programming and reuse of previously developed software modules. Microelectronics In microelectronics, we have witnessed the exponential growth of the familiar Moore Curve as the number of components per chip moves steadily toward known physical limits. In the early part of the next century, familiar “bulk effect” solid-state devices may mature with transistors that measure about 400 atoms by 400 atoms each— the smallest such transistor likely to operate reliably at room tem- perature. The new frontier then will not involve making the devices smaller, but in creatively and economically using the vast increase in complexity and power made possible by this remarkable technology. The amazing progress of microelectronics represents a micro- cosm of the broad thrust of information technology and the other key driving forces made possible by information technology—all the most vital forces that are leading to the multimedia revolution and the evolution of the NII. Let me examine the progress and impact of these related forces. After the invention of the integrated circuit, every time the com- plexity of silicon chips increased by a factor of a thousand, some- thing had to be re-engineered. The first re-engineering that was done, as we headed toward that first thousandfold increase, was to change all of our design processes, which had been based on discrete compo- nents. When we reached 1,000 components per chip, we used the new digital circuitry to re-engineer our products from analog to digital, as 

EVOLUTION OF INFORMATION INFRASTRUCTURES 3 did many other industries. Let me stress that this early progress toward digital products, enabled by microelectronics and software, has brought about the digitalization of most systems and services domestically and, more and more, internationally. Digitalization is a powerful force driving us toward multimedia communications and the NII. About a decade ago, it became possible to make chips approach- ing 1 million components, and this brought us powerful microcom- puters, along with all the peripherals related to microcomputers and the necessary software systems. This led, in turn, to an explosion of advanced telecommunications services that forced AT&T to re-engi- neer itself from a company that provided largely voice and data-on- voice telecommunications services to one focused on universal infor- mation services, or the provision of voice, data, and images anywhere, anytime, with convenience and economy. Providing advanced ser- vices on an increasingly intelligent global network was the beginning of multimedia communications, which has now become the revolu- tion of the 1990s and beyond. We are currently experiencing another thousandfold increase in components per chip. Re-engineering has now extended beyond our company and is leading to the merging of communications, comput- ers, consumer electronics, and entertainment. The convergence of these four industries is being accomplished through joint projects, joint ventures, mergers, acquisitions, and some start-up companies. This industrial re-engineering appears to be the next-to-last stage of the information revolution brought on by the invention of the transistor. The last stage, and one that may go on forever, is the re-engi- neering of society—of how we live, work, play, travel, and commu- nicate. Education will change with distance learning and home schooling; virtual offices and working at home will transform our work lives; routine tasks such as visiting and shopping more and more will be done from home. Let me add, however, that many of these changes will be generational; social change, as well as technol- ogy, is needed to make them happen. Speech Processing The trend toward increasingly powerful silicon chips is enabling exponential increases in the processing power of fixed-point digital 

4 JOHN S. MAYO signal processors. This trend makes it possible to use much less processing power to process a specific speech-recognition or speech- synthesis algorithm. Today, for example, a single chip can recognize a 100-word vocabulary, a function that would have required 20 to 30 chips 5 years ago. Based on these advances, we can project that speaker-indepen- dent automatic speech recognition and synthesis will become com- monplace. These technologies will mature in a variety of service capabilities based on the ability of intelligent machines to talk and listen much as people do. In addition, speech in one language will be automatically translated into a second language, which might then be synthesized with the voice characteristics of the original speaker. Virtually unrestricted recognition vocabulary will permit natural lan- guage interaction—with humans, machines, and databases. Automatic speech recognition is vital to the multimedia revolu- tion not because it replaces human operators, but because it is a powerful technology for making multimedia systems easy to use. Ease of use is an imperative for broad marketplace success. Image Processing The emerging technology of image processing is related in many respects to speech processing and is another driving force toward multimedia communications and the evolution of our information infrastructure. The same key information technologies that are en- abling progress in speech processing are also fueling rapid progress in image processing, especially in the important areas of image and video compression. Consider video compression. At the network capacity of two Integrated Services Digital Network (ISDN) Basic Rate Interface channels, or 128 kilobits per second, we can have video that is com- pressed using the so-called p X 64 international standard for video- telephony, for person-to-person communications. This standard was designed for sending videotelephony over phone lines in multiples of 64 kilobits per second. At this capacity, we have the ISDN video- phone, which is likely to be a highly attractive product as long as copper wires make up the local telephone plant. At the network capacity of 1.5 megabits per second, we can have high-quality desk- top video compressed with the maximum level of the p X 64 stan- 

EVOLUTION OF INFORMATION INFRASTRUCTURES 5 dard. This network capacity can support National Television System Committee (NTSC) color television with VHS quality, using the Moving Picture Experts Group (MPEG) standards, which are de- signed for capturing moving images on CD-ROM and for storing and transmitting movies and broadcast television. However, with avail- able compression technology, a capacity of 1.5 megabits per second would not deliver broadcast quality for scenes with rapid motion and detail. The network capacity of 45 megabits per second will readily support high-definition TV (HDTV) compressed to about 20 mega- bits per second, using variations of the MPEG-2 coding algorithm. That is a considerable reduction (about 50:1 compression) from the roughly 1 billion bits per second required for uncompressed HDTV. That same network capacity will also support near studio-quality video (HDTV or NTSC) that has been compressed by using varia- tions of the identical coding algorithm. Common Standards Another force driving the progress of multimedia communica- tions and the evolution of the information infrastructure is the world- wide push toward common standards that will encourage global net- working. Photonic transmission facilities, for example, will be based on the evolving international standard called SDH, for Synchronous Digital Hierarchy. Because SDH defines standard network inter- faces, service providers and end users will be able to use equipment from many different vendors without worrying about compatibility. SDH will provide efficient transport of broadband services and will simplify networks. Similar standards in domestic networks will al- low digital communications into the workplace and home, and they will make possible services dependent on high rates of data transmis- sion. Broadband Integrated Services Digital Network (B-ISDN) is a new digital format as well as an international standard that supports multiple services, such as voice, data, and new video services, using fiber-optic transmission facilities. B-ISDN is currently defined at interface rates of 155 megabits per second and 622 megabits per second. Based on the fast-packet technology of asynchronous trans- fer mode (ATM), B-ISDN could herald an exciting new era in global 

6 JOHN S. MAYO communications networking, as equipment vendors and service pro- viders adopt compatible standards to provide sophisticated high-band- width services. Market Demand The demands of the marketplace, more than technology, set the pace for the multimedia and infrastructure revolution. For the greater part of this century, the customer willingly accepted whatever tech- nological capabilities we were able to achieve. Thus, the telecommu- nications industry was supplier driven, and the suppliers managed the evolution of the industry and the NII. However, the technology became so rich that it made possible many more products and ser- vices than the user could accept or was willing to pay for. That marked the transition from a supplier-driven industry to today’s cus- tomer-driven industry—from supplier push to marketplace pull. Competition The global transfer and assimilation of information technology, along with political and regulatory forces, such as the move to priva- tize telecommunications around the world, are leading to strong in- ternational competition in providing communications products and services. Such pervasive competition is another powerful force driv- ing the evolution of both multimedia communications and informa- tion infrastructures. The public policy challenge is to provide a frame- work in which that evolution may occur. THE MULTIMEDIA REVOLUTION Let us look a bit deeper into these subjects, starting with the multimedia revolution. After all, the pursuit of multimedia is creat- ing social pressures on the NII and the information superhighways. So, what is “multimedia”? A reasonable working definition is that the term multimedia refers to information that combines more than one medium, including speech, music, text, data, graphics, fax, im- age, video, and animation. At AT&T, we tend to focus on multime- dia products and services that are networked, or connected over a communications and information network. Examples of this range 

EVOLUTION OF INFORMATION INFRASTRUCTURES 7 from videotelephony and videoconferencing; to real-time video on demand, interactive video, and multimedia messaging; to remote collaborative work, interactive information services such as elec- tronic shopping, and multimedia education and training. Eventually, we will have virtual reality, which will enable people to experience a place or an event indirectly and remotely—and do so in all dimen- sions. We are excited about multimedia because public switched net- works, or information infrastructures, can currently accommodate a wide array of networked multimedia communications. Given the di- rections in which those networks are evolving, they will be able to handle an increasingly vast range of such communications. More- over, a potentially enormous market is out there for multimedia hard- ware and supporting software. Although estimates differ widely, the most commonly quoted projection for the total worldwide market for multimedia products and services is roughly $100 billion by the year 2000. AT&T is playing a major part in facilitating the emerging multi- media revolution: We provide services, network products to local service providers, and products to end users. These are familiar roles for AT&T, so let me briefly describe another perhaps less well known aspect of the multimedia revolution that we are studying. That role is as the host for a wide variety of digital content and multimedia appli- cations developed by others. Hosting is a function that connects end users to the content they seek. Customers will gain easy and timely access to personal communications, transactions, information ser- vices, and entertainment via wired and wireless connections to tele- phones, handheld devices, computers, and, eventually, television sets. Sources for this digital content will range from publishers and large movie studios to small software houses. Global standards and open critical interfaces are vital to this complex hosting function. The entertainment industry, for example, must have software systems that are compatible with those of the hosting industry. These software systems must, in turn, be compat- ible with those of the communications and information-networking industry, which must be compatible with customer premises equip- ment and systems. In the age of multimedia communications, consumers who are geographically separated from each other will do more than just play 

8 JOHN S. MAYO games together over networks. According to the AT&T vision, people will increasingly seek new relationships based on “tele- presence”—a new type of community and social experience indepen- dent of geography. This potential for interactive networks is quite unlike what will result from the 500 preprogrammed cable TV chan- nels proposed for the United States. The beauty of interactive net- works is that consumers will have the freedom to choose any subject or service from the intelligent terminals in their homes. And they will be able to network clusters of friends or associates to enjoy such services as a group. Although I have focused heavily on the impact that multimedia will have on the consumer, networked multimedia communications will dramatically change the nature of work and will therefore have a broad effect on business. Videoconferencing, for example, can en- hance productivity, save time, and reduce travel. Moreover, current developments in multimedia telephony are making realistic the possi- bility of remote collaboration. In a few years, a person could be working in real time with colleagues in New York, Washington, Hong Kong, Paris, and Sydney. They could, for example, accom- plish the task of producing printed materials, presentation slides, and a videotape introducing a new product line. EVOLUTION OF THE NII The quest for multimedia is driving social issues that relate to the NII and the information superhighway. The NII might be viewed as the superhighway plus all the terminals and databases connected to it. What is AT&T’s vision of the NII? It is to bring people together, giving them easy access to each other and to the information and services they want and need, any time, anywhere. In AT&T’s view, the NII is a seamless web of communications and information net- works, computers, databases, and consumer electronics, which will put huge quantities of information at the fingertips of a variety of users. Quite simply, we see the communications component of the NII as a vast interoperable network of local, long-distance, and glo- bal networks; wireless; broadcast and cable; and satellites. In addi- tion, the NII encompasses the Internet as well as the test beds associ- ated with the High-Performance Computing Initiative, such as the 

EVOLUTION OF INFORMATION INFRASTRUCTURES 9 Blanca test bed with which AT&T is affiliated. However, the NII is not a uniform end-to-end network developed and operated by gov- ernment or any one company. It is the totality of networks in our nation, interconnected domestically and internationally. These networks consist of a wide variety of technologies that provide a complete range of features and transport speeds. In the United States, competitive backbone information superhighways of optical fibers are already in place and expanding rapidly. However, access to the backbone is still largely via copper wire pairs, espe- cially for homes, schools, and small businesses. Many consumers and small firms will be able to get substantial value from the NII using recent technological advances, such as ISDN. Genuine and effective competition in the local exchange is a long-term possibility and is the key to advancing the deployment of needed digital and higher-bandwidth access services. We also believe that the coaxial cables of the CATV networks can become important elements of the nation’s broadband superhighways. One key to making this network of networks a true global infor- mation superhighway for multimedia and other communications is a system of open, user-friendly interfaces and global standards. Such a system both promotes maximum interoperability and connectivity and supports a multivendor environment that allows maximum cus- tomer choice of equipment and services. To ensure that people can use whatever NII capabilities they need when they need them, a range of bandwidth offerings should be made available. Several policy issues are associated with the NII: (1) The government should help provide the vision for the way the NII will evolve and operate. Private industry should build, own, and operate the NII under competitive conditions. (2) The government should provide incentives and opportuni- ties, such as tax credits for investment in equipment, R&D, or worker training, to encourage the private sector to invest in and deploy new information technology. The government should continue to focus its direct support on precompetitive projects or applications that dem- onstrate and test new NII technologies. (3) In order to realize the broad benefits of the multimedia revo- lution and the NII, there must be full and fair competition in all 

10 JOHN S. MAYO sectors of our communications and information industry, both do- mestically and internationally. (4) Private industry must continue to play the lead role in work- ing to define the interoperability standards of the NII. The govern- ment must continue to work with and represent industry in some international treaty organizations that develop standards. Jointly, we need to speed up the standards-setting process. (5) Perhaps the broadest policy challenge is that of facilitating public acceptance of and comfort with the benefits of multimedia communications and the NII. This will require setting ground rules to protect electronically available intellectual property, creating a framework to ensure privacy and security of information, and estab- lishing mechanisms to make these benefits available to the largest number of users. This last point is vital if we are to avoid having a nation of information “haves” and “have nots.” The challenge, sim- ply put, is to develop a new definition of universal service. Fortunately, most of the key issues and challenges surrounding multimedia and the NII are already being addressed by cooperative government and private-sector efforts. Much more remains to be done. The ultimate resolution of these issues will require the support of all of us. IMPACT ON R&D At the beginning of this paper, I noted that a solution that has worked well in the past might be totally wrong for the problem at hand. This applies broadly to R&D and R&D environments. The forces behind the emerging multimedia revolution and the evolution of information infrastructures, along with the benefits associated with the information age, are also bringing about a new paradigm for R&D and a new R&D environment. The traditional R&D approach of past decades was to use proto- typing and redesign: to do, then learn, and then redo as often as necessary to produce a good product, and to do so through serial handoffs from one function to another. Today, that approach is not competitive because of the increasingly complex and rapid advances of information technology, multimedia communications, and infor- mation infrastructures. The marketplace demands more and cheaper 

EVOLUTION OF INFORMATION INFRASTRUCTURES 11 products and services with ever-shorter delivery intervals. In addi- tion, an increasing demand exists for the customization of products to ever-smaller buying units. Thus, in contrast to the traditional R&D approach, we must pur- sue concurrent engineering, together with the important integration function of deciding and planning precisely what to do before we build anything. Instead of talking to the customer after we complete a project, we do so before starting, because the customer is really in control. The vital integration function, which helps ensure that the product or service works the first time, is becoming an ever-larger piece of the R&D effort. Functionally structured R&D has yielded to customer-focused, multifunctional teams. This shift has helped generate a new R&D paradigm and a new R&D environment with very different processes and demands on people. Let me illustrate with the typical challenge of trying to pursue a number of R&D projects at once. Within today’s lean, cost-effective R&D organizations, there are usually many more projects than there are expert staff. Therefore, dedicating these ex- perts to specific projects would result in poor overall project perfor- mance. Instead, support systems must facilitate the training and shar- ing of expertise, most often through well-orchestrated multifunctional teams and new techniques for learning, including a strong focus on intergroup learning. Another key element of the new R&D paradigm and its associ- ated environment is a focus on reusable assets, such as previously designed and tested software modules and hardware platforms. Re- usable assets facilitate the creation of products and services from higher-level elements. They enable organizations to share key skills across projects and create world-class experts, while also developing rapid and cost-effective solutions, customizations, and competitive advantage. Reusable assets are the most powerful mechanism we have today for improving R&D productivity. Proper use of such assets can make possible perhaps as much as a tenfold improvement in design productivity. The reusable-asset concept also extends to families of reusable processes, which unify architectures and facili- tate efficient realization of products and services. Structuring R&D around reusable assets produces an environ- ment in which research is more important than ever. A major part of doing good research is finding good problems on which to work. 

12 JOHN S. MAYO Bell Labs has never had to look far; every day we face many real problems for which we do not have good solutions. Moreover, re- search results are themselves reusable assets; they help reduce the development effort and shorten the development interval. The new R&D approach also yields lower costs and higher quality. The new R&D paradigm and R&D environment cannot by them- selves fulfill the demands of the new marketplace paradigm of growth through new corporate structures, new customers, and new competi- tors. R&D, however, must play the important role of identifying the new opportunities and parameters that a corporation must pursue with its total resources. SUMMARY Rich information technology, the worldwide push toward global standards, ever-increasing customer demands, and growing interna- tional competition are key forces behind the emerging multimedia revolution and the evolution of national information infrastructures. The growth of multimedia communications and the further competi- tive evolution of our National Information Infrastructure, as well as the Global Information Infrastructure, raise some difficult issues and challenges, but these advances promise a broad range of information- age benefits to virtually every citizen of our nation. Already, we find that yesterday’s R&D solutions are no longer adequate for the revolution at hand. A new R&D paradigm for a new R&D environment is developing rapidly. And each and every one of us, especially those in R&D, must understand and assimilate this new paradigm. We must ensure that it both serves society well and cre- ates an environment in which the best minds of the world seek ca- reers in science and engineering. Long-term progress depends on it, and it can be done.