I. INTRODUCTION

--------------------- * This presentation is a modified version of a talk given at the National Geographic Society Seminar, Washington, D.C., January 26, l988 our work focused on efforts to control a galloping scientific enterprise and its interaction with the affairs of state. They emerged again later when I helped create the Congressional Office of Technological Assessment

our attention focused on two major problems--the arms race and the undesired impacts of technology on the society. Much of what I will say today stems from thinking about those pathological conditions in the exploitation of science in technology. I have studied communication processes in man-made and living systems for most of my professional career. So it was natural that many aspects of the behavior of social systems appeared to be understandable, even obvious, if one just considered them as complicated information systems with all the normal characteristics of an information system such as bandwidths, time constants, positive and negative feedback, information processing abilities, including the ability to learn, and degraded performance in the presence of noise. When I found that I could not come here I put the plan to develop these ideas aside. Recently, however, I was invited to participate in a symposium celebrating the one hundredth anniversary of the National Geographic Society, a symposium in which the focus was on the deteriorating condition of the planet and what might be done to rescue it. I was asked to talk about the impact of technology on our global home, and that gave me the needed stimulus to explore my ideas. For many years, my own special efforts have been to understand and attempt to eliminate the nuclear arms race. To pursue my own interests I would have focused especially on that problem, but because of the geographic and ecological focus of the symposium I was addressing, I gave special attention to the impact of technology on the environment. But, as you will see, I concluded that the arms race is a major piece of the problem. As part of that paper, I also explored a slightly whimsical but hopeful scenario that flows from understanding what is going on.

The planet with its billions of people and countless other living creatures, and the complex ecosystem and social systems that supports them, is in chaos. The extent of the global disarray is only dimly perceived because individuals everywhere tend to regard the troubles they face as local or national phenomena, rather than as the worldwide problems that many of them are. Scientists are just beginning to realize that the earth in reality is a single complex organism made up of living creatures, the land, the sea, and the sky woven together in delicate balance by a combination of natural and man-made links. The recognition of this oneness, this interdependence, is so recent that we don't have a name for the collective "us". This is the most positive thing that the space flight has given humanity. We are just beginning to realize, too, that this wonderful oneness is critically ill. How ill? No one knows.

Humanity is suffering from a series of diseases most of which have no ethology, no previous history, and few well-informed doctors. Many concerned individuals believe that the world is approaching a critical point of no return where conditions will run relentlessly downward. Fortunately, people are also beginning to sense the extent of the threats we face. Also, these global diseases, ranging from the nuclear danger which promises sudden and total devastation, to the many ecological and social diseases, could create extremely inhospitable environments and declining conditions of life, but life, even human life, will go on in some condition. At what stage, if at all, the situation crosses the threshold into a state of steady decline, as some people fear, is not yet clear. In parallel, there is a growing awareness that although the scientific culture, through the large-scale applications of technology, is responsible for most of the threats to survival, the only solutions will be found in a judicious mixture of restraint and new technologies focused on improved life-support systems and global resource management.

The challenge, I concluded, is for those of us now alive, to mobilize; to start building a world that is in ecological and human equilibrium. The resources exist, but the desire and knowledge to do it do not. Scientists and engineers can, I believe, provide the "how to". Only an informed and dedicated population can provide the "want to"; a response that needs to be the moral equivalent of the nuclear arms race in its intensity and dedication; a response that starts by eliminating the arms race. None of this can be done without an adequate understanding of the processes at work, how the applications of technology occur, and how they could be moderated, and equally important, how human beings fit into the process.

When I began planning this paper, as I said earlier, I intended to develop some ideas that I have long harbored about society as a learning system that are based on cybernetic concepts. As I proceeded, I realized that the understandings that this provided could also indicate how it might be possible to plan a rescue mission for the planet and arrangements for a better world. I began to outline how this might be done and soon discovered two things, first that I had begun what should be a major interdisciplinary study and, second, that I was far exceeding the time available to me. Rather than omit the second part, I have shortened the entire paper by eliminating some details and examples. I have found the ideas so interesting, particularly those relating to the role of human beings in the system, that I plan to continue this exploration.

I realize that the plan I am going to present is incomplete. I am also aware of the fact that as vital as the problems that I am talking about are, you have a more urgent agenda here in Israel. I believe that some of the same characteristics of the human mind that lead to frozen positions in the nuclear arms race and other social pathologies that I will mention briefly, are at play in the inability to find alternate solutions to the awful situation that exists here.

My original intention was, as I mentioned, to use information and learning theory concepts to explore why these problems--what we might call man-made plagues--occur and why the emergence of corrective measures is so slow and inadequate. These ideas will help understand why so many inventions and discoveries that are so promising initially, lead to the kind of difficulties I have listed, and why both capitalist and socialist states have become increasingly unmanageable as the extent of their technologies and their systems for controlling the society have grown. Obviously, in Israel size is not a problem unless it is smallness. But like it or not, we are all closely coupled with the other inhabitants of the planet. Also, if the truth were known, most of us contribute a proportionate share of poison to the ecosphere, and the regional conflicts are certainly potential sources of global conflagration.

II. EVOLUTION AND LEARNING

Basically, I believe that the use of technology to improve the human condition is an experimental process, i.e., one of trial and error, that needs to be better understood if it is to be properly directed. For too long we took advantage of the positive contributions of technology and ignored the dark side. In fact, the dark side wasn't always evident, perhaps didn't even exist. When there were only a few thousand automobiles, for example, pollution was not a problem and those who projected the environmental and social impact of their mass deployment were regarded as cranks. But there has been growth in many dimensions and with it an exponential increase in the destructive impacts of technology, which has brought humanity to a crisis point. All experimental activities are learning processes, inherently involving trial and error. Many unanticipated problems could be avoided, or at least recognized and countered more rapidly, if their experimental character was widely appreciated.

Even in those societies that appear to be traditional and untouched by technology, it has often had a disruptive effect. For example, just a small dose of public health was enough to cause a dramatic increase in populations in many traditional societies.

Charles Darwin's theory of biological evolution, by now very well established, maintains that living organisms adapt to their environment as a consequence of random changes, mostly small changes of genetic origin, in which natural selection bestows dominance on those species that are best adapted for life in the prevailing environment. This process of random selection requires many generations. Man-made interventions via technology have displaced the slower natural process, vastly speeding up the rate of change, making biological adaptation unable to compensate for the effects of technological change, at least in the short term. Cultural anthropologists maintain that there is a parallel cultural adaptation through which individuals learn to adapt to many man-made changes and could obviously continue to do this in the face of deteriorating conditions. The extent to which this is true remains a question of considerable controversy.

Humans have a singular advantage in the evolutionary trek--intelligence--which confers the ability to adapt quickly, particularly through the invention of mental models, tools, and techniques. As providential as this gift has been, it must also be regarded as dangerous because until now such purposeful adaptive activities have left a trail of ever more difficult problems. It is my hope that with a proper attention in the design stage the need for large-scale, after the fact, correction can be minimized.

III. GENERAL DISCUSSION OF INFORMATION SYSTEMS: PROBLEM SOLVING AND LEARNING IN A SOCIAL SYSTEM

By a society's information processing capability, I don't mean its communication technology, but rather the way in which its individuals and groups are organized to do their work and the ease with which they can interact, learn, and change based on their experiences.

The two concepts, technology as the source of a man-made evolutionary process, and a technological society as a learning system, together serve to explain why problems arise from the deployment of new technologies. Seeing the problems in this light helps to understand how with more effective feedback their effects could be minimized or even avoided. They also make evident why the steady stream of new technologies shapes and reshapes civilization, altering the physical environment, and constantly posing new intellectual and psychological hurdles for people.

The collective task then is that of guiding the man-made evolutionary process, one of such complexity that its future state can never be predicted.

An important point to keep in mind about this situation is that the use of technology to enhance human capabilities is not entirely a random process, but neither is it entirely a logical process. In fact, it appears to be a combination of the two. At the micro level new initiatives or experiments can be well ordered, but even then the outcomes often contain surprises. Well done scientific research clearly demonstrates this point. The goal is usually obvious. So may be the path to follow. But rarely do these starting points remain invariant. Some of the most important scientific discoveries have been the consequence of pursuing very different goals. This is usually the case in other human activities as well, including business and government. Too often individuals making decisions in all of the foregoing situations do not appreciate the uncertainties in their undertakings.

Each sub-group in a society has tasks on which it focuses and strives to fulfill. Sometimes a task is simple organizationally and relatively independent of the rest of the system, as in the case of a piece of scientific research; sometimes the task is hopelessly dominated by the conflicting demands of the complex social system. The overall performance of a society as a learning machine, then, is determined by the performance of all of its parts, interacting and influencing each other. Little wonder that outcomes, large and small, have a substantial degree of uncertainty and surprise. While it is impossible to predict with confidence the actual outcome of any initiative in this complicated, interactive world, it is possible to judge qualitatively the effectiveness of organizations involved in creative tasks by determining how well they satisfy the conditions for an effective learning system.

IV. PROPERTIES OF A GOOD LEARNING SYSTEM

1. Feedback. An essential element in any learning system is feedback. In fact, any goal-seeking system, living or man-made, must have a mean for comparing the present state of the system with its ultimate goal, i.e., a means of measuring the error between the present situation and the ultimate goal. Consider a person reaching for an object. As his arm moves toward the object, the eyes constantly compare the position of the hand as it moves and adjusts the muscle forces to lead to ultimate closure. In the same way a robot, whose task is to orient parts on a machine, compares the position of the part with the desired position--usually using a camera system for determining the error on location--and then with feedback through an actuating device takes corrective action until the error is as small as the system is capable of detecting. All learning systems require these elements, a goal, a means of comparing the present situation with that goal and a method for moving toward the correct situation. Intellectual activities that we call "learning" all involve this process. That is true whether the goal is learning the contents of a textbook or involves an effort to gain a new piece of information about nature. In the robot I have described, the task is clearly defined but the path followed to the objective will depend upon the starting conditions so in a very primitive sense the machine is learning to do one task over and over. Unlike living systems, the robot cannot change its basic method of operation although some machines can indeed adapt to changing task conditions. A sophisticated telephone network is an example of this.

It is evident that for optimum performance the feedback system should be as responsive as it is possible to make it. i.e. its measuring system and computer that determines the appropriate path should be able to detect small errors. In practice there are limits to the degree to which this goal can be met without generating troubles, i.e., noise in the communication channels might be regarded as error signals if the system is too sensitive. In social systems, public opinion, new media, voting, public interest study and action groups, special interest lobbyists, the performance of the economic system, and many other matters serve as feedback. The social organization determines the responsiveness of the system to each of the feedback signals.

2. New initiatives. i.e., experiments, should be kept as small as is consistent with achieving a chosen goal. This is desirable both so that a trial can be made quickly and the consequence of any error is minimized. Doing this, however, slows down the ultimate implementation of new techniques, so every undertaking represents a compromise between the two objectives.

3. Size of experiments. Experiments should be as small as is consistent with effective learning. Because most learning systems combine elements of directed or logical behavior with trial-and-error learning activities, a number of course changes will undoubtedly be needed to reach a satisfactory goal. New initiatives should be arranged to perform individual experiments quickly or, alternatively, do many of them at the same time. The more urgent the need for an answer, the larger the number of experiments that should be done at the same time. (It is important to realize that a scale change for an old technology is also an experiment. Factors of two often totally change the performance or response of human systems.) It is quite common for hundreds of people to be working independently on a major scientific question at the same time. To spur the development of socially-relevant technology, many independent groups should be encouraged to seek solutions to the problems. To my knowledge, this factor has never deliberately taken into account the design of social experiments. In fact, most of the time people involved in making social innovations don't appreciate the degree to which their undertakings are experiments with perhaps more possibility of going wrong than right. Multiple experiments occur automatically in the private sector of a capitalist society where the market forces bring large numbers of competitors into any promising field. The experiments are essentially the products they develop and the measure of success is the market response. Typical of this phenomenon is the situation in the computer industry at the moment where hundreds of independent companies are producing small computers, and those that satisfy a sufficient need will survive while many of them will fade away. The situation is very different in the socialist countries where a few large organizations are responsible for meeting the society's need for such machines. An important element of Chairman Gorbachev's Perestroika is the encouragement of independence in individual enterprises. 4. Noise. Noise in the feedback system introduces random responses. Consequently, every effort should be made to keep it to a minimum so that a sensitive error-detecting system can be employed without the danger of provoking unwanted fluctuations in the system's behavior. The problem is tractable and manageable in man-made systems but has not been systematically considered in the study of societal problems. 5. Time Constant. A very important aspect of any feedback or control system is its time constant, the time it takes for it to detect an error, send a message about this error to the computer, which then computes what corrective action should be taken, and sends a command signal to the operating mechanism, which initiates the corrective action. In artificial systems this corrective process is usually continuous. In general, the goal of an equipment designer is to keep this time as short as possible to insure effective corrective action. If the time constant is too great, the error signal may come too late to initiate corrective actions. In this situation the response to errors will be sluggish, and under certain conditions may even cause the mechanism to amplify rather than correct the error.

These criteria show why very large systems, whether governmental or private, tend to be very inefficient. The feedback loops are very long and slow, error-detection is poor and requires large errors to stimulate a correction. In government, the ultimate feedback system is the citizen's response, and this tends to be a very ineffective process because even deciding what issues should be used for the feedback signal is difficult, whether it is the quality of government provided services, or military budgets, or economic decisions, or ethics, or many other issues of public concern.

IV. PEOPLE ISSUES

With regard to the survival issues we are considering, the majority of the feedback signals have come from the efforts of individuals and private groups who typically go into action only after situations become relatively threatening, in other words when the error signals become big. Furthermore, the messages from such groups have no place to go. That is, the error-correction focused on the survival issues is not an integral part of an organized society and has to be created de novo whenever a significant problem arises. This is a slow and ineffective process. That is why the problems we are concerned about have become so serious.

V. TECHNOLOGICAL RESPONSES

VI. HALTING THE ARMS RACE

From the very invention of the atom bomb, knowledgeable people like Dwight D. Eisenhower and Winston Churchill realized and said that there was no military use for nuclear weapons and that the only way to protect the world was to eliminate them. But instead of doing that, after the Soviet Union produced its bomb, American planners decided that they could best insure national security through more and better bombs and delivery systems. The rationalization for nuclear arsenals became deterrence--security under a nuclear umbrella. Even the present administration agrees with this proposition in its public statements, though it still seeks funds for a war-fighting capability.

Repeated studies in the United States and elsewhere have shown that a single nuclear weapon would suffice to devastate almost any city in the world. One hundred nuclear weapons properly targeted on either superpower will eliminate it as a functioning society. For most people, the destruction of a single city would be a deterrent against any form of aggression. Joseph Stalin's behavior in the Berlin crisis was restrained by the knowledge that the United States had a few primitive nuclear weapons. But that was before the explosion of the first Soviet atomic bomb. With that, an increased fear of Stalin and intelligence information that enormously exaggerated the Soviet bomber force caused a large buildup in the American nuclear forces. Since then, both participants in the arms race have continued to add weapon system to weapon system until the enormous stockpiles defy all understanding. Why is this? The military strategies of the superpowers have built in positive feedback.

Traditionally, nuclear forces have been designed to satisfy what is called worst case analysis, i.e., in the case of the American forces, to have the capability to deter an insane opponent bent on dominating the world even if the cost were the destruction of his country. This leads to sophisticated strategies and arcane analyses, based upon pure imagination, and conclusions like, "it require the capacity to destroy forty or fifty percent of the potential aggressor's population and property to constitute an adequate deterrent." Unfortunately, there are two players in this game. Why should a Soviet leader believe that an enormous American arsenal is designed to prevent something he has neither the capability nor intention of doing. So his worst case analysis requires him to match the original buildup. Because of the positive feedback nature of this encounter, there is no end to the spiral it generates until one, or the other, or both, run out of resources or interest. If civilization is lucky this may be happening.

But suppose we ask "Is this really the worst case?" The answer is clearly "no." To protect against a very unlikely event, the worst case scenario requires a very costly competition which gets increasingly dangerous.

The solution is simple. Abandon worst case analysis and its positive feedback syndrome and substitute a less dangerous strategic objective, a small secure deterrent force. Actually President Reagan and Chairman Gorbachev both articulated such goals in their talks at the Reykjavik summit meeting. Their objectives were correct. Drastically reduce the strategic nuclear forces and substitute information, i.e., feedback, to provide the needed security. This is easier to do than most people believe. The nuclear forces are so large that either side could unilaterally make substantial force reductions without violating the requirements of a prudent deterrent. Such an action would, in fact, be very stabilizing. Only after very major reductions would intensive monitoring of each by the other be necessary. There is one particularly difficult hurdle to be overcome on both sides, that is, the vested interests that have enormous stakes in the continued development, deployment, and operation of the military forces. To overcome this obstacle, one could involve the military-industrial-research community in the survival problems. The scientific, technological, production, and operational challenges in this field will prove to be greater, more interesting, more satisfying, and probably equally profitable to those of the arms race, and have even more growth potential-related problems.

VII. CHANGING THE GLOBAL AGENDA

I also assumed that if the arms race were halted some form of super-national authority would emerge--undoubtedly slowly-- to monitor the arms-limitation agreements and to provide the degree of cooperation and mutual assistance between nations required to successfully cope with the remaining survival problems. I did not explore what form that international authority should take. That, I think, must evolve, perhaps out of the United Nations, as the arms race waxes and international cooperation to save civilization becomes a major effort of the global community. This may be happening right now. Ultimately, this survival effort will require the involvement of all of the major industrial nations. Indeed, many of the problems that must be faced can only be handled through global cooperation. This is essential. It will be impossible for any single nation to maintain a healthy society in a collapsing world. Furthermore, the industrial nations produce much of the pollution that is poisoning the ecosystem, and, finally, they also are faced with the question of how to achieve an equitable and stable society with adequate educational and social opportunities.

The task of providing replacement and new infra- structures, will strain the available research, development and production capacity of the world. Among such demands will be: adequate and universally available educational systems, universal health care facilities, efficient and less polluting fossil fuel power plants, including plants that consume coal without emitting carbon dioxide, safe nuclear plants, super computers to model global climate, pollution, and even social problems, efficient solar cells, education technologies and software, and large-scale desalinization systems, among others. On the other hand, without these demands the industrial nations will be faced with a growing surplus manufacturing capacity and labor force. Advancing technology has brought the world to the point where its continued development requires the challenge of the global rescue project.

Do I think this can happen? Not now I fear. Perhaps when the dangers are more evident, when it is obvious to all that humanity was approaching the irreversible thresholds of the natural systems that sustain the human family. But then it may be too late. In spite of this small attack of reality, I began to list the things that should be started now and in the not too distant future, so that if civilization survives, and we meet again in one hundred years, there will be general agreement that the present time was indeed the low point in man's irresponsible use of his great powers of mind and its handmaidens, science and technology. As I said at the start, this won't be an operational plan for making this dream come true. I will only, in effect, provide some road signs.

I propose that this effort be planned and managed by a group of new agencies, the Departments of Development, or the D.O.D.'s, in the participating countries. In the early stages, much of their effort and resources would support research and policy studies to provide the basis for the subsequent action programs.

Are there adequate resources to sustain the action programs? I don't know, but a quick estimate shows that there is the potential for financing vast efforts.

If one assumes that the nations will make a transition from the present level of spending for military programs, i.e. between 5 and 8 percent of national GNP's, to a security force costing perhaps 2 or 3 percent, while the funds for the survival tasks including the appropriate research and development would grow, perhaps to 5 or 6 percent, this would provide more than one half trillion dollars. If one further assumes that worldwide excess production capacity would be used to help provide the new infra-structure, perhaps fifty billion dollars per year could be made available for research and development and at least two trillion dollars for new infra-structure. Some of these funds would be used by governments, but much of it could be used to build new capabilities in the private sector as well. Two trillion dollars sounds very large, but it only amounts to about four hundred dollars per inhabitant of the planet -- not much money for the task I have laid out but ever so much more than is now available.

One problem I will carefully avoid is how to provide the resources that will be needed, i.e. how to share the large economic burden. In World War II, such problems were avoided by the straightforward recognition that the allies would survive or perish together. While the situation the world faces today may be equally urgent, this is not recognized. Today we are not even able to use surplus food, or goods, or manufacturing capacity to help the poor nations. This is a political problem, not one of resource limitations. Perhaps the overriding necessity of the task will lead to ingenious solutions and even a survival mentality.

Some of this is already occurring. For example, experts believe that one of the major causes of the greenhouse effect is the destruction of the rain forests caused by the desperate energy and land needs of the people who have access to them. Imaginative environmentalists have suggested that the more affluent nations of the world pay their owners fees to leave them in their natural condition. This is a wonderful example of an experiment that could easily be done. The rent would provide the needed food and fuel and equipment for the country that owned the forests. It would be a steady source of income. Perhaps, too, the local people would be given employment maintaining and improving the forests. Perhaps as a one-world mentality develops, the nations, functioning together, will buy and preserve the essential lands around the world just as many of them do now within their boundaries. Once a dedication to maintain the ecosystem develops, there will be many such needs or opportunities around the globe.

VIII. CHANGING DIRECTIONS

To oversee, support, and evaluate this major new effort in the United States, the Congress should probably establish a joint committee not unlike the Joint Committee on Atomic Energy that guided the U.S. nuclear program for so many years. A comparable international oversight group should also be created. This level of attention and priority is needed to sustain this massive undertaking and to protect it from inevitable bureaucratic diseases. Because there will be no visible enemy to keep the world's attention and commitment focused on building a sustainable planet, there will be need for such groups to keep reminding us of the urgency of the task.

The new D.O.D.'s should rapidly select those areas that pose the most serious threats to survival and begin research and development projects focused on them. As basic knowledge is accumulated, and new ideas for coping with the problems emerge, it will undoubtedly be necessary to experiment with new technical solutions, just as is done in the present D.O.D. Great new industries will be needed and competition for both research and production contracts will be fierce. It will be proud work, and students will once again flock to study science and engineering. The excitement will again come from making things that improve the quality of life. As indicated earlier, the complex nature of the interaction of new technology with the already existing infra-structure of a society precludes making a detailed plan for solving these global or even many local problems, or even for organizing the new D.O.D.'s. So humanity's task for the foreseeable future will be that of steering the man-made evolutionary process with its major elements of chance. Since there can be no comprehensive blueprint for the future, humans and their governments will have to do a balancing act, continuously monitoring the state of the world, detecting problems, and then developing new technologies and new institutions to meet the changing human conditions.

Just appreciating the uncertain prospect of the interactions and being prepared to change means and even goals will permit the emerging difficulties to be recognized and addressed more quickly. Currently, neither individuals nor governments approach a new enterprise with an adequate degree of tentativeness or alertness for signs of unanticipated effects. They are conditioned to believe that deviations from plans are human mistakes rather than inherent in life. Thus rather than watch for errors, they try not to see them. Rather than look for corrective changes, they look for excuses until the errors become glaring and destructive. Although many nations, including the United States, have established agencies to detect and prevent environmental abuses, until now these groups have only been able to slow down a bit, not halt or reverse the damage being done.

A new world order is needed. International, interdisciplinary study activities should be started to raise the level of concern and to point the way. If this happens, the day will come when society will focus on the needs and possibilities and concentrate on solutions, not threats.