workings of automation, the magic carpet of our free economy heads for distant and undreamed horizons. Just going along for the ride will be the biggest thrill on earth.

But the somber and complex difficulties produced by cybernation, which already are beginning to plague some aspects of our society and economy, are only beginning to be recognized. Thus, although this paper will describe, first, the advantages of cybernation, which make its ever-expanding application so compelling, it will, on the whole, emphasize the less obvious, sometimes acutely uncomfortable aspects of this development with which we must successfully contend if we are to enjoy the benefits of both cybernation and democracy.

THE ADVANTAGES OF CYBERNATION

In recent years deteriorating sales prospects, rising production costs, increased foreign competition, and lower profits have led business management to turn to our national talent for technological invention as the most plausible means of reducing costs and increasing productivity, whether the product is an engine block or tables of sales figures. And the Government, faced with the need to process and understand rapidly increasing masses of numerical facts about the state of the Nation and the world, is already using 524 computers and is the major customer for more of them.

What are the advantages of cybernated systems that make Government and private enterprise turn to them to solve problems?

In the first place, in a competitive society a successfully cybernated organization often has economic advantages over a competitor using people instead of machines. As U.S. News & World Report says:

In one line of business after another, the trend is the same. Companies are spending millions of dollars to mechanize their operations, boost output, and cut costs * * *. Says an official of a big electrical company: "It is no longer a question of whether or not to automate, but rather it is how far to go and how fast to proceed. If you don't, your competition will."

Not only must many organizations automate to compete, but the same principle probably holds for competing nations. We are by no means the only semicybernated society. Europe and Russia are well underway, and their machines and products compete with ours here and in the world market. The U.S.S.R. is making an all-out effort to cybernate as much of its planning-economic-industrial operation as it can.

In the second place, reducing the number of personnel in an organization reduces the magnitude of management's human relations tasks, whether these be coping with overlong coffee breaks, union negotiations, human errors, or indifference.

In the third place, cybernation permits much greater rationalization of managerial activities. The computers can produce information about what is happening now, as well as continuously updated information about what will be the probable consequences of specific decisions based on present and extrapolated circumstances. The results are available in a multitude of detailed or simplified displays in the form of words, tables of figures, patterns of light, growth and decay curves, dial readings, etc. In many situations, built-in feedback monitors the developing situation and deals with routine changes, errors, and needs with little or no intervention by human beings. This

frees management for attention to more basic duties. There is, for example

an automatic lathe ** which gages each part as it is produced and automatically resets the cutting tools to compensate for tool wear. In addition, when the cutting tools have been worn down to a certain predetermined limit, the machine automatically replaces them with sharp tools. The parts are automatically loaded onto the machine and are automatically unloaded as they are finished. These lathes can be operated for 5 to 8 hours without attention, except for an occasional check to make sure that parts are being delivered to the loading mechanism.

Another example, combining built-in feedback with a display capability, adds further illumination:

The Grayson-Robinson apparel chain, which has more than 100 stores throughout the country, receives print-punch tags daily from its stores and converts them to full-size punchcards. The complete merchandise and inventory control function is then handled on a computer. What styles are to be processed first are determined at the computed center. During any given week about 60 percent of the sales data are received and summarized. On the following Monday morning the remaining 40 percent of the sales data are received. The computer can then begin running style reports immediately after the tickets have been converted to cards. By this time the company can run up style reports by departments and price lines in order to obtain the necessary merchandise information. The entire reporting job is completed by Wednesday afternoon of each week, including reports on all inactive stockpiles.

Freeing management from petty distractions in these ways permits more precise and better substantiated decisions, whether they have to do with business strategy, government economic policy, equipment system planning, or military strategy and tactics. Thus, management in business or government can have much better control both over the system as it operates and over the introduction of changes into future operations. Indeed, the changes themselves may be planned in conformity with, and guided by, a strategy that is derived from a computer analysis of the future environment.

In the fourth place, cybernation allows government and industry much greater freedom in locating their facilities efficiently in relation to the accessibility of raw products, markets, transportation, and needed (or cheaper) human and material resources. Distance is no longer a barrier to control and coordination. The computers that control automated processes need not be near the factories nor the data-processing computers near their sources of information or users if other considerations are more pressing. Widely dispersed installations can be coordinated and controlled from still another place, and the dispersed units can interact with each other and affect one another's performance as easily, in many cases, as if they were all in the same place.

In the fifth place, some degree of cybernation is necessary to meet the needs of our larger population and to maintain or increase the rate of growth of the gross national product. An estimated 80 million persons will be added to our population in the next 20 years. Beyond increases in productivity per man-hour to be expected from the projected 20-percent growth in the labor force during this same period, productive growth will have to be provided by machines.

If the criteria are control, understanding, and profits, there are strong reasons why government and business should want to, and indeed would have to, expand cybernation as rapidly as they can.

The

versatility of computers and automation is becoming better understood all the time by those who use them, even though, as with the human brain, most present users are far from applying their full potential. Cheap and general purpose computers or modular components applicable to many types of automatic production and decisionmaking are now being manufactured. In good part, they are cheap because they themselves are produced by automated methods. Techniques for gathering the field data that serve as the "inputs" to the machines are being refined and themselves automated or semiautomated. For example, a large shoe distributor is planning to attach a prepunched IBM cards to each shoebox. When a sale is made, the card is returned to a central facility to guide inventory adjustment, reordering, and sales recording and analysis. Techniques for quickly implementing the "outputs" from the machines are also being invented. Methods are being developed for systematically establishing the precise kind and degree of cybernation required in specific situations as well as the changes needed in the rest of the institution or organization using cybernation.

These are the advantages for management, for government, and for those parts of the work force whose status has been enhanced because of cybernation. But as cybernation advances, new and profound problems will arise for our society and its values. Cybernation presages changes in the social system so vast and so different from those with which we have traditionally wrestled that it will challenge to their roots our current perceptions about the viability of our way of life. If our democratic system has a chance to survive at all, we shall need far more understanding of the consequences of cybernation. Even the job of simply preserving a going society will take a level of planning far exceeding any of our previous experiences with centralized control.

The balance of this paper will point out some of the implications of cybernation that we must recognize in our task of developing a society and institutions in which man may be allowed to reach his full capacities.

[From "Automation, the Impact of Technological Change" (American Enterprise Institute for Public Policy Research)]

AUTOMATION, THE IMPACT ON TECHNOLOGICAL

CHANGE

(By Yale Brozen)

Worries about automation have made headlines, occupied congressional committees, been the subject of many conferences, and produced a deluge of reports. The name is new, but the condition and the fear are old. In the 1920's, mechanization and electrification were felt to threaten us. In the 19th century, the Luddites rioted and smashed equipment because of their fear of the steam engine and power machinery. In the 15th century, hostelers and innkeepers opposed the growing use of the stage coach. In the 12th century, hand fullers dreaded the spread of mechanical fulling.

The worries about automation are mixed. Some fear that it is not coming rapidly enough to solve our balance-of-payments and economic growth problems. Others fear it is spreading with such a rush that workers are unable to learn new skills rapidly enough to avoid becoming as obsolete as horses.

HISTORICAL PERSPECTIVE

Before looking at the specifics of this decade's problem, we should place automation in the perspective of the continuing stream of history. Although the techniques now being applied are the result of an engineering revolution, they are, in an economic and cultural sense, nothing more than a continuation of an evolution that has been going on for centuries. Beginning with a world in which men could barely gather and capture enough to keep themselves alive, we went through a first technological revolution, ending about 3000 B.C. It raised productivity sufficiently that societies could afford priests, an aristocracy, armies, and a bureaucracy. We even reached the stage of the Greek democracies. There were democracies of the few, however, resting on a large class of slaves.

The second technological revolution did away with slavery, since it became cheaper to use mechanical power than human muscle, and created the base on which mass democracy now rests. The economic gains of the first revolution went to the few; those of the second have gone to the many; perhaps the automation revolution will carry us on from a mass democracy to a mass aristocracy.

In the past quarter century, we have doubled the productivity of the average man. If we can do the same in the next quarter century, average family income will go to $14,000 a year instead of the present $7,000. Median educational attainment, which rose from the grade school graduate to the high school graduate level, will move to that of the junior college graduate. Similar movements in the avaliability of leisure, the consumption of world travel, music, fine art, and medical services, will occur if productivity and income trends continue. Essentially, this is what automation promises.

THE PACE OF AUTOMATION

Technology is felt to have moved at a sedate pace in the past compared with the recent rate of advance. In the 1930's, it was fashionable to argue that the country was stagnant because, among other factors, technology was not producing the great new techniques it formerly did. For this reason, it was said, there was little use for additional capital. Unemployment then was attributed to the lack of technological advance and the resultant lack of outlets for the application of new capital.

Too much technological change, however, may be as bad as too little. In the late 1940's, we were told that automation was coming with such speed that factories would be fully automated in 10 to 20 years and that we were faced with "a decade or more of ruin and despair." It is more than 12 years since that forecast was made by Norbert Wiener, the popularizer of the term "Cybernetics" and one of the developers of the mathematics of communication which underlies much of the new control technology. Yet, we still employ over 12 million produc

tion workers in factories, about the same number as at the time of Wiener's prediction. Total employment in manufacturing industries has grown by more than a million men since 1950.

How rapidly we automate depends upon the availability of capital and the rapidity of the rise in real wage rates. Present rates of capital formation are not much higher than the amount required to equip additions to the work force with the same quantity of capital as that used by the presently employed work force. If we add 1,300,000 workers a year to our labor force, the rate forecast for the next decade, nearly $20 billion a year will be needed to equip them sufficiently to be productive enough to be employable at current wage rates. If wage rates increase, a larger amount of capital will be required. If a larger amount does not become available in this situation, unemployment will grow.

Total spending on business plant and equipment is less than $10 billion a year. Almost half of this amount is required to replace plant and equipment worn out by use, rendered economically obsolete, and destroyed by fire or other catastrophe. There is very little left to increase the quantity of capital per man employed after providing for additions to the work force. In these circumstances, the spread of automation will be slow.

Automation is spreading somewhat more rapidly than is indicated by these figures because of the overpricing of labor in some industries. Increases in wage rates make it economic to automate when it otherwise would not be so. The cost of the new equipment required to replace a man is about $35,000. If the annual wage of a man is less than $7,000, it does not pay to replace him at this cost since property and corporate earnings taxes, insurance, depreciation, and interest costs amount to about $7,000 on $35,000 worth of equipment. If annual employment costs rise above $7,000, it then pays to replace a few men. The rate at which men will be released by an industry to other occupations or to the manning of additional capacity depends upon how rapidly wage rates rise. (If they rise too rapidly, men will not be completely absorbed by expanding occupations and unemployment will rise.)

As men are replaced, it becomes increasingly expensive to replace additional men. In the late 1940's, when coal mining employment exceeded 400,000, the cost of replacing a man was about $20,000. Now that coal mining employment has dropped below 170,000, the cost of replacing a man is close to $50,000.

To automate as completely as possible with present technology, only one major segment of the American economy, manufacturing, would require an expenditure of well over $2.5 trillion, assuming output is not increased. Even to modernize manufacturing to the levels of the new plants built in the 1950's would require over $500 billion. Since total spending on new plant and equipment in manufacturing amounts to about $15 billion per year, American manufacturing could not be modernized even to the level of the technology of the 1950's for over 30 years. And this is under the extreme assumption that all the expenditure is used for modernization and none for expansion. With the current division of capital outlays between modernization and expansion, modernization to the level of new plants of the variety built in the 1950's would require about 50 years.