In November 1962 electrical engineer David L. Johnson and clinical-social psychologist Arthur L. Kobler, both at the University of Washington, Seattle, published "The Man-Computer Relationship. The potential contributions of computers crucially depend upon their use by very human human beings," Science 138 (1962) 873-79. The introductory and concluding sections of the paper are quoted below:
"Recently Norbert Wiener, 13 years after publication of his Cybernetics, took stock of the man-computer relationship [Science 131, 1355 (1960).] He concluded, with genuine concern, that computers may be getting out of hand. In emphasizing the significance of the position of the computer in our world, Wiener comments on the crucial use of computers by the military: 'it is more than likely that the machine may produce a policy which would win a nominal victory on points at the cost of every interest we have at heart, even that of national survival.'
"Computers are used by man; man must be considered a part of any system in which they are used. Increasingly in our business, scientific, and international life the results of data processing and computer application are, necessarily and properly, touching the individuals of our society significantly. Increasing application of computers is inevitable and requisite for the growth and progress of our society. The purpose of this article is to point out certain cautions which must be observed and certain paths which must be emphasized if the man-computer relationship is to develop to its full positive potential and if Wiener's prediction is to be proved false. In this article on the problem of decision making we set forth several concepts. We have chosen decision making as a suitable area of investigation because we see both man and machine, in all their behavior actions, constantly making decisions. We see the process of decision making as being always the same: within the limits of the field, possibilities exist from which choices are made. Moreover, there are many decisions of great significance being made in which machines are already playing an active part. For example, a military leader recently remarked, "At the heart of every defense system you will find a computer." In a recent speech the president of the National Machine Accountants Association stated that 80 to 90 percent of the executive decisions in U.S. industry would soon be made by machines. Such statements indicate a growing trend-a trend which need not be disadvantageous to human beings if they maintain proper perspective. In the interest of making the man-machine relationship optimally productive and satisfactory to the human being, it is necessary to examine the unique capabilities of both man and machine, giving careful attention to the resultant interaction within the
"The levels of human knowledge of the environment and the universe are increasing, and it is obviously necessary that man's ability to cope with this knowledge should increase—necessary for his usefulness and for his very survival. The processes of automation have provided a functional agent for this purpose. Successful mechanized solution of routine problems has directed attention toward the capacity of the computer to arrive at apparent or real solutions of routine-learning and special problems. Increasing use of the computer in such problems is clearly necessary if our body of knowledge and information is to serve its ultimate function. Along with such use of the computer, however, will come restrictions and cautions which have not hitherto been necessary. We find that the computer is being given responsibilities with which it is less- able- to cope than man is. It is being called on to act for man in areas where man cannot define his own ability to perform and where he feels uneasy about his own performance- where he would like a neat, well-structured solution and feels that in adopting the machine's partial solution he is closer to the "right" than he is in using his own. An aura of respectability surrounds a computer output, and this, together with the time-balance factor, makes unqualified acceptance tempting. The need for caution, then, already exists and will be much greater in the future. It has little to do with the limited ability of the computer per se, much to do with the ability of man to realistically determine when and how he must use the tremendous ability which he has developed in automation. Let us continue to work with learning machines, with definitions of meaning and 'artificial intelligence.' Let us examine these processes as 'games' with expanding values, aiming toward developing improved computer techniques as well as increasing our knowledge of human functions. Until machines can satisfy the requirements discussed, until we can more perfectly determine the functions we require of the machines, let us not call upon mechanized decision systems to act upon human systems without intervening realistic human processing. As we proceed with the inevitable development of computers and means of using them, let us be sure that careful analysis is made of all automation (either routine-direct, routine-learning, or special) that is used in systems of which man is a part) sure that man reflects upon his own reaction to, and use of mechanization. Let us be certain that, in response to Samuel Butler's question, "May not man himself become a sort of parasite upon the machines; an affectionate machine tickling aphid?' We will always be able to answer 'No.' "