numbers of variables so organized that the mathematics of probability and statistics do not altogether apply. We need new logic for these types of problems. Most of the problems dealing with living prototypes fall in this latter category. Because of these difficulties on the living side and the advanced capabilities on the physical side, man has pushed his physical understanding far out into space. Yet, within his own environment and right in his own backyard are many problems that still remain unsolved. Increased Industrial Research on Living Prototypes In the industrial research program, consider the extent of industrial research on physical phenomena as compared to corresponding research on the living world, and you might say to yourself: "This is natural because industry is interested in material things. Its products are made of material substance so why should industry do research on living prototypes?" As a challenge to industry, I firmly believe that a significant advance in material capabilities, in material analogs, can be made by research on living prototypes and exploiting the resulting know-how to achieve new orders of material and device advances. Industry can therefore make a large contribution by emphasizing investigation of the living world as well as that of the physical world in its research program. COMPLEXION OF TODAY'S TECHNOLOGY Having seen the complexion of bionics and where we stand today, why do we think that there are new possibilities through renewed interest in this area? These stem from, first, considerations as to our capabilities of modern machine analysis and simulation. We can solve problems of a complexity that we never dealt with before. Second, there are rapid advances being made in new logic, the development of logic of self-organizing systems, selfadaptive systems, of systems that come closer than ever before to behaving in the manner of living prototypes. Third, we are in an era of exploitation of molecular know-how. We have accomplished a great deal of research on molecular phenomena at the atomic and nuclei level, and today there is a significant payoff being achieved in terms of using this know-how toward applicational interests. Finally, we are placing key emphasis on new and more fundamental approaches to solutions of engineering problems. Today, instead of using heat to generate steam in a boiler, the steam to drive a turbine which in turn drives an alternator to generate electricity, we are applying the heat to a plasma, and through the resulting motion of charged particles, producing electrical power directly. Through molecular electronics, we are achieving fundamental building block techniques that provide major circuit functions by synthesized molecular and charged particle relations. Bulk and repetitive structures of large numbers of very small elemental components to produce such functions are evolving. They are becoming more similar in character to their biological counterparts. Whether they are exact analogs does not matter as long as they provide properties of equivalent logic. There is a significant trend today toward cross-fertilization of technical and scientific disciplines. In the aeronautics and astronautics area, we have seen the rapid development of magnetohydrodynamics--the marriage of aerodynamic and fluid dynamic know-how with fundamental electricity and magnetism. Hypersonic flow is heavily involved with charged particle motions as affected by the presence of electrical fields, either naturally present or imposed for purposes of obtaining a problem solution or performance advantage. A new breed of aeronautical engineer is required to deal with this new flavor of technolɔgy. Electrical propulsion has brought the fundamentals of nuclear physics and electrical engineering into the propulsion field. In molecular electronics there is a union of the physical chemist, the solid state physicist, the materials engineer and the electronics expert; it takes all of these people working together to achieve the advances being made. Similar crossfertilization of technical disciplines of an even greater degree is of extreme importance to bionics. You cannot solve the problems of the living world and exploit such know-how without cross-fertilization of our medical and biological disciplines with those of the engineering and physical sciences. With deliberate cross-fertilization of the required disciplines and interests, using advance machines and logic, I am sure that accelerated progress will result. SUMMARY I am convinced that Bionics has a tremendous promise of highly significant pay-offs and that we are on a threshold of revolutionary advances in scientific understanding and technological achievements concerning problems of organizational complexity. Further, we can achieve these goals through major expansion of research on the living world, deliberate cross-fertilization and cross-breeding of our technical and scientific disciplines, and particular emphasis on new logic. 1. 2. REFERENCES Wiener, Norbert, Cybernetics, New York, N.Y., John Wiley and Sons, Inc., 1948. Weaver, Warren, "Science and Complexity," The Scientist Speaks, New York, N.Y., Boni and Gaer, 1947. ATTITUDES TOWARD INTELLIGENT MACHINES Paul Armer "A bird is an instrument working according In this paper, I will attempt to analyze some of the attitudes and arguments which have been expressed in dealing with questions like "Can machines think?" or "Can machines exhibit intelligence?" I do so with a single purpose a hope to improve the climate which surrounds research in the field of machine or artificial intelligence. For those of you who would answer such questions negatively, my goal is not to convince you that you are wrong and the positivists are right but merely to attempt to show that most of the disagreement is a matter of semantics. (I will attempt to refute some of the negative arguments.) I do hope to convince those negativists, who argue that research on artificial intelligence is wrong, that they should be tolerant of such research. Those of you who would answer such questions affirmatively need no convincing, but if you share my views concerning the importance of research in this field, then I believe we can profitably spend some time discussing these attitudes, for the negativistic attitudes existent today are inhibiting such research HISTORY [1]. Before examining the substance of these arguments and attitudes, a look at some of the history of this discussion is in order, for the question of machines exhibiting intelligence is one which has been around for a long time. Samuel Butler (1835-1902) dealt with the question in Erewhon and Erewhon Revisited [2], wherein a civil war takes place between the "machinists" and the "antimachinists. (Incidentally, victory went to the "antimachinists.") Butler stated "There is no security against the ultimate development of mechanical consciousness in the fact of machines' possessing little consciousness now" and speculated that the time might come when "man shall become to the machines what the horse and dog are to us." Discussion of the question apparently took place in Babbage's time (1792-1871), for the Countess of Lovelace commented on it, negatively, in her writings of Babbage's efforts [3]. The topic came into prominence in the late 1940's when Babbage's dreams became a reality with the completion of the first large digital computers. When the popular press applied the term "giant brains" to these machines, computer builders and users, myself among them, immediately arose, almost to the man, to the defense of the human intellect. We hastened to proclaim that computers did not "think"; they only did arithmetic quite rapidly. Discussion of this question died down (but not out) in the early and mid 1950's but has come back in the last several years stronger than ever before. fact, it has recently invaded the pages of Science [1,4,5]. In THE NEGATIVE ARGUMENTS An examination of the arguments advanced by the negativists reveals that many of them are not arguments at all but only statements. Many just dismiss the notion out of hand, saying things like, "Let's settle this once and for all, machines cannot think!" or "A computer is not a giant brain, in spite of what some of the Sunday supplements and science fiction writers would have you believe. It is a remarkably fast and phenomenally accurate moron." [6] Others have advanced arguments which turn out to be fallacious. Many are of the type "machines will never be able to do this, because they have (or lack) such and such a property. Falsity creeps in from a variety of sources. The attribution of specific properties (or lack thereof) may be in error or it may not logically follow that the presence (or absence) of the specific properties implies that the machine will not be able to do what the arguer states it will never be able to do. Or it may be that the negativist has erroneously assumed that such properties are invariant. Let me give you some examples along these lines. "The Manchester machine which was set to solve The The statements about machine behavior in the above are just not true. While it is true that some of the early approaches to chess-playing machines were of the nature of attempting to review all possibilities in limited depth [8], this is not the only way in which the problem can be approached. The chess-playing routine of Newell, Shaw, and Simon [9] does not examine all possibilities. And those which it does consider it examines in varying detail. The routine rejects moves which appear to be worthless; it selects moves which appear to be good ones and examines them in depth to ascertain that they are indeed good. The earlier routine developed by this same team to prove theorems in logic [10] did not examine all possible proofs to do so with today's computers would literally take eons of time. Rather, it searched through the maze of possible proofs for ones which looked promising and investigated them. It relied on knowing what approaches had worked before. An example of an erroneous assumption that machine properties are invariant occurs in an article by John H. Troll: "The human memory is a filing system that has a |