I am quite ready to subscribe to the already mentioned slogan that "whatever a human being can do,an appropriate machine can do, too"; but I do this only because.I regard the slogan as utterly trivial. At the moment, I am not talking about what maohines could do in principle but only about what actually existing or blueprinted machines could do, and it Is with regard to these that I utter my definite opinions. If someone wishes to write sciencefiction about information-processing centres of the (undetermined) future, let him do so and I shall discuss it with him over a glass of beer and even offer some startling suggestions of my own. If he is interested in improving the literature search process today, I would strongly advise him to forget about mechanizing abstracting or indexing. May I add that it is with a good amount of sorrow that I have come to this conclusion which is quite counter, to my temperament and my convictions (never published) of a few years ago.

Particular attention is given to processes involving pattern recognition, learning, planning ahead, and the use of analogies or?models!. Also considered is the question of designing "administrative" procedures to manage the use of these other devices. The paper begins with a discussion of what is meant by "Intelligence" and concludes with a sec-- tion concerned with some techniques through which a machine might further improve itself by adding to Its collection of problem--solving methods. I. INTELLIGENCE I feel that it would not be useful to lay down any absolute defini-- tion of "intelligence" or of "intelligent behaviour". For our goals in trying to design "thinking machines" are constantly changing in relation to our ever--increasing resources in this area. Certainly there are many kinds of performances which if exhibited by a man we would all agree, today, require or manifest intelligence. But would we agree tomorrow?

... the technique of heuristic programmingis under detailed investigation as a means to the end of applying largescaledigital computers to the solution of a difficult class of problems currentlyconsidered to be beyond their capabilities; namely those problemsthat seem to require the agent of human intelligence and ingenuity fortheir solution. It is difficult to characterize such problems further, except,perhaps, to remark rather vaguely that they generally involve complexdecision processes in a potentially infinite and uncontrollable environment.If, however, we should restrict the universe of problems to those thatamount to the discovery of a proof for a theorem in some well-definedformal system, then the distinguishing characteristics of those problems ofspecial interest to us are brought clearly into focus.Proceedings of an International Conference on Information Processing. Paris:UNESCO House, 273-282.

The paper examines the extent to which a less mechanistic approach may be possible and suggests limitations that may be imposed not only by human limitations but by difficulties of correspondence and significance between machine and manual situations. 1. INTRODUCTION LET us assume that automatic data processing (A.D.P.) can do the things that we are planning for it to do at present, such as payroll, stores accounting, and statistical analyses. There will, of course, be many problems to be solved before these tasks can be regarded as satisfactorily completed, and before we can speak with confidence out of experience. But these problems do not appear to have any fundamentally insuperable content. The difficul-- ties are manmade rather than intrinsic. They originate in part from the difficulty of adjusting the organisms of office life to new rhythms, new environments, new relationships, in part from imperfect understanding and appreciation of the power and range of new techniques, and in part from a lack of perception of the limitations and deficiencies of these systems. We may reasonably suppose that, during the course of the next five years, these difficulties will be overcome and that, throughout Government Departments and Industry, there will be a growing number of installations at work on these jobs. With this perhaps over--simplified premise, it is not too early to start thinking about a possible future form of A.D.P. in Government Departments in, say, ten or fifteen years' time.

TABLE OF CONTENTS FOR the two volumes of papers and discussions of papers from the 1959 conference held at the National Physical Laboratory, sometimes known as "The Teddington Conference". Officially these two volumes are the Proceedings of the Symposium on Mechanisation of Thought Processes. Many of the classics are downloadable separately, the remainder are available from the longer downloads of the complete volumes:Vol 1 = http://aitopics.org/sites/default/files/classic/TeddingtonConference/Mechanisation of Thought Processes Vol. 1.pdfVol. 2 = http://aitopics.org/sites/default/files/classic/TeddingtonConference/Mechanisation of Thought Processes Vol. 2.pdfTeddington Conference

Report VG-1196-G-1, Cornell Aeronautical Laboratory, Buffalo, NY

"We ask first whether we need a theory of creative thinking distinct from a theory of problem solving. Subject to minor qualifications, we conclude there is no such need -- that we call problem solving creative when the problems solved are relatively new and difficult. Next, we summarize what has been learned about problem solving by simulating certain human problem solving processes with digital computers. Finally, we indicate some of the differences in degreee that might be observed in comparing relatively creative with relative routine problem solving."RAND Corporation Paper P-1320, Santa Monica, Calif

From the The Teddington Conference– D. V. Blake and A. M. Uttley (Eds.). Proceedings of the Symposium on Mechanisation of Thought Processes, National Physical Laboratory, Teddington, Middlesex, England, London: H. M. Stationary Office.

Proceedings of the Symposium on the Mechanisation of Thought Processes.Teddington, Middlesex, England: National Physical Laboratory.

This is a case study in problem-solving, representing part of a program of research on complex information-processing systems. We have specifieda system for finding proofs of theorems in elementary symbolic logic, and by programming a computer to these specifications, have obtained empirical data on the problem-solving process in elementary logic. The program is called the Logic Theory Machine (LT); it was devised to learn how it is possible to solve difficult problems such as proving mathematical theorems, discovering scientific laws from data, playing chess, or understanding the meaning of English prose.The research reported here is aimed at understanding the complexp rocesses (heuristics) that are effective in problem-solving. Hence, we are not interested in methods that guarantee solutions, but which require vastamounts of computation. Rather, we wish to understand how a mathematician, for example, is able to prove a theorem even though he does not know when he starts how, or if, he is going to succeed.Proceedings of the Western Joint Computer Conference, 15:218-239. Reprinted in Feigenbaum and Feldman, Computers and Thought (1963).