The purpose of this section is to review our own research on the use of a computer to solve diagnostic and treatment problems in medicine. A major result of this research has been the development of a computer program that is intended to serve as a consultant in a number of medical problem areas. Here the considerations that underlie the program are discussed. The basic functions of the program are outlined in a nontechnical way, and an example of the use of the program is given. Then the results of the use of the program for several different medical problems are reviewed. Finally, an attempt is made to ascertain the potential of programs such as this in the delivery of appropriate medical care. Detailed reports on various aspects of this research are available in the literature (Gorry, 1967; 1968; Gorry and Barnett, 1968a; 1968b), and so the emphasis here will be on providing a general overview of the work and results obtained to date.

At the present stage of research in artificial intelligence , machines are stil l remote from achieving a level of intelligence comparable in complexity to human thought. As computer applications become more sophisticated, however, and thus more influential in human affairs , it becomes increasingly important to understand both the capabilities and limitations of machine Intelligence and its potential impact on society. To this end, the artificial intelligence field was ex­amined in a systematic manner. The study was divided into two parts : (1) Delineation of areas of artificial intelligence, and postulatio " of hypothetical prod­ucts resulting from progress in the field , and (2) A judgmental portion, which involved appli­cations and implications of the products to society . For the latter purpose, a Delphi study was conducted among experts in the artificial intelligence field to solicit their opinion concerning prototype and com­mercial dates for the products, and the possibility and desirability of their applications and implications .In IJCAI-73: THIRD INTERNATIONAL JOINT CONFERENCE ON ARTIFICIAL INTELLIGENCE, 20-23 August 1973, Stanford University Stanford, California.

A computer program has been written that successfully discovers syntheses for complex organic chemical moleculeB. The definition of the search space and strategies for heuristic search are described in this paper. It is not growing like a tree... ...In small proportions we just beauties see; - Ben Jonson. Introduction The design of application of artificial intelligence to a scientific task such as Organic Chemical Synthesis was the topic of a Doctoral Thesis completed in the summer of 197I. Chemical synthesis in practice involves i) the choice of molecule to be synthesized; ii) the formulation and specification of a plan for synthesis (involving a valid reaction pathway leading from commercial or readily available compounds to the target compounds with consideration of feasibility regarding the purposes of synthesis); iii) the selection of specific individual steps of reaction and their temporal ordering for execution; iv) the experimental execution of the synthesis and v) the redesign of syntheses, if necessary, depending upon the experimental results. In contrast to the physical synthesis of the molecule, the activity in ii) above can be termed the'formal synthesis'. This development of the specification of syntheses involves no laboratory technique and is carried out mainly on paper and in the minds of chemists (and now within a computer's memory!). Importance and Difficulty of Chemical Synthesis The importance of chemical synthesis is undeniable and there is emphatic testimony to the high regard held by scientists for synthesis chemists.

And-or graphs and theorem-proving graphs determine the same kind of search space and differ only in the direction of search: from axioms to goals, in the case of theorem-proving graphs, and in the opposite direction, from goals to axioms, in the case of and-or graphs. Bidirectional search strategies combine both directions of search. We investigate the construction of a single general algorithm which covers unidirectional search both for and-or graphs and for theorem-proving graphs, bidirectional search for path-finding problems and search for a simplest solution as well as search for any solution. We obtain a general theory of completeness which applies to search spaces with infinite or-branching. In the case of search for any solution, we argue against the application of strategies designed for finding simplest solutions, but argue for assigning a major role in guiding the search to the use of symbol complexity (the number of symbol occurrences in a derivation).

"In this paper we describe some major new additions to the STRIPS robot problem-solving system. The first addition is a process for generalizing a plan produced by STRIPS so that problem-specific constants appearing in the plan are replaced by problem-independent parameters.The generalized plan, stored in a convenient format called a triangle table, has two important functions. The more obvious function is as a single macro action that can be used by STRIPS—either in whole or in part—during the solution of a subsequent problem. Perhaps less obviously, the generalized plan also plays a central part in the process that monitors the real-world execution of a plan, and allows the robot to react "intelligently" to unexpected consequences of actions.We conclude with a discussion of experiments with the system on several example problems."Artificial Intelligence 3:251-288

The research reported here concerns the principles used to automatically generate three-dimensional representations from line drawings of scenes. The computer programs involved look at scenes which consist of polyhedra and which may contain shadows and various kinds of coincidentally aligned scene features. Each generated description includes information about edge shape (convex, concave, occluding, shadow, etc.), about the type of illumination for each region (illuminated, projected shadow, or oriented away from the light source), and about the spacial orientation of regions. The methods used are based on the labeling schemes of Huffman and Clowes; this research provides a considerable extension to their work and also gives theoretical explanations to the heuristic scene analysis work of Guzman, Winston, and others. A condensed version appears in Patrick Winston (ed.), The Psychology of Computer Vision, pp. 19{91, New York: McGraw-Hill, 1975. Direct link to . MIT AI Lab Technical Report No AITR-271, November 1

"Problem-solving methods using some sort of heurstically guided search process have been the subject of much research in Artificial Intelligence. This paper groups these problem-solving methods under three major headings: the State-Space Approach, the Problem-Reduction Approach and the Formal-Logic Approach." New York: McGraw-Hill.

Early programs were written to demonstrate that a particular task could be accomplished and could not periorm other tasks, even if quite similar, without being extensively rewritten. Generality unnecessary for the task at hand was sacrificed to keep the programs as *Currently on leave to The University of Jerusalem **Now at Computer Science Department, Rutgers University ***Is now at NIH, Bethesda, Maryland ****With Lockheed Palo Alto Research Labs //This research was supported by the Advanced research Projects Agency of the Department of Defense under Contract No. SD-183. The views and conclusions contained in this document are those of the authors and should not be interpreted as necessarily representing the official policies, either expressed or implied, of the Advanced Research Projects Agency of the U.S. Government. Bmall as possible so they would fit the core limitations of our computer. The main result of this research was the development of programs which could find and stack cubes, either sorting them by size (1), or ordering them by voice command (2).

Some of the world's most advanced artificial intelligence (AI) systems, at least the ones the public hear about, are famous for beating human players at chess or poker. Other algorithms are known for their ability to learn how to recognize cats or their inability to recognize people with darker skin. But are current AI systems anything more than toys? Sure, their ability to play games or identify animals is impressive, but does this help toward creating useful AI systems? To answer this, we need to take a step back and question what the goals of AI are.