The major limitations in building large software have always been (a) its brittleness when confronted by problems that were not foreseen by its builders, and (by the amount of manpower required. The recent history of expert systems, for example highlights how constricting the brittleness and knowledge acquisition bottlenecks are. Moreover, standard software methodology (e.g., working from a detailed "spec") has proven of little use in AI, a field which by definition tackles ill- structured problems. But decades of work on such systems have convinced us that each of these approaches has difficulty "scaling up" for want a substantial base of real world knowledge.

A variety of proposals for preferred methodological approaches has been advanced in the recent artificial intelligence (AI) literature. The article presents a review of such perspectives discussed in the existing literature and then considers a descriptive and relatively specific typology of these differing research perspectives. It is argued that researchers should make their methodological orientations explicit when communicating research results, to increase both the quality of research reports and their comprehensibility for other participants in the field. For a reader of the AI literature, an understanding of the various methodological perspectives will be of immediate benefit, giving a framework for understanding and evaluating research reports.

Further, new application domains such as computer -aided design (CAD), computer- aided manufacturing (CAM), and image understanding based on formal logic require novel concepts in knowledge representation and inference beyond the capabilities of current production rule systems. Fundamental research in artificial intelligence is concentrated at Corporate Research and Development (CR&D), with advanced development and applications pursued in parallel efforts by operating departments. The fundamental research and advanced applications activities are strongly coupled, providing research teams with opportunities for field evaluations of new concepts and systems. This article summarizes current research projects at CR&D and gives an overview of applications within the company.

An alternative approach allows an automaton to learn to solve problems through iterative trial-and-error interaction with its environment, much as humans do. To solve a problem posed by the environment, the automaton generates a sequence or collection of responses based on its experience. The environment evaluates the effectiveness of this collection, and reports its evaluation to the automaton. The principles underlying this paradigm, known as collective learning systems theory are explained and applied to a simple game, demonstrating robust learning and dynamic adaptivity.

The fifth Distributed Artificial Intelligence Workshop was held at the Schlumberger-Doll Research Laboratory from October 14 to 17, 1984. It was attended by 20 participants from academic and industrial institutions. It included brief research reports from individual groups along with general discussion of questions of common interest. This report summarizes the general discussion and contains summaries of group presentations that have been contributed by individual speakers.

Research in AI within the Computer Science Department at the University of California, Los Angeles is loosely composed of three interacting and cooperating groups: (1) the Artificial Intelligence Laboratory, at 3677 Boelter Hall, which is concerned mainly with natural language processing and cognitive modelling, (2) the Cognitive Systems Laboratory, at 4731 Boelter Hall, which studies the nature of search, logic programming, heuristics, and formal methods, and (3) the Robotics and Vision Laboratory, at 3532 Boelter Hall, where research concentrates on robot control in manufacturing, pattern recognition, and expert systems for real-time processing.

Over the past few years, the notion of a "prototype" (e.g., TYPICAL-ELEPHANT) seems to have caught on securely in knowledge representation research. Along with a way to specify default properties for instances of a description, proto-representations allow overriding, or "canceling" of properties that don't apply in particular cases. This supposedly makes representing exceptions ( three-legged elephants and the like) easy; but, alas, it makes one crucial type of representation impossible-that of composite descriptions whose meanings are functions of the structure and interrelation of their parts. This article explores this and other ramifications of the emphasis on default properties and "typical" objects.

One reason is that the requirements-oriented methods and intuitions learned in the development of other types of software do not carry over well to the knowledge engineering task. Another reason is that methodologies for developing expert systems by extracting, representing, and manipulating an expert's knowledge have been slow in coming. At Tektronix, we have been using step-by-step approach to prototyping expert systems for over two years now. This methodology has helped us collect the knowledge necessary to implement several prototype knowledge-based systems, including a troubleshooting assistant for the Tektronix FG-502 function generator and an operator's assistant for a wave solder machine.

Artificial Intelligence Research at the University of Michigan

The Artificial Intelligence in Medicine (AIM) Workshop has become a tradition. Meeting every year for the past nine years, it has been the forum where all the issues from basic research through applications to implementations have been discussed; it has also become a community building activity, bringing together researchers, medical practitioners, and government and industry sponsors of AIM activities.