COMMENT ⊗   VALID 00002 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002	title: Some Expert Systems Require Common Sense
C00006 ENDMK
C⊗;
title: Some Expert Systems Require Common Sense

abstract: Human common sense involves certain knowledge and ability
that is only beginning to be identified by artificial intelligence
research.  Only part of it has been formalized enough to be usable
by computer programs, and even less is involved in current expert
systems.  Many current expert systems are quite impressive and even
useful without it.  However, many important tasks require common sense
knowledge and reasoning ability. This lecture describes some of this
knowledge and ability and discusses what kinds of tasks require it.

	Common sense knowledge includes the following:

	1. General facts about how events occur in time and affect situations.

	2. Facts about the ability of actions to achieve goals.

	3. Concurrent events.

	4. Knowledge and belief and ways of obtaining knowledge.

	5. Physical objects, their shapes, their construction and
destruction, and their motion.

	6. Naive or common sense physics.

	Little of this knowledge is included any present expert systems,
and that is usually represented in such a way that it can only be
used to achieve specific goals.  Much human common sense knowledge is
available in purpose-independent forms, and this will be required in
many future AI systems as well.

	Common sense reasoning includes deductive logical reasoning,
inductive reasoning, non-monotonic
reasoning (a new area of research), and obtaining facts by observation.

Biographical sketch:

	John McCarthy is Professor of Computer Science at Stanford
University.  He has been interested in artificial intelligence since
1949 and coined the term in 1955.  His main artificial intelligence
research area has been the formalization of common sense knowledge.
He invented the LISP programming language in 1958, developed the concept of
time-sharing in the late fifties and early sixties, and has worked on
proving that computer programs meet their specifications since the early sixties.
His most recent theoretical development is the circumscription method
of non-monotonic reasoning since 1978.  McCarthy received the A. M. Turing
award of the Association for Computing Machinery in 1971 and was elected
President of the American Association for Artificial Intelligence for
1983-84.