This preview shows pages 1–3. Sign up to view the full content.
This preview has intentionally blurred sections. Sign up to view the full version.View Full Document
Unformatted text preview: EPISTEMOLOGICAL PROBLEMS OF ARTIFICIAL INTELLIGENCE John McCarthy Computer Science Department Stanford University Stanford, CA 94305 email@example.com http://www-formal.stanford.edu/jmc/ 1977 1 INTRODUCTION In (McCarthy and Hayes 1969), we proposed dividing the artificial intelli- gence problem into two partsan epistemological part and a heuristic part. This lecture further explains this division, explains some of the epistemolog- ical problems, and presents some new results and approaches. The epistemological part of AI studies what kinds of facts about the world are available to an observer with given opportunities to observe, how these facts can be represented in the memory of a computer, and what rules permit legitimate conclusions to be drawn from these facts. It leaves aside the heuristic problems of how to search spaces of possibilities and how to match patterns. Considering epistemological problems separately has the following advan- tages: 1. The same problems of what information is available to an observer and what conclusions can be drawn from information arise in connection with a variety of problem solving tasks. 1 2. A single solution of the epistemological problems can support a wide variety of heuristic approaches to a problem. 3. AI is a very difficult scientific problem, so there are great advantages in finding parts of the problem that can be separated out and separately attacked. 4. As the reader will see from the examples in the next section, it is quite difficult to formalize the facts of common knowledge. Existing programs that manipulate facts in some of the domains are confined to special cases and dont face the difficulties that must be overcome to achieve very intelligent behavior. We have found first order logic to provide suitable languages for express- ing facts about the world for epistemological research. Recently we have found that introducing concepts as individuals makes possible a first order logic expression of facts usually expressed in modal logic but with important advantages over modal logicand so far no disadvantages. In AI literature, the term predicate calculus is usually extended to cover the whole of first order logic. While predicate calculus includes just for- mulas built up from variables using predicate symbols, logical connectives, and quantifiers, first order logic also allows the use of function symbols to form terms and in its semantics interprets the equality symbol as stand- ing for identity. Our first order systems further use conditional expressions (nonrecursive) to form terms and -expressions with individual variables to form new function symbols. All these extensions are logically inessential, because every formula that includes them can be replaced by a formula of pure predicate calculus whose validity is equivalent to it. The extensions are heuristically nontrivial, because the equivalent predicate calculus may be much longer and is usually much more difficult to understandfor man or...
View Full Document