Potential of Artificial Intelligence

To understand what artificial intelligence is, it is necessary to understand what it is not. Today's computers perform basically one task: processing data. They add, subtract, multiply, divide and compare numerical information. In order to do this, they must be programmed with step-by-step instructions, usually written in a computer "language."

Artificial intelligence systems can process information much more rapidly than can conventional computers. Instead of being limited to numerical information, they can understand symbolic information as well. This combination of rapid processing and the ability to understand symbols enables the artificial intelligence machines to process information simultaneously. In this way, artificial intelligence machines are able to compare facts and rules to make responses.

In 1956, the first working artificial intelligence machine was programmed to process symbols at a time when all other computers used numbers (Leepson, 1985, p. 626). The year 1956 also was the first time a wor

. . .
, p. 622). "Expert" systems got their name because the rules and facts on which they act come from recognized human experts in the field. When Mutual of New York decided to design an expert system for underwriting, for example, the system's designer spent 40 days with the company's chief underwriter. In trying to make the computer work as much like this underwriter as possible, the designer observed everything about how the job was performed. An application is 12 pages long, yet the underwriter looked at the last page first. Consequently, the artificial intelligence program looks at the last page first, as well. This same method is used in designing artificial intelligence for medical diagnosis and for manufacturing. Designers sit with human experts, either physicians or manufacturing engineers, and observe them doing their job. The experts are forced to analyze how they perform everyday tasks in such detail as they never previously considered. With diagnostic physicians, this means which questions are asked first about given symptoms. With manufacturing engineers, it means determining how a design becomes a product. In finance, expert systems can evaluate positive and negative indicators and reach a conclusion without bei
. . .

Some common words found in the essay are:
York American, FORTRAN COBOL, Mutual York, , Dartmouth College, artificial intelligence, Pacific Bank, Bank Seattle, expert system, expert systems, Peter Childers, Automation Corporation, Winter Real, intelligence systems, artificial intelligence systems, leepson 1985, gross income, credit risk, artificial intelligence focus, intelligence machines, conventional computers, developers artificial intelligence, intelligence focus, understand artificial intelligence, artificial intelligence machines,
Approximate Word count = 1659
Approximate Pages = 7 (250 words per page)

More Essays on Potential of Artificial Intelligence