AI - Background Info | AI Characteristics | Testing and Evaluating AI | Implications of AI

Testing and Evaluating Artificial Intelligence

There are several different ways that scientists have developed to help "test" the effectiveness or how realistic Artificial Intelligence really is. One of the more well-known tests, developed by Alan Turing, called the "Turing Test," states that "...if a machine acts as intelligently as a human being, then it is as intelligent as a human being." A test subject communicates with something, and if the individual is unable to correctly determine whether or not he or she is communicating with a person or a machine, then the machine effectively passes the Turing Test.

There are several other evaluative measures used in reference to Artificial Intelligence. A conventional system for describing the apparent effectiveness of Artificial Intelligence have been developed. The following terms describe the varying levels of how artificially intelligent machines compare to humans: