Turing Test Is a Strong, Valid Test

Turing Test is a strong, valid test for human-like intelligence in machines, and has long been viewed as the "benchmark" (Berrar et al., 2013, p. 241) for determining this sort of intelligence. There are a number of different aspects of this assessment measure that makes it so. Its history alludes to its validity, as it was developed near the midway point of the 20th century by Allen Turing, who helped to provide some of the seminal work in (at the time) the emerging field of Artificial Intelligence.

Therefore, the fact that it was designed by one of the pioneers in this field helps to provide a degree of credibility about this particular assessment instrument, which created by someone who definitely knew how to distinguish artificial intelligence from the human-defined variety. Moreover, another reason why this test is a valid means of determining if machines have intelligence similar to humans is because of its requirements. Specifically, machines are expected to elicit responses to a number of inquiries in a manner that is decidedly human.

In fact, the test is actually structured to emphasize this human quality that machines possess to pass it. A human is the one who actually decides whether or not the computer can pass the test. To do so, the aforementioned person asks a series of questions to both another human being and to a computer in what amounts to a sort of "game" (Warwick et al., 2013, p. 163). This sort of querying is usually based around a particular subject and lasts for a specified duration.

The computer can pass the examination and be certified of possessing human-like intelligence if the human issuing the questions is unable to determine which response was elicited by the other human, and which was given by the computer, with a degree of regularity. In this way, what is actually being tested is not necessarily knowledge of a particular subject matter. Rather, the Turing test assesses a human's ability to mimic human behavior in the sort of responses that are given to questions.

It is important to note that answering a question correctly is not the key determinant of this sort of assessment. There are certain questions or subjects that some people do not have knowledge of in general. Still, how those questions are answered will reveal if the answerers have human-like intelligence or not. It is important to realize that computers cannot simply answer every question and be deemed that they have human-like intelligence.

There are certain factors such as the length of time it takes to answer a question and how it is phrased that is relevant for determining if a computer can pass such an assessment. Another reason that the Turing assessment -- whhich was initially unveiled in Turing's paper "Computing Machinery & Intelligence" (Vardi, 2014, p. 5) is particularly efficacious for determining if computer truly have human-like intelligence is due to the nature of the questions that one can choose to ask it.

It is not sufficient to ask a computer a question in which there are limited responses -- such as yes or no. Rather, it is important to ask machines open ended questions to see how they are able to answer them. These latter question types are included in the Turing test and help to reveal a lot about a computer's ability to duplicate human intelligence. When asked certain questions people might equivocate,.