security

encyclopedia

S (36)
Explore Encyclopedia

Turing Test

The Turing Test is a straightforward test of machine intelligence developed in 1950 by British multidisciplinary scientist Alan Turing.

In this test, if a human interacting with a computer is unable to discern whether they are communicating with a computer or a person, the computer has thus demonstrated human intelligence. 

Today, artificial intelligence (AI) is commonplace but AI as we know it does not meet the rather high threshold for passing a Turing Test. The computer being tested must not only be able to respond to queries, it must also be able to carry on a two-way conversation that, even if simplistic, is sensible. 

Objections to the pass/fail nature of Turing’s test abound, indeed Turing listed a dozen such objections that would cause people to cancel the test’s validity on rational or purely emotional grounds. Some other scientists, for example, have asserted that a computer passing the test reveals that it communicates like a human indistinguishably, but that it doesn’t think like a human. 

In all, there is a great deal of debate surrounding the limits or potential of AI. The Turing Test sets a benchmark for how these technologies are rigorously evaluated.

Example:

“The 2014 sci-fi thriller Ex Machina is a good example of Hollywood’s depiction of a Turing Test. In the film, the highly advanced humanoid AI ‘Ava’ passes the basic test of machine intelligence and indeed demonstrates self-awareness.” 

hypr_trial_ad
hypr_trial_ad