What is the Turing Test? De­f­i­n­i­tion and function explained

The Turing Test was developed by the math­e­mati­cian, Alan Turing, in 1950 and is supposed to be able to prove the in­tel­li­gence of machines through an ex­per­i­men­tal procedure. The supposed proof is provided by a question-and-answer game that is meant to prove human and ar­ti­fi­cial in­tel­li­gence are in­dis­tin­guish­able because human in­ter­roga­tors are unable to dis­tin­guish between a human and ar­ti­fi­cial in­tel­li­gence. Whether this is actually objective proof of machines with human-like in­tel­li­gence remains disputed.

Is that a human or a bot? Anyone who spends a lot of time on social media or browsing the comment sections of online articles fre­quent­ly asks them­selves this very question. Social bots imitate human users as opinion bots, steer dis­cus­sions and make automated comments. Often un­in­tel­li­gi­ble from humans, they are based on al­go­rithms that use ar­ti­fi­cial in­tel­li­gence and machine learning to imitate human-like com­mu­ni­ca­tion. This is precisely where the Turing Test, which is designed to determine whether we are dealing with humans or machines, comes into play.

The Turing Test was invented and developed by the eponymous British math­e­mati­cian, computer scientist and logician Alan Turing in 1950. He in­ten­sive­ly studied the problem of ar­ti­fi­cial in­tel­li­gence and its criteria during his work on one of the first legendary tube computers called Man­ches­ter Mark I at the Uni­ver­si­ty of Man­ches­ter.

In his article “Computing machinery and in­tel­li­gence”, published in the journal “Mind”, Turing outlined the basic features of an ex­per­i­men­tal set-up now famous as the Turing Test, but known at the time as the “Imitation Game”. Since ar­ti­fi­cial neural networks according to the principle neural network did not yet play a major role in the debate about ar­ti­fi­cial in­tel­li­gence and objective sci­en­tif­ic proof of thought processes was a long way off, ob­serv­able analyses of com­mu­ni­ca­tion with machines were used for this purpose. The goal was and is to be able to speak of ar­ti­fi­cial in­tel­li­gence or machine in­tel­li­gence in the case of machine com­mu­ni­ca­tion behavior that is in­dis­tin­guish­able from humans.

The structure and procedure of the Turing test couldn’t be any simpler. The test uses a simple question-answer procedure between a human ques­tion­er and two anonymous answerers who are not visible to the ques­tion­er. The free, un­spec­i­fied questions are asked by the human without any visual or auditory contact with the in­ter­locu­tors via an input tool such as a keyboard or a screen. At the end of the test, if the human ques­tion­er cannot determine from the answers which of the two answerers is the machine, the in­tel­li­gence of the machine can be defined as human-like.

To date (March 2022), no official examples can be cited of machines passing Turing tests. Nev­er­the­less, the ex­per­i­men­tal setup is still relevant for the de­vel­op­ment of ar­ti­fi­cial in­tel­li­gences today, e.g. in the context of deep learning, re­in­force­ment learning and su­per­vised learning, re­spec­tive­ly. In the future, human-level machine com­mu­ni­ca­tion based on neural networks will not only play a role on social media and in customer service. Fields such as medicine, di­ag­nos­tics, agribusi­ness, security, sur­veil­lance, marketing, trans­porta­tion and pro­duc­tion will also be in­creas­ing­ly char­ac­ter­ized by ar­ti­fi­cial in­tel­li­gent com­mu­ni­ca­tion.

It is still ques­tion­able whether the Turing Test can be used to provide credible or objective proof of ar­ti­fi­cial in­tel­li­gence at all. Much of the criticism voiced about the test questions above all was whether the de­cep­tive­ly genuine imitation of human com­mu­ni­ca­tion actually suggests an in­de­pen­dent ar­ti­fi­cial in­tel­li­gence or is rather nothing more than a de­cep­tive­ly genuine imitation. The ob­ser­va­tion of machine behavior, which suggests or ap­par­ent­ly pre­sup­pos­es ar­ti­fi­cial in­tel­li­gence, is not to be equated with ob­jec­tive­ly existing ar­ti­fi­cial in­tel­li­gence. Intention and thinking ability could thus neither be depicted nor proven by the question-answer game of the Turing Test.

The machine learning test called Winograd Schema Challenge (WSC) is often mentioned as an optimized counter design. This uses a pre­de­fined question scheme that requires active knowledge ap­pli­ca­tion, general knowledge and rational thinking for correct answers. Based on Terry Winograd’s Winograd Scheme, answering the questions requires an un­der­stand­ing of context, human behavior, cultural back­ground, and reasoning. Other al­ter­na­tives include the Marcus test, which asks ar­ti­fi­cial in­tel­li­gences about their un­der­stand­ing of a tele­vi­sion show they “watched”, and the Lovelace Test 2.0, which examines AI’s potential creative abilities.

Despite all the points of criticism mentioned, the central idea of the Turing Test, the de­cep­tive­ly authentic imitation of human com­mu­ni­ca­tion, still plays a major role in dig­i­ti­za­tion today.

Three usage examples il­lus­trate the unchanged con­tem­po­rary sig­nif­i­cance of the Turing Test:

• Human In­ter­ac­tion Proof (HIP): The CAPTCHA query can be described as a negative Turing Test. As a human in­ter­ac­tion proof, it is used to dis­tin­guish machines from humans as quickly as possible and to ef­fi­cient­ly filter bots through automated text and image queries before they visit a website. CAPTCHA has the Turing Test in its name: Com­plete­ly Automated Public Turing Test to tell Computers and Humans Apart.
• Bots: Bots are digital tools that offer positive or negative functions depending on how they are used. They are used, for example, as chatbots to ef­fi­cient­ly automate customer service, but are also used as social bots or spam bots to spread fake news or malware. In both cases, forms of Turing testing are used to promote the de­vel­op­ment of the bots and make them as in­dis­tin­guish­able as possible from humans.
• Voice as­sis­tants: Voice as­sis­tants are probably one of the de­vel­op­ments that come closest to Alan Turing’s basic idea. Voice-con­trolled, human-like as­sis­tants such as Alexa or Siri are based on the question-answer principle and are intended to automate everyday functions and user needs. Although none of the ap­pli­ca­tions currently come close to passing the Turing test, the in­tel­li­gent voice functions are con­stant­ly being optimized through machine learning and analyses of user behavior, making them more human-like.