The man attempts to trick the interrogator into believing they are female, while the interrogators must decide which of the two is truly the woman. Turing then raises the proposition of a machine taking the man’s place, questioning “will the interrogator decide wrongly as often when the game is played [with a machine] as he does when the game is played between a man and a woman?” This introduced a new way of measuring a machine’s ability through exhibition of intelligent human behaviour. Half a century later, R.M. French explored how attitudes towards solving the Turing test shifted over the years in relation to the development of Artificial Intelligence in his paper “The Turing test: the first 50 years” (2000). He notes that cognitive scientist Marvin Minsky initially approached the problem with optimism, writing in 1967 that “Within a generation the problem of creating ‘artificial intelligence’ will be substantially solved”; however, by 1982 his viewpoint took a major paradigm shift, quoting “The AI problem is one of the hardest ever undertaken by …show more content…
von Ahn et al. in the paper “CAPTCHA: Using Hard AI Problems For Security” as a challenge-response authentication, which requires the user to complete a task that can only be solved with human intelligence; the purpose being to prevent automated software from accessing or submitting data. CAPTCHA can be described as a “reverse” Turing test, in that it is being administered by a computer to determine if the subject is human, and therefore the Turing test can be used as a suitable means of measuring the effectiveness of a CAPTCHA system. Interestingly, CAPTCHA has seen recent usage in aiding machine learning with the now Google-owned reCAPTCHA. reCAPTCHA tests poses the user in the identification of images from a given series that match a provided description, some of which are known to the system, and others unknown. With Google’s expansive dataset, such as StreetView images, the reCAPTCHA system can use human effort to analyse a vast number of images for visual components it may contain, such as street signs or house numbers, and employ pattern recognition on images that have been confirmed to contain a certain object so that it may more accurately predict the presence of these objects in other images. On the other side of the spectrum, some computer scientists have been inspired to develop their own machine learning