The Turing test stipulates a human assessing a conversation between a machine (made to speak as if it were human) and a human being, whereby the assessor is aware that one of the parties is a machine. All participants have to be separated from one another to ensure control and the conversation would be conducted only using some electronic text system so that the device is not evaluated based on its ability to “speak.” The machine will have passed the Turing Test if the human assessing the conversation cannot give the difference between man and machine. Turing at first had made the suggestion that in 7 out of 10 cases, the machine would be able to convincingly pass the test after 5 minutes of conversing. This does not show the machine’s capacity to give correct answers when asked; it purely checks the machines capacity to answer back as if it were human.
While working at the University of Manchester, Turing first introduced this test in his research paper called, “Computing Machinery and Intelligence.” The paper opens with a bold statement where Turing poses the question, “Can machines think?” Cognitive functioning in humans varies, and the term “thinking” becomes tricky to define, so Turing chose rather to insert another closely related question in place of the original one with precise wording. His new query delving into this topic is: “Are there imaginable digital computers which would do well in the imitation game?”