“I propose to consider the question, ‘Can machines think?’” With this deceptively simple declaration Alan Turing started a conversation that has resonated for more than 70 years, and has decisively shaped the development of artificial intelligence

Alan Turing was an English-born mathematician, computer scientist, philosopher and theoretical biologist. Turing’s early research provided the theoretical framework for a universal computing machine. As a cryptologist at Bletchley Park during World War II, he showed the power of computing in mechanizing expert human procedures and judgements. In the latter years of the War, he began to speculate that machines could simulate the operation of human brains. 

The “Imitation Game” is the name of a test that Turing sketched out in his landmark paper, “Computing Machinery and Intelligence” (1950). Taking a simple party game as the starting point for his analysis, Turing described a scenario involving three subjects: a person, a machine and an interrogator. The interrogator is placed in a separate room and asked to question the machine and the person in a way that will help the interrogator decide which respondent is human.

The seeming simplicity of this test was not lost on Turing, and in his paper, he also considers the validity of various possible objections to his proposition: the Theological Objection, the Mathematical Objection, the Lady Lovelace Objection, and so on. Turing concluded his paper with the observation that new developments in computing machine design—including the possibility of a “learning machine” based on the model of child learning—should be actively pursued.

Ultimately, Turing intended the “Imitation Game” as a tool to think about the possibility of a future computing machine, and the conditions that might be needed to achieve that goal.

Turing’s “Imitation Game,” or the Turing Test as it was later called, affirms the human mind as the ideal model for the formulation of a thinking machine. This isn’t a surprising conceit: humans have long imagined themselves as the ideal model for most things. But Turing chooses to illustrate his argument with a test that is also based in doubt, confusion and ambiguity. This may be a game, he seems to suggest, but what are the implications of a thinking machine that could fool us into believing it is human?

Text To Speech

Alan Turing, “Computing Machinery and Intelligence," Mind: A Quarterly Review of Psychology and Philosophy, Vol LIX, no. 236, October 1950

Alan Turing, by Elliott & Fry, bromide print, 29 March 1951, Given by the sitter's mother, Ethel Sara Turing (née Stoney), 1956, Photographs Collection, NPG x27078