Thinking Machines: Some Reflections on the Turing Test

Abstract

In 1950, the British mathematician Alan Turing, who first developed the idea of a programmable computing device, suggested a new way of answering the question Can machines think? (1964). First he described a game, the Game, to be played by a man, a woman and an interrogator. The man and woman are hidden from the interrogator and all communication between him and his two partners is done by teletype. The object for the interrogator is to determine, after questioning each partner, which is the man and which is the woman. The man's goal is to convince the interrogator that he is the woman; the woman's task is to convince him of the truth. The man wins the game if the interrogator makes the wrong identification. Suppose, Turing said, we replace the man or woman with a computer. In this new version, the interrogator's task is to determine which is the human being and which is the computer. Of course, no computer currently available can fool an interrogator. But suppose there were one that would do well in the Imitation Game. Turing claimed that, in every theoretically interesting sense, a machine with this capacity would be a thinking system. This conception of how to decide whether a machine can think has come to be known as the Turing Test. In describing his test, Turing intended to specify a sufficient condition for thinking. There could not be, he thought, a system capable of passing his test yet incapable of genuine thought and reason. The following reflections will probe this claim. I do not have to offer anything

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 25 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.