Turing Test
FoundationsEvergreen Landmarks
The Turing test is a historical benchmark asking whether a machine can exhibit human-equivalent conversational intelligence. Proposed by Alan Turing in 1950, it holds that a machine passes if a human evaluator cannot reliably distinguish its natural language responses from those of a real human. It remains an iconic conceptual landmark in the history of artificial intelligence.
In practice
The Turing test frames discussion of machine conversation, but it is not a procurement criterion. Sounding human and being right are different properties, so judge a system on evaluation against your own data and your own tasks rather than on how natural the demonstration sounds.