Turing Test

The Turing test is a proposal by Alan Turing, published in 1950 as "the imitation game", for judging whether a machine can exhibit human-like intelligence. A human judge converses in text with a hidden human and a hidden machine; if the judge cannot reliably tell which is which, the machine is said to have passed.

Why it matters

It is the best-known benchmark idea in artificial intelligence and a common reference in debates about chatbots and language models. Its limits are widely discussed: it measures the ability to imitate human conversation, not understanding, consciousness or general reasoning, and results depend heavily on the judges, the time allowed and the topics. Modern researchers rely on broader benchmarks rather than treating it as a formal standard.

Example

Judge: "What did you do last weekend?"
Entrant A: "Went hiking, my knee still hurts."
Entrant B: "I hiked near Shimla and slept badly after."
If the judge guesses wrong about half the time, the machine is
considered indistinguishable under that protocol.

Common misconceptions

• "Passing it proves a machine thinks." It only shows convincing imitation.
• "It is an official certification." No single agreed formal test exists.
• A short chat fooling someone does not equal passing a rigorous version.

Related

• Artificial intelligence
• Chatbot
• AGI
• Benchmark

This page is general information written by our editorial team and updated when documentation changes. Servers and providers differ, so test changes safely first. See our editorial policy and disclaimer.