The Turing test evaluates whether a machine can convincingly imitate human conversation and thought. A human evaluator communicates with a hidden human and a hidden AI through printouts; the machine passes when the evaluator cannot reliably distinguish them. section 9.11
The test measures human resemblance rather than intelligence in every possible form. An AI might think, remember, or communicate differently from humans and still possess advanced or comparable abilities. section 9.11
Hidden-conversation setup
Place a human and an AI behind a curtain, have an evaluator exchange written messages with both, and ask whether the responses can be distinguished reliably. section 9.11
