AI Glossary

Turing Test

A test of machine intelligence proposed by Alan Turing in 1950, where a human evaluator tries to distinguish between a machine and a human based solely on conversation.

The Original Proposal

In 'Computing Machinery and Intelligence,' Turing proposed the 'imitation game': if a machine can fool a human evaluator into thinking it's human through text-based conversation, it demonstrates intelligence.

Modern Relevance

Modern LLMs can pass simple versions of the Turing Test. But the test is now considered an insufficient measure of intelligence -- a machine can be convincingly human-like while lacking genuine understanding, reasoning, or consciousness.

← Back to AI Glossary

Last updated: March 5, 2026