(artificial intelligence) A test of a computer's ability to demonstrate intelligence, as measured by whether a human judge can reliably distinguish the computer from another human being by conversing with both entities in written natural language over a terminal.