Turing test 
Named for computing pioneer Alan Turing, this test evaluates the robustness of artificial intelligence. Back in the 1950s, Turing predicted that by the end of that century a computer with the right program could engage in a written conversation, and for five minutes pass for a real human about 70 percent of the time. The Turing test (now held annually for the Loebner Prize) involves participating in a written conversation via e-mail or online chat. If the human participant believes the computer correspondent is another person, the software is classified as possessing true artificial intelligence.