The Friedman Inverse Test
The Turing Test is a famous method for determining if a computer exhibits intelligent behavior. The idea, basically, is to have a human judge engage in an natural language conversation via keyboard and screen with two entities: another human and a computer. If the judge can’t tell the difference between them, the computer is said to exhibit intelligence. (Try to construct an effective definition of “intelligence” from first principles to see why the Turing Test is such a clever idea.)
It is now possible to devise an updated version of the Turing Test, which I call the Friedman Inverse Test. A judge gives a computer a set of writings, each of which is either a Thomas Friedman column, or a very precise parody of a Thomas Friedman column. If the computer cannot correctly distinguish between the two, then it is intelligent.