Over sixty years ago, Alan Turing proposed to test the hypothesis that a machine might think in the following way. Put either a machine or a person into a closed room, and communicate from outside this room in a manner that either the person or the machine can understand. (Let us say, by typing on a keyboard.) If one poses questions to whatever or whoever is in the room and if the responses do not allow one to distinguish between human and machine, then according to Turing, we can say that the machine (or the person) thinks.

Suppose, however, the entity in the room answers all questions with “I don't know,” or “I don't understand what you are saying.” Human being or machine? Could be a programmed computer. Could be one of our denser colleagues. Suppose the questions concern next best moves in the game of tic-tac-toe.