Intelligence Testing Party Games

Party games are just for fun - nothing serious about them at all. A party game certainly wouldn't be behind one of the most important research papers of the 20th century would it?

Err, actually yes. A Victorian parlour game called the Imitation Game lies at the heart of research on Artificial Intelligence and the quest to make machines think.

Back in the 1940s and 1950s, Alan Turing, one of the most influential computer scientists ever, was interested in giving computers intelligence. Intelligence is a tricky thing though that is very hard to tie down even in humans never mind animals or computers. Turing started to wonder if ever a computer deserved to be called intelligent how could we test it? He turned to the Imitation Game for inspiration.

The Imitation Game involves two people, a man and a woman, both trying to convince the rest of the party that they are the woman. They go into a different room and are asked questions. Their answers are read out by a neutral referee. The woman must answer truthfully. The man answers in any way he believes will convince everyone else he is really the woman.

The Imitation game tests for femaleness, but a similar test could be used to test for other things too, and in particular, Turing suggested it could be used as a test for intelligence in a machine. If on questioning a hidden machine at length it can convince you that it is a person then, given you accept that a person is intelligent, the computer must be too. Turing's version of the imitation game, therefore replaced the man by a computer and the aim of both person and computer was to convince everyone that they were the human and so intelligent.

The paper helped launch the field of Artificial Intelligence (AI). It pulls together biologists, computer scientists and psychologists in a quest to understand and replicate intelligence. Turing thought machines would pass his test before the 20th century was out, but Intelligence has proved more illusive than that. AI techniques have delivered some stunning results though - computers that can beat the best human at chess, diagnose disease, and invest in stocks more successfully than humans, for example. Not bad given it all started with a Victorian parlour game.

So next time you find yourself playing a game of Musical Chairs, don't just sit on your backside. Do some serious thinking about it and you too might launch a new research area.