The Illusion of Intelligence

The Turing test is a test proposed as a way of testing for whether machines are really intelligent. If a machine can pass itself off as a human then we should credit it with human-level intelligence. Not everyone agrees that the Turing test is really about intelligence. Many researchers believe it is just about the illusion of intelligence. They claim it will be possible for machines to pass the Turing test without being able to do anything remotely like thinking at all - just like chess computers can beat humans without thinking like a grandmaster. After all just because a man passed himself off as a woman in the original Victorian parlour game on which it was based that wouldn't mean he was a woman.

There is in fact a competition run every year based on the Turing Test. The Loebner Prize for artificial intelligence will award $100,000 and a Gold Medal for the first computer "whose responses are indistinguishable from a human's". Many of the entrants each year are created by people with no intention of building intelligent machines: just ones able to pass the test. Just like students sometimes "learn to a test" which they then pass without gaining any real understanding of the subject, so it is argued chatbots will eventually win the Loebner prize without any understanding of what they are talking about. So far no chatbot has come close to winning the prize though - so maybe it's a difficult illusion to pull off after all.