Recent developments in AI technology and its spawning of personalized chatbots has renewed attention in the Turing test.
The Turing test, originally called the imitation game by Alan Turing in 1949, is a test of a machine’s ability to exhibit intelligent behaviour equivalent to that of a human. In the test, a human evaluator judges a text transcript of a natural-language conversation between a human and a machine. The evaluator tries to identify the machine, and the machine passes if the evaluator cannot reliably tell them apart. The results would not depend on the machine’s ability to answer questions correctly, only on how closely its answers resembled those of a human.
It seems clear that, by and large, these AI chatbots can simulate human conversation pretty well, apart from the tendency to occasionally hallucinate, to make up stuff and say it will confidence, a trait that many humans also exhibit. If one is not aware of the details of what it says, one might easily be convinced that the made-up fact is genuine. For example, a mathematician friend of mine said that his son (also a mathematician) tested out an AI system. He multiplied two very large integers together and then fed the result into the AI system and asked whether the number was ‘prime’ i.e., not divisible by any number other than 1 and itself. The machine said ‘yes’.
But despite those well known problems, people are treating the AI bots as being at least partly sentient, as demonstrated by those who seem to have formed long-term ‘relationships ‘with them and consider them as friends or even more. So at least as far as these people are concerned, these bots seem to have passed the Turing test.
[Read more…]

