Alan Turing, 20th century mathematician and Bletchley Park codebreaker, devised an imitation game commonly known as the Turing test.
How a Turing test is set-up depends on the question you want to answer, the basic one being: ‘Can a human interrogator, using text-based questions only, determine if they are speaking to a human or a machine when they can’t see or hear their chat partner?’ – Think of the scenario something like the chat interface here on ‘I’m a Scientist’.
At the moment, as in 2016, typical results are:
1. Most humans can tell the difference between human and machine from the answers.
2. Sometimes some human interrogators get it wrong and:
– a. think a machine is human
– b. think a hidden human (acting as a foil for the machine), is not human
-c. some are uncertain and can’t say from the answers whether they were put by a human or a machine.
Comments