Imitation Game

This paper discusses the “Imitation Game” proposed by Alan Turing in his paper titled “Computing machinery and intelligence”. As quoted by Turing “Can machines think? This should begin with definitions of the meaning of the terms machine and think”. He believed this question should be addressed in a statistical manner. Rather he proposes a test known as “Imitation Game”, which involved three participants, a man A, a woman B and an interrogator C. The interrogator asks certain number to question to determine whether A or B is real woman. And A and B both try to convince C that they are real woman. This was the initial version of the game; Alan replaces A with a machine and now C has to determine which one among A and B is real human. In order to convince C, the machine has to think like a human, which led to the creation of his research paper. Can a machine think like a human and can it be intelligent enough to fool the interrogator?

Strong Arguments

The best thing about his paper is that he presents his argument by providing several counterexamples and refutes the expected objections with convincing reasoning. Like to the argument of Lady Lovelace that machines can only operate on the instructions given to them and lack originality. He agrees that at that time it might have seemed impossible for a machine to think on their own, but it does not imply that it will not be possible in the future. And also claims that machines have the ability to surprise their creator, as there are still some areas of its working methodology we fail to understand completely. The other arguments opposed by Turing were – various disability argument, mathematical objection, extrasensory perception etc., to which he responds that once the capacity, processing power and storage is available, building a machine with such diverse behavior and human like mind is not impossible.

I also liked his notion of designing an intelligent machine with a human brain like capabilities. His approach of training a child-program with education process and training it with experiments similar to an evolutionary process. Overall, he believes that with powerful enough computers and strong algorithms, all the limitations of a machine can be overcome in future.

Less Likely to be True

I found few of his supporting claims less substantial like to the argument made by Professor Jefferson, that machine is unemotional and unconscious and cannot feel the pleasure at its success or grief at its loss. Turing’s views were, the only way to know that a man thinks, is to be that particular man. So, we will never know, if the machine is actually thinking or not, until we are a machine. Now, I agree that idea of his test was just to fool the interrogator but actually not to test whether a computer can think or feel. But still, I believe it is a long way, before we will witness a machine with consciousness and emotions.

Parting thoughts

Turing’s work has certainly played a major role in the creation of artificial intelligence and motivated many of the researchers to build human-like machines. But even though lot of ground has been covered in the world of artificial intelligence, the Turing test is still very difficult to pass. It covers a wide range of human behavior, a machine should have, in order to pass the test. I believe that the test cannot be considered an ideal measure of intelligence in machines.