One of the hottest topics that modern science has been focusing on for a long time is the field of artificial intelligence, the study of intelligence in machines or, according to Minsky, "the science of making machines do things that would require intelligence if done by men." (qt d in Copeland 1). Artificial Intelligence has a lot of applications and is used in many areas. "We often don't notice it but AI is all around us. It is present in computer games, in the cruise control in our cars and the servers that route our email." (BBC 1). Different goals have been set for the science of Artificial Intelligence, but according to Whitby the most mentioned idea about the goal of AI is provided by the Turing Test. This test is also called the imitation game, since it is basically a game in which a computer imitates a human.
In an analysis of the Turing Test I will focus on its features, its historical background and the evaluation of its validity and importance. First of all, the Test itself doesn't really have any complex features. As described by Haugeland, the procedure of the game is simple. Suppose that we have a person, a machine, and an interrogator. The interrogator is in a room separated from the other person and the machine. The participants in this game use teletypewriter to communicate with one another -- to avoid clues that might be offered by tone of voice, etc.
The object of the game is for the interrogator to determine which of the other two is the person, and which is the machine. The interrogator knows the other person and the machine by the labels 'X' and 'Y' -- but, at least at the beginning of the game, does not know which of the other person and the machine is 'X' -- and at the end of the game says either 'X is the person and Y is the machine' or 'X is the machine and Y is the person'. (29-30) The object of the machine is to try to cause the interrogator to mistakenly conclude that the machine is the other person; the object of the other person is to try to help the interrogator to correctly identify the machine. If a machine passes the test, then it is clear that for many ordinary people it would be a sufficient reason to say that that is a thinking machine. And, in fact, since it is able to with a human and to actually fool him and convince him that the machine is human, this would seem to us quite reasonable. The Turing Test seems to be able to give a simple answer to the complex problem of whether machines can think.
In fact, its simplicity is one of the things that have made the Turing Test resist time and history. The Turing Test has had a rich history since its creation by Alan Turing. But first of all, who was Alan Turing? As Whitby notes Alan Turing was a superb British mathematician. During World War II Turing worked in secrecy for the British military to break the German military codes together with some other scientists using some machines that had some characteristics of the modern computers.
(12) After the War, a machine was built in Manchester from which "all modern computers are descended." (12) In 1948, Alan Turing was writing programs for this machine and was also writing the paper "Computing machinery and intelligence", published in 1950, from which the concept of the Turing Test was derived later on. (Whitby 13). An important part of the test history are the occasions of computers being tested and whether or not they passed the test. In the years following its creation and until our days, many computers have taken the test. Turing made himself the prediction stated below: I believe that in about fifty years' time it will be possible to programme computers, with a storage capacity of about 109, to make them play the imitation game so well that an average interrogator will not have more than 70 per cent chance of making the right identification after five minutes of questioning. (qt d in Hodges) We must say that although today the storage capacity of digital computers has reached the levels predicted by Turing, the level of today's programs isn't quite what Turing expected.
A quite recent event in which computer programmes are tested is the Loeb ner Prize Competition that takes place annually and in which computer programmes are investigated by a "limited version" of the Turing Test. These programs are still very far from the standard predicted by Turing. (Oppy et al) In the competition of year 2000 - the year predicted by Turing as a deadline for machines having passed the test -- a programme called A. L. I.
C. E. was the winner but none of the programs of the competition came close to really passing the Turing Test. (Hodges) On the other side, there are some that think that there have been computers that have passed the test.
Floridi argues than in a certain way, Turing was "over-pessimistic", because since 1967 there was a program called ELIZA that was able to and "improvise" around a topic. (135) In fact ELIZA was once "inadvertently" tested. A business executive was trying to contact a member of his staff in the same building by typing messages using a computer terminal but without realizing it he was answered by ELIZA. And he made a discussion with her as if he was talking to a human. (Copeland 39) Because of this happening several people claim that ELIZA passed the Turing Test. However Copeland argues that this is not correct because the procedure was not repeated and some specifications of the Turing Test were not met: the executive had no idea he could have been talking to a computer, there were only 2 participants rather than three, etc.
(39) Undoubtedly, in all its history the Turing Test has always been a matter of discussion with scientists emphasizing its virtues or others criticizing it. As Millar mentions in his article, one of the main virtues is that Turing Test is an "operational definition" that evaluates a computer and can be used as a "criterion" in evaluating intelligent machines. (1) The importance that the test has had in the field of Artificial Intelligence is certainly indisputable. There have been thousands of programs that have tried to past the test and there is no book on Artificial Intelligence that doesn't mention the Turing Test. However there are also many critics and objections to the Turing Test. Copeland notes that we can summarize them into four main objections: The chimpanzee objection is based on the fact that other creatures (e.
g: chimpanzees) which are able to think, fail the test. From this arises that a thinking computer can fail the test because its responses are not human. (44) The sense organs objection says that the test evaluates only the computer's verbal abilities without considering whether the computer knows what the words mean in the real world. Those that support this objection argue that computers should be equipped with sense organs (e.
g: a television eye). (45) The simulation objection is based on the fact a simulated thing is not the actual thing. A computer passing the test doesn't show that it thinks, it shows that is an excellent simulation of a thinking subject. (46-7) The black box objection comes from the concept of a device whose inner operations are unknown. The computer taking the Test is like a black box because the test evaluates it only from the outside.
There is still much debate around Turing Test and scientists of AI have different opinions about it. However there are some facts of which we can be sure of. The Turing Test was invented by a great scientist, it has had a long and rich history of 55 years and has played an important role in the science of Artificial Intelligence. Bibliography Artificial Intelligence. BBC. Mar.
3, 2005 Copeland, Jack. Artificial Intelligence: A philosophical introduction. Oxford: Blackwell, 1993. Floridi, Luciano. Philosophy and Computing: An introduction. London: Routledge, 1999.
Haugeland, John, ed. Mind Design II: Philosophy, Psychology, Artificial Intelligence USA: MIT Press, 2000. Hodges, Andrew. Alan Turing and the Turing Test Mar. 15 2005 < web P. H.
"On the point of the Imitation Game." Mind, New Series, Vol. 82, No 328 (Oct. 1973): 595 par 1. Mar.
20, 2005 Oppy, Graham, Down, David, The Turing Test, The Stanford Encyclopedia of Philosophy (Summer 2003 Edition), Mar. 10, 2005. Whitby, Bay. Artificial Intelligence: a beginner's guide. Oxford: Oneworld, 2004.