| Philosophy of Mind | |||||
| Introduction | Dualism | Behaviourism | Identity Theory | Functionalism | Dennett |
| The Turing Test | |
|
The English mathematician Alan Turing (1912-54) developed a test whereby he argued the question of artificial intelligence could be settled. The test is based on something called "The Imitation Game" in which 3 people, each in separate rooms, communicate by Teletype (or some other mechanical means). Each of the 3 people has a specific role: one acts as an interrogator whose job it is to find out what sex the other two people are; the other two (one man and one woman) one whose role it is to answer honestly and the other dishonestly. The interrogator must therefore must answer the interrogator's questions - one honestly and one dishonestly. Turing's version of this game involves replacing one of the people with a computer that has been programmed to deceive the interrogator. If, as with a human subject, the interrogator was deceived a certain percentage of the time (say 70%) then the machine can be said to have passed the test (it is conscious and intelligent). However, there
currently exists no computer that can get anywhere near passing
this test. This is because any interrogator would find it easy to
devise questions which a machine would have difficulty answering.
For instance, even if we only ever asked literal questions - that
is, ones not involving any figures of speech (such as "driving
someone up the wall") - the machine would need to be programmed
with a massive amount of information about language structure, syntax,
grammar, history, culture, etc. Although this does not present a
theoretical barrier, it certainly presents a practical one. |