Artificial Intelligence is covered in it's own site section.
Intelligence is one of those amazing terms that everyone knows the meaning of until you ask them to explain. Many people are happy to accept that IQ (Intelligence Quotient) is an accurate measure of intelligence but there is considerable evidence which will disagree with the conclusion. Intelligence is a multi facetted concept. Some general agreement as to what intelligence is would include:
It is also believed that intelligence has a cultural context. What is considered intelligent in one culture may not be considered so in another. Whether in the final analysis, artificial or machine intelligence will be judged using the same measures as human intelligence is unknown; but unlikely.
With the emergence of the science of artificial intelligence, the parallel question of whether computers will ever equal or surpass the intelligence of man has raged. It is the two opposing camps that get the most publicity. Those who feel that computers will never be as intelligent as man discuss some unique properties of man which are fundamentally unachievable in machines. Some of these properties relate to the uncertainty principal known by most physicists.
Clearly, present day computers perform their tasks in an entirely different way to human brains. Even attempts to model brain cells or neurons using computer simulations falls a long way short of the work done in a real neuron.
Yet, the science of artificial intelligence is developing and is pushing back the boundaries of what is possible. Will these boundaries continue to be pushed back or will we reach some important point where no further development is possible?
Are machines destined to become the next dominant form of life on earth?
"The capacity for understanding,The ability to perceive and understand meaning."
The Collins English Dictionary
Other works follow more detailed models of intelligence such as a detailed grid of properties and a hierarchical model of intellectual abilities.
Psychometrics means: Measurement of the mind.
IQ tests are an example of psychometric testing that provide a measure of how good people are at solving problems of certain kinds under test conditions. They are not intended to provide an account of the process by which people solve problems.
Psychometric testing has been used in various selection tasks for people. Cyril Burt used IQ testing to identify children from disadvantaged backgrounds who might benefit from a more academic, grammar school education. In both world wars, mental testing was used to identify the most appropriate personnel for difficult jobs, eg fighter pilots. Mental testing was extensively used by the US armed forces during the first world war.
Other psychometric tests such as EPI and MMPI concentrate on personality testing and may not be relevant to the study of thinking and reasoning.
Tests with a specific objective, such as selecting a good fighter pilot, are fairly easy to validate. Validation of a general IQ test is more difficult because objective criteria for determining intelligence are difficult to find. Because of this, there has been some attempt to show that psychometric testing itself is a valid measure of intelligence.
There was then debate concerning the validity of testing for a general intelligence (such as IQ) (Spearman) or whether the separate components of intelligence such as spatial and verbal skills should be tested independently (Thurstone). Attempts to find correlation between the two approaches proved difficult.
In the 1950s and 60s, psychometric testing was used to find out what creative people were like and how they differed from other people. Roe (1952) examined American scientists who were deemed to be creative by peers. Cattell & Drevdahl also studied creative scientists in the same way (1955) and artists (1958). These tests provided some generic characteristics for each group and (not surprisingly) found artists to be more sensitive and have more inner tension than scientists. Subsequent studies (Tardif & Sternberg 1988) have identified a wide variety of cognitive characteristics in creative people.
Guilford and others have highlighted the difference between convergent thinking (required to answer most IQ tests) and divergent thinking. Many believe that divergent thinking is linked to creativity although this has been hard to prove. Those who believe in divergent thinking campaign against rote learning and the use of IQ test.
Attempts to use psychometric tests to assess creativity had questionable success. In some cases the only success was the correlation between these tests and IQ tests, which had already been shown to be poor measures of creativity.
Some more recent testing has shown correlation between 0.4 and 0.6 for creativity assessed by peers. Such test include (Barron 1988) the symbolic equivalence test where people have to produce metaphors for images described to them verbally. Responses are scored for acceptability and originality (how many people produce the same response).
Chapter 14 of "Cognitive Psycology: Robert J. Sternberg: Harcourt Brace. 1996" provides an excellent discussion of intelligence in general and human and artificial intelligence in particular