Logical Methods for Testing the Effectiveness of Artificial Intelligence Prototypes

Perhaps the most famous test of artificialgiven as results of the use of the axioms and
intelligence is the Turing test, elaborated in 1950.theorems present in the brain. Behaviorism sees
As part of his argument Turing put forward theevery decision made by the mind as the choice
idea of an 'imitation game', in which a human beingof one alternative over another, where the
and a computer would be interrogated underpreferred output is the response that has
conditions where the interrogator would not knowrepeatedly been followed by a reinforcing
which was which, the communication being entirelystimulus.We can see now that a method that
by textual messages. Turing argued that if thewould test the effectiveness of such a machine
interrogator could not distinguish them bywould be entirely based on a chosen interpretation
questioning, then it would be unreasonable not toof the mind. Moreover, not only the test should
call the computer intelligent.There are severalbe based on this interpretation, but also the whole
ways of interpreting and discussing the Turingway the machine is built. It is obvious that the
test, which in effect all lead to the discussing ofTuring test is entirely based on functionalism
some key philosophical views on the mind.because the way computers were built, especially
Considering the mind as an input-output devicein the beginning was based on
with the exterior world, the four mainfunctionalism.Nevertheless, the Turing test has
metaphysical takes on the mind would have fourbeen severely criticized, especially by logicians.
different views of how the decision of choosingThere has even been invented a counter-test, or
an appropriate output to a given input is mediated.experiment to prove its lack of validity. This test,
The dualistic view on the mind would see asthe Chinese Room Experiment, developed by John
responsible for this mediation the right consciousSearle has the main quality that assumes the
experiences, the mind-brain identity theory wouldpremises of the Turing test, namely that the
name the right neurophysiological mechanisms,human mind works according to theory that it has
functionalism the right procedures while thebuilt in. Shortly, the test says that an English
behaviorist would not identify any. Accordingly,native speaker that has no knowledge of Chinese,
these four views would view the mentalis locked in a room and given a set of instructions
processes as phenomenological, neurophysiological,and then two batches of Chinese writings.When
computational, while the behaviorist would see thethe third one is given, he is able to correlate one
mind as lacking any processes but just havingset of formal symbols to another set of formal
qualities resulted from behavioral dispositions.Whatsymbols and thus, using the rules, he is able to
these four views state effectively about the waygive back certain sorts of Chinese symbols with
the mind works can be explained in an easiercertain sorts of shapes in response to certain
way, as follows. The dualistic view implies asorts of shapes given him in the third batch. Thus,
certain awareness of the self, of the pastthe person would be able to communicate in
experiences and is probably the closest to imply aChinese, and convince the interviewer of his
freedom of choice. The mind-brain identity seesabilities, without actually knowing any word of
the mind and the brain as a whole and as the onlyChinese. He produces the answers by manipulating
defining element of humans. All the processes ofuninterpreted formal symbols, but as far as the
thinking can be explained through chemical andChinese is concerned, he simply behaves like a
biological reactions. Functionalism is the closest tocomputer; he performs computational operations
the Turing test and to the most basic way weon formally specified elements. For the purposes
perceive computers. It states that the mind actsof Chinese, he is simply an instantiation of the
as an axiomatic system and all the outputs arecomputer program.