Inductivist Foundation for the Turing Test: A Peircean-Hempelian Perspective.

Asodun, F.O (2014)

A Thesis Submitted to the School of Postgraduate Studies, University of Lagos.

Thesis

There is ample evidence in Artificial Intelligence (AI) research indicating that appropriately programmed machines are capable of performing functions that require intelligence when performed by humans. However, there is no available correlative record that any such machine has been able to pass the Turing Test; an intelligence test purposefully designed by Alan Turing to actualise his objective of justifying machine intelligence. This work attempts to resolve this paradox. Consequently, employing the critical method of conceptual analysis native to philosophical inquiry, the work subjects the Turing Test to rigorous philosophical investigation. Findings show that inappropriate criterion resulting from the misinterpretation of the test by some scholars has all along been used to misjudge machine’s performance in the test. The criterion stipulates that machines need to possess the human properties of brain, mind and consciousness in order to pass the test. Since the criterion could not be met, machines could not pass it. However, further findings show that the appropriate criterion is actually anchored on machine’s “imitative” ability. Turing believes that if in the course of imitating human functions machines exhibit appropriate intelligent behaviours that make them indistinguishable from humans, that could serve as sufficient evidence to infer that they are also intelligent irrespective of the availability of evidence that they possess brain, mind and consciousness or not. This work interprets Turing’s conclusion to be based on inductive reasoning, and argues that machines, if appropriately programmed, might pass the test on inductive grounds. Consequently, in a bid to put the Turing Test in its appropriate theoretical perspective so as to attain Turing’s objective, our research presents a model of induction that will serve as an inductivist foundation for the test. The model, referred to as Peircean-Hempelian inductive reasoning, offers the logical template both for inductive generation of hypotheses on human intelligent behaviour and inductive confirmation of machine’s imitative ability of such behaviour as a proof of intelligence in the Turing Test. The template also offers an appropriate interpretation of the Turing Test. Incidentally, in actualising Turing’s objective, the work aspires to raise awareness on the importance of exploiting the idea of machine intelligence for human benefit. Pursuant to that, our research shows that exploiting the potentiality of ICT - an AI paradigm for processing, maintenance and transmission of information - in the area of educational development, shall have positive impact on the field of education, philosophy of education and the educational system of Nigeria generally.

Collections: