The Library
Machine guessing II
Tools
Miller, David (2012) Machine guessing II. In: J. S. Mill’s Ideas on Induction and Logic of the Humanities and Social Sciences in the Cognitive Research and in Artificial Intelligence Systems, Moscow, Jun 2011. Published in: Voprosy Filosofii, 2012 (8). pp. 117-126. (Unpublished)
|
PDF
WRAP-Machine-Guessing-II-Miller-12.pdf - Other - Requires a PDF viewer. Download (220Kb) | Preview |
Abstract
According to Karl Popper, the evolution of science, logically, methodologically, and even psychologically, is an involved interplay of acute conjectures and blunt refutations. Like biological evolution, it is an endless round of blind variation and selective retention. But unlike biological evolution, it incorporates, at the stage of selection, the use of reason. Part I of this two-part paper begins by repudiating the common beliefs that Hume’s problem of induction, which compellingly confutes the thesis that science is rational in the way that most people think that it is rational, can be solved by assuming that science is rational, or by assuming that Hume was irrational (that is, by ignoring his argument). The problem of induction can be solved only by a non-authoritarian theory of rationality. It is shown also that because hypotheses cannot be distilled directly from experience, all knowledge is eventually dependent on blind conjecture, and therefore itself conjectural. In particular, the use of rules of inference, or of good or bad rules for generating conjectures, is conjectural. Part II of the paper expounds a form of Popper’s critical rationalism that locates the rationality of science entirely in the deductive processes by which conjectures are criticized and improved. But extreme forms of deductivism are rejected. The paper concludes with a sharp dismissal of the view that work in artificial intelligence, including the JSM method cultivated extensively by Victor Finn, does anything to upset critical rationalism. Machine learning sheds little light on either the problem of induction or on the role that logic plays in the development of scientific knowledge.
Item Type: | Conference Item (Paper) | ||||||
---|---|---|---|---|---|---|---|
Alternative Title: | Машинное угадывание II | ||||||
Subjects: | B Philosophy. Psychology. Religion > BC Logic B Philosophy. Psychology. Religion > BD Speculative Philosophy Q Science > Q Science (General) |
||||||
Divisions: | Faculty of Social Sciences > Philosophy | ||||||
Library of Congress Subject Headings (LCSH): | Artificial intelligence -- Philosophy, Machine learning, Logic, Science -- Philosophy, Rationalism | ||||||
Journal or Publication Title: | Voprosy Filosofii | ||||||
Official Date: | 2 August 2012 | ||||||
Dates: |
|
||||||
Volume: | 2012 | ||||||
Number: | 8 | ||||||
Page Range: | pp. 117-126 | ||||||
Status: | Peer Reviewed | ||||||
Publication Status: | Unpublished | ||||||
Copyright Holders: | D.W. Miller | ||||||
Description: | Original version of the article prior to its translations into Russian and Spanish, which has been subsequently updated |
||||||
Conference Paper Type: | Paper | ||||||
Title of Event: | J. S. Mill’s Ideas on Induction and Logic of the Humanities and Social Sciences in the Cognitive Research and in Artificial Intelligence Systems | ||||||
Type of Event: | Conference | ||||||
Location of Event: | Moscow | ||||||
Date(s) of Event: | Jun 2011 | ||||||
Related URLs: | |||||||
Contributors: |
|
Request changes or add full text files to a record
Repository staff actions (login required)
View Item |