CMU Artificial Intelligence Repository
 
   
   
   
   
  
Chaitin: The Limits of Mathematics
pubs/books/online/math/chaitin/
This book features a definitive reformulation of algorithmic
information theory with new more constructive definitions of
program-size complexity. It is a revised version of the course notes
given to the participant at the limits of mathematics short course,
University of Maine, Orono, Maine, June 1994.
The book discusses a new definition for a self-delimiting universal
Turing machine (UTM) that is easy to program and runs very quickly. It
then uses this to provide a new foundation for algorithmic information
theory (AIT), which is the theory of the size in bits of programs for
self-delimiting UTM's.  This new self-delimiting UTM is implemented
via an extremely fast LISP interpreter written in C. Source code for
the LISP interpreter is included in the book.
This directory includes a copy of a 4 page summary of the book and the
270-page book itself, in LaTeX, dvi, and PostScript formats.
Three different versions of the book (and source) are included, in the 
three directories lm_1, lm_2, and lm_3. The difference between the
three is as follows:
   An N-bit formal axiomatic system cannot enable one to exhibit any
   specific object with program-size complexity greater than N+c.
   An N-bit formal axiomatic system cannot enable one to determine more
   than N+c' scattered bits of the halting probability Omega.
   LM I:
  c = 2359 bits and c' = 7581 bits.
   LM II:
  c = 1127 bits and c' = 3689 bits.
   LM III:
  c =  994 bits and c' = 3192 bits.
   LM IV: 
         c =  735 bits and c' = 2933 bits.
Origin:   
   Email to chao-dyn@xyz.lanl.gov with Subject line
      get 9407003
Version:      Draft of 7-JUL-94
Requires:     LaTeX or PostScript
Updated:      14-JUL-94
CD-ROM:       Prime Time Freeware for AI, Issue 1-1
Author(s):    Gregory J. Chaitin 
              IBM, P O Box 704
              Yorktown Heights, NY 10598
Keywords:
   Authors!Chaitin, Books!Limits of Mathematics, 
   Interpreters!Lisp, Lisp!Implementations
Contains:
   lm_?/lisp_src.tgz	source code from the book
   lm_?/limits.tgz      250 page book, with LaTeX, dvi, ps
   summary.tgz          4 page summary, with LaTeX, dvi, ps
References:
   Gregory J. Chaitin, "Algorithmic Information Theory", Cambridge
   University Press, 1987.
   
   Gregory J. Chaitin, "Information-theoretic incompleteness", Applied
   Mathematics and Computation 52:83-101, 1992.
   
   Gregory J. Chaitin, "Information-Theoretic Incompleteness", World
   Scientific, 1992. 
Last Web update on Mon Feb 13 10:38:30 1995 
AI.Repository@cs.cmu.edu