home *** CD-ROM | disk | FTP | other *** search
- Xref: sparky comp.theory:2445 sci.math:15104 sci.crypt:4908
- Newsgroups: comp.theory,sci.math,sci.crypt
- Path: sparky!uunet!charon.amdahl.com!pacbell.com!ames!agate!stanford.edu!CSD-NewsHost.Stanford.EDU!Sunburn.Stanford.EDU!pratt
- From: pratt@Sunburn.Stanford.EDU (Vaughan R. Pratt)
- Subject: Re: Cryptography and P=NP
- Message-ID: <1992Nov17.043640.5525@CSD-NewsHost.Stanford.EDU>
- Sender: news@CSD-NewsHost.Stanford.EDU
- Organization: Computer Science Department, Stanford University.
- References: <1992Nov15.110945.19939@ringer.cs.utsa.edu> <15115@ember.UUCP> <Nov.16.16.59.47.1992.6436@remus.rutgers.edu>
- Date: Tue, 17 Nov 1992 04:36:40 GMT
- Lines: 36
-
- In article <Nov.16.16.59.47.1992.6436@remus.rutgers.edu> clong@remus.rutgers.edu (Chris Long) writes:
- >Sure, but P=NP and P<>NP are certainly not equally likely. The a priori
- >evidence that P<>NP is *overwhelming*; to claim that P=NP has a chance
- >of being true is worse than claiming that there is a chance of only a
- >finite number of twin primes existing. Both are unproven, but there are
- >*very* strong reasons for believing them to be true. So strong that
- >anyone who believes otherwise must be either ignorant or a crackpot.
-
- If one perturbs the definition of primality slightly while retaining
- its basic statistical characteristics (mean, variance) then one can
- actually prove such theorems as "There are infinitely many twin primes"
- and "Every even number > 3 is the sum of two primes" (Goldbach's
- conjecture) without much difficulty. (I believe "lucky" numbers
- provide an instance of this general principle.) In these proofs the
- statistical evidence is overwhelming, encouraging one to assign
- extraordinary odds to the failure of Goldbach's conjecture.
-
- (I will with no misgivings bet you ten thousand dollars at a thousand
- to one odds that no counterexample to Goldbach's conjecture will appear
- in the next ten years. If one is found I owe you ten thousand dollars,
- otherwise you owe me ten dollars. Anyone who's played for a while with
- Goldbach's conjecture will see that this is very easy money for me.)
-
- There is *no* comparable statistical evidence for either P=NP or
- P!=NP. The evidence to date for P!=NP is that a lot of people have
- looked for a fast algorithm for the generic NP problem and failed. If
- you ask me what evidence there is to date for P=NP I will reply that an
- even larger set of people have looked for a proof of P!=NP and failed.
- If we judge these arguments only by the extent of the search in each
- case, an otherwise disinterested bettor would say that the a priori
- evidence that P=NP is *overwhelming*.
-
- Moshe Vardi has made P=NP his odds-on favorite, I admit I'm a
- conservative old coward in sticking to even odds.
- --
- Vaughan Pratt A fallacy is worth a thousand steps.
-