Re: Slashdot - The Computational Requirements for the Matrix

From: Hal Finney (hal@finney.org)
Date: Mon Jun 02 2003 - 12:42:36 MDT

  • Next message: Lee Daniel Crocker: "Re: Buffalo sentence (sidebar to Yakyakian Sentence)"

    Robin wrote:
    > Hal Finney wrote:
    > >The idea is that the strings are self-delimiting, so that the part
    > >after the end doesn't count. Therefore, among all infinite strings, the
    > >fraction of them which correspond to a particular finite, self-delimited
    > >string of length n is equal to 2^-n.
    >
    > Is the claim that it is impossible to define say a uniform distribution
    > over all self-delimited strings, or is it that we know the "true"
    > distribution over all infinite strings and that distribution implies the
    > feature you describe?

    I would say it is the latter. We assume that the true distribution over
    the infinite strings is the one you get if each character in each position
    has an equal probability. That seems like the simplest assumption.

    Hal

    This archive was generated by hypermail 2.1.5 : Mon Jun 02 2003 - 12:56:59 MDT