RE: Evolution of Beliefs and Preferences

From: Lee Corbin (lcorbin@tsoft.com)
Date: Sun Jun 15 2003 - 22:00:34 MDT

  • Next message: Mike Lorrey: "RE: Dollars and donuts, was: Re: Was Re: PHYSICS: force fields (RANT)"

    Robin writes

    > I don't mind discussing this with you; I just didn't see where you
    > were going with your questions. Care to rephrase?

    Yes, I'll give it another try.

    > > > > 2. What proportion of human beings in the United States would
    > > > > you guess to be Bayesians?
    > > >
    > > > There has never been and will never be an exact Bayesian - it
    > > > is computationally intractable.
    > >
    > >Ah yes, I should have said *non-exact* Bayesians!
    >
    > It all depends on how close you look. From a distance people are
    > roughly Bayesian in many ways.

    Yes. But it's difficult to conceive how it could have
    turned out differently.

    > But if you look real close at particular decisions you can
    > certainly see a lot of error, and sometimes you can also see
    > systematic deviations. So how big a deviation is "big"?

    All right. So no Bayesians per se exist, but it's an ideal
    that one speaks about, such as "rational". No exactly rational
    people exist either. But for rational, we are able to reasonably
    assert that, for example, one nation's population is more rational
    than another nations (on perhaps a particular restriction).

    But I don't get the parallel for "Bayesian". I use Bayes theorem
    a lot in solving probability problems, have solved all the ones
    I've seen on this list both intuitively (mathematically), and
    formally (mathematically). I have always preferred the point
    of view that readily accepts or postulates, or even agitates
    for prior distributions, and goes right ahead with probability
    calculus. So who is more Bayesian than I am?

    What is the utility that you and the researchers find in referring
    to a class of such agents by that description? (Eliezer once wrote
    "[Vorlon voice:] We are all Bayesians now.")

    A semantic analysis by context appears to show that it is being
    used as a substitute for "clear-thinking", or "rational", or some
    blend of such familiar concepts. It frankly sounds silly to me,
    but (a) I did miss out on a lot of earlier Extropian and other
    list discussions about this (b) if enough people I trust really
    say the Emperor has clothes, then it's probably so, but *I* still
    can't see them.

    Lee



    This archive was generated by hypermail 2.1.5 : Sun Jun 15 2003 - 22:09:41 MDT