Re: Decision theory

From: Jacques M Mallah <>
Date: Wed, 30 Dec 1998 16:41:47 -0500

On Tue, 29 Dec 1998, Wei Dai wrote:
> I realize now the problem with decision theory is really about the absence
> of free parameters in a physical theory, and the problem is practical, not
> metaphysical. So let me redescribe it. Decision theory depends on a
> physical theory to compute the consequences of actions, statements like
> "If I do X, Y will happen." But a physical theory with no free parameters
> cannot be used for this purpose. Either the theory says I in fact do X, in
> which case Y will happen but it's the same Y for every X, or it says in
> fact I don't do X, in which case it doesn't give any predictions about Y
> at all.

        The problem would seem to arise when it is not possible to
calculate the effect of a decision within the physical theory. However,
if that did happen, I doubt it would be a problem because with only one
physically possible option, a decision is hardly needed.
        In most cases, of course, and certainly in practical cases, it is
not a problem in principle. Both courses of action can be treated
without having to assume violation of any physical laws.
        Certainly any theory with no free parameters will predict many
cases in which a being like you will be faced with that decision, and in
some of them he will make one decision, in the rest he will make the
other. You can find out which of those cases you are in by making the
decision, and you can make the decision because the theory predicts the
results for both cases. Of course there must then be a small chance that
a 'random factor' might cause you to make the 'wrong' decision; QM
certainly predicts that there is such a chance, ditto a theory with 'all
possible structures'.

                         - - - - - - -
              Jacques Mallah (
       Graduate Student / Many Worlder / Devil's Advocate
"I know what no one else knows" - 'Runaway Train', Soul Asylum
            My URL:
Received on Wed Dec 30 1998 - 13:45:13 PST

This archive was generated by hypermail 2.3.0 : Fri Feb 16 2018 - 13:20:06 PST