On Fri, 3 Mar 2000 GSLevy.domain.name.hidden wrote:
> jqm1584.domain.name.hidden writes:
> > > Infinity x2 = infinity, infinity/2 = infinity.
> >
> > That's one of the first topics I discussed on this list. My
> > answer now is the same as it was then. First, we know that some
> > observations do have more measure than others, because if not we'd see
> > only 'white rabbits'.
>
> I kind of agree with you that some macroscopic observations are more probable
> than others but only because the microscopic "white rabbits" tend to cancel
> out. (In fact, if it wasn't for the microscopic white rabbits, there wouldn't
> be any world - white rabbits make the world go round :-) ) So, I agree
> that, from the first person point of view, the probabilistic measure could
> be higher for some macroscopic event or object than for others.
Why do you say WRs cancel out? I disagree completely.
> > Second, in physics we deal with ratios of infinite
> > quantities all the time. The key is to use a limiting process in the
> > definition. One can renormalize the measure distribution to make it into
> > an effective probability distribution. If you like you can take the
> > latter to be the fundamental quantity, but it's easier to work with
> > measure so we don't have to worry about the normalization.
> >
> Precisely. In physics we use ratios of infinite quantities because we cannot
> deal with the infinite quantities themselves. Same thing should apply to
> measure. Let's say that the measure of a particular object, say the computer
> in front of you, is infinite. How would you renormalize it? Well, you'll have
> to take a ratio. Divide the measure of the computer by something. By what?
Obviously, you need some general way to take the limit. In the TM
program ensemble, it's easy because one can consider strings of length N
(2^N being finite) and let N go to infinity. I'd say that's the most
sensible way.
In the quantum wavefunction case, where I am comparing numbers of
implementations, the best way seems to be discreting the system in some
natural coordinate system and taking the limit as the step size goes to
zero. (This being in wavefunction space, not 3-space or even 3N-space.)
> >From an observer point of view, the only thing that makes sense is to divide
> the measure of your computer by your own measure - rather than by the measure
> of your shoes or your girlfriend. Using your own measure as the common
> denominator for calculating the renormalized measures of all objects around
> you, will keep your observations consistent and relativistic.
That's total stupidity and bullshit. There is not even an
absolute definition of "you" so it's not even well defined. And what's
needed is an overall measure distribution of observer moments, which that
method could never provide thus giving NO predictions and having NO value.
> In the absolute sense, it may very well be that measure is lost and gained.
> However, from the first person relativistic point of view, measure of self is
> always conserved, and the first person point of view is all that matters.
There is no such thing. The measure distribution of observer
moments is all that matters. If absolute measure is lost, that's
effectively a chance for death.
- - - - - - -
Jacques Mallah (jqm1584.domain.name.hidden)
Physicist / Many Worlder / Devil's Advocate
"I know what no one else knows" - 'Runaway Train', Soul Asylum
My URL:
http://pages.nyu.edu/~jqm1584/
Received on Fri Mar 03 2000 - 14:45:31 PST