Why is it assumed that a multiple "runs" makes any difference to the
If the computation is reversible we could run the simulation backwards -
even though the initial state make seem contrived because it leads to a
low entropy at the end of the computation. Given that the simulated
beings don't know the difference (their subjective time runs in the
direction of increasing entropy) the fact that the simulation is done in
reverse is irrelevant to them.
Would a simulation done in reverse contribute to the measure?
Once we say that all possible computations exist in the Platonic sense,
it seems to me that running them is irrelevant. Of course it is agreed
that the existence hypothesis tells us nothing about their relative
measure. Does anyone have some principles to go by?
I presume a theory of measure along the lines described by Jesse would
need to account for the measure of mappings between computations.
Presumably a simple correspondence would have higher weighting than some
complicated mapping between two computations.
> -----Original Message-----
> From: Jesse Mazer [mailto:[EMAIL PROTECTED]
> Sent: Saturday, 17 January 2004 4:56 AM
> To: [EMAIL PROTECTED]
> Subject: Re: Is the universe computable
> Eugen Leitl:
> >I have trouble alternating between the internal and the external
> >view. So we have a machine crunching bits, sequentially falling from
> >state. This spans a continous trajectory. We can make a full record
> >trajectory, eliminating a time axis. When does the subjective
> >existence assemble into place? The first time the computation was
> >I have trouble seeing my subjective observer experience as a sequence
> >frames, already computed. Is the first run magical, and the static
> >dead meat? I'm confused.
> I think the most common theory on this list is that there is nothing
> about the first vs. the second run--rather, the total number of runs
> determine the measure of that subjective experience. If I scan my
> a computer knowing that my first experience after being uploaded will
> on what environment is created for my simulated brain, I should make
> friends do lots of runs where the upload wakes up in an idyllic
> and that my enemies don't get their hands on the program and do a lot
> runs where the upload is used as a slave or something. If my enemies
> manage to get a copy and do a few of those runs before they are caught
> stopped by the Society for the Prevention of Cruelty to Uploads, my
> can at least try to minimize the damage by doing so many runs of the
> environment that the probability of having that experience after I
> will be much greater than the probability that my first experience
> being uploaded will be waking up as a slave to my enemies (according
> particular theory of how measure works, anyway).
> Jesse Mazer
> Get a FREE online virus check for your PC here, from McAfee.