From: Eugene Leitl (Eugene.Leitl@lrz.uni-muenchen.de)
Date: Thu May 31 2001 - 03:51:25 MDT
On Thu, 31 May 2001, Emlyn wrote:
> Well, wait a second. From inside the sim they are identical... from
> outside they are clearly not; they might even each map onto something
Of course they're identical, whether from inside or the outside, just
do a bit level comparison: the diff is zero.
> as crude as separate process ids. From each being's subjective
If you add additional bits to the bit vector describing a complete person
state you can make them differ, but not in a functional sense. The label
is not used in processing. As soon as the label bits enter the algorithm,
and as soon as label bits are different you have a bifurcation. Boom.
Once again: the label is an external state. It is not used in the
algorithm. As soon as it is used in an algorithm, it ceases being just an
external label.
> position, they are not identical... that being can say for certain "I
> am me, and the other guy is not". Only to an external observer inside
If you're in perfect synch, you _can't_ bifurcate. You can be aware of it,
but this doesn't matter: the other guy is also aware of it. Whatever you
think, you both think exactly the same thoughts in unison.
> the sim are the processes (intelligences) identical, and only given a
> highly contrived environment. That's a pretty unreasonable basis on
> which to declaim the processes "identical".
Identity is a boolean metric, which only occurs in such contrived
environments. This is similiar to two systems both being in the exact the
same quantum state: they both are nondistinguishable, but in practice you
can't put two macroscopical systems in exactly the same state (well, okay
Bose-Einstein condensate does come close).
In practice, people are talking about similiarity when they're saying
"identity". Of course two cloned-state people are very similiar
immediately after bifurcation, but it only goes downhill after that. Even
a single bit diff introduced during noncompressed state space evolution
flow will result in exponential deviations. After fork, assuming the diff
doesn't erase itself (only possible under very weird circumstances, and
for very small perturbations), you both are different persons, in fact
drifting away from each other with each subsequent time tick. After a few
years, you'd even look different.
> Me too... in fact, I'd object regardless of whether I knew we'd forked
> or not.
If we haven't forked, and I know it, I would still object on general
principle. I would demand from the operator that he will remove the synch
boundary condition, and allow the other clone to become a person (or me a
person, whatever).
> > I don't get it. This topic is so dead, it's not even smelling
> > anymore.
>
> heh heh!!! I apologise in advance to you Eugene, and commiserate, that
> the world will continue to inflict this topic on you, for a long, long
> time...
But this doesn't mean I have to deal with that topic. I already wanted to
deal with this thread via the DEL button, but not for tedium of the
dayjob.
This archive was generated by hypermail 2.1.5 : Sat Nov 02 2002 - 08:07:51 MST