Re: QTI & euthanasia (brouillon)

From: Bruno Marchal <>
Date: Fri, 7 Nov 2008 18:07:48 +0100

On 07 Nov 2008, at 03:27, Jason Resch wrote:

> On Wed, Nov 5, 2008 at 4:52 AM, Bruno Marchal <>
> wrote:
> Hi Jason,
> Le 04-nov.-08, à 23:21, Jason Resch a écrit :
> > although I agree with Brent, if the simulated world in the
> computer is
> > entirely cut off from causal effects of the physical world where the
> > computer is running, then you have also created an entirely new
> > world/reality.
> I agree with this too. The only thing necessary to understand step 6,
> is that you do survive there like if it was teleportation. And in
> that
> context, the calculus of probability remains the same as in the five
> preceding steps. For example, if you understand step 5, you know that
> if a instantenous of you is done, and is not detsroyed, and if that
> copy is reinstantiated in the virtual Moscow tomorrow, and in the
> virtual Washington in one billion of years, the probability that you
> will "stay here" (and not find yourself in the virtual realities) is
> 1/3 (assuming 1/2 for perfect duplication). It means that Nozick's
> closer continuer identity theory fails with comp.
> My only reservation with the above is I am not sure probabilities in
> expecting your next observer moment work this way. From a third
> person perspective I have a 100% chance of experiencing all 3
> extensions, and as you say, this is interesting because from a first
> person POV you do not experience all 3 locations at once. I think
> this is where I disagree, you _do_ experience all 3 locations at
> once, but due to the isolated locations and lack of communication
> between the 3 different brains, they are unable to merge the
> experience into a memory of being in all 3 locations.

I do agree with you. But in that sense I am already Jason Resch. We
come from the same splitting amoeba. This is true, at some level, but
it does not seem to me relevant for the understanding that physics
*has to* be extracted from probability/credibility measure on

> This is the same reason that though Einstein says we exist in a 4-
> dimensional block where time is only subjective, we never experience
> being in all times at once, due to the limited, low-bandwidth,
> communication from past memories to the present, and complete lack
> of communication from future states to the present. The vast
> majority of information within our brain state at any one time is
> chiefly information of the present and very recent past, giving us
> the feeling of living in the present, when of course our true nature
> is that of a 4-dimensional snake stretching from the time of birth
> to death.
> This lack of communication between individual brains is what fools
> us into believing we are each a unique conscious entity, when truly
> there is nothing to differentiate one observer from any other,
> except for the current content of their experience.

You are right. But when you look for the true reason why apples appear
to fall on the ground. UDA(+AGF, that is 1...8) explains why the comp
correct way to predict the behavior of the apple consists in looking
in the universal deployment, and then looking at all computations
going through your actual states (the one you have (by comp) once you
observet the apple before dropping it), and, looking at the normal
most probable stories/computations going through that state.
  Hmm... perhaps you have a problem with the UD? It does not just
generate OM (Observer Moment). The DU generates all third person
"observer moments" (as instantaneous state of universal self-
observing machine) by generating all the stories (singular
computations) going through that state. There is a continuum (from a
first person pov) of such stories. The first person moment are
different modalities.

> I think it is a mistake to use the memories one has access to as a
> means to delineate observers, for the vast majority of ones memories
> are not in the content of ones OM at any one time.

You are right, but not at the level needed to understand why and how,
assuming comp, we *have* to explain why apples falls from "pure
(mathematical) computer science". (or perhaps you are and there is a
misunderstanding, to be sure). The reasoning should show comp
testable. So we have to take into account all computations going
through each observer moment to have normal relative "expected
values". A bit like already with QM and Everett.

> I think the importance of a particular computational history in
> defining an "observer moment" is not as important as memory/
> communication isolation.

You are right if your goal is to discover who you really are. But not
if your goal is to understand why we have to derive Schoedinger
Equation from computer science and number theory. It makes comp
testable, and it provide a fundamental theory which does not eliminate
the person. On the contrary it explains how person creates reality by
history sharing.
Somehow the memory/communication is a key, to delineate sharable
stories, to extract eventually the right notion of arithmetical
entanglement. But this is not easy matter.

> The Universal Dovetailer shares a single computational history as
> one well-defined short program, and it implements all possible
> observer moments.

Careful. For an infinitely deep reason related to Church Thesis and
the closure of the set of computable functions for the diagonal
procedure (the real difficulty of the seventh step), the Universal
Dovetailer does not generate just all observer moments, it generate
all the computational histories going through or accessing those
observer moments. Those computations obeys the laws of, well ..., of
computations, meaning we have to take at computer science. Eventually
we have to interview the relatively most probable machine).

> Yet would not all OM's it generates be considered the same since
> they share a single computational history?

Gosh! I was suspecting you have a problem with the UD work. The UD
does not generate the third person OMs. The UD generate the many
stories (computations). It generates all programs AND their many
relative implementations. Well this includes it generates also his own
history many times, but also all other stories (Universal!).
Most stories are infinite, that is what it has to zig-zag recurrently
on the stories it generates. Those stories themselves generate the 3-
OM,( or not (depending on the story and also of the precise definition
of OM). Third person OMs are "causally related" through their singular
First person OMs are related to their probable histories which depend
on the infinitely many stories going through them.

> I think it is better to track the flow of information that goes

Yes. (Well your sentence is interrupted) but I agree we have to track
the information going through the stories of those OMs. I am just
saying that with UDA(1...7) we know that we have to take account of
ALL the stories. And with UDA.8 we have to to take account of only the
mathematical even arithmetical 3-stories. But the 1-histories will
explode in the full first person plenitude. Arithmetic as seen from
inside is already bigger than Cantor Paradise or any mathematical
approximation of Plato Heaven. But let us not anticipate perhaps (some
have get the point I think, and it is a remind for them, let us say).

> (Experiencingo ne moment because of no communication from the future
> tothe past, and very low bandwidth rate from the past to the
> present) Surely if I was duplicated to 100 locations (only one of
> which was Moscow), I could wager $1 that I will not appear in
> Moscow, and 99 of my copies will be richer but


> One OM that travels across all OM's, the UD? ...Share computational
> history, same program, it is jsut that the information doesn't get
> linked between them.

All 3-OMs are linked by one computation. But reappears in other
computations as well. Then by UDA: all 1-OMs are linked by infinities
of (and measure on) computations.

Just remember this: the UD generates all computations. For each OMs
there is an infinity of computations going through it. It is really
because he first person cannot be aware of the delays (defined here by
number of steps of the UD to get some OM), that the 1-person
indeterminacy bears on infinite sets of computations.

Don't confuse the (abstract) running of the UD, and the computations
on which the UD is running. I have not the yime now, but I could try
to draw a picture.

Tell me if this has helped a bit, or not.


You received this message because you are subscribed to the Google Groups "Everything List" group.
To post to this group, send email to
To unsubscribe from this group, send email to
For more options, visit this group at
Received on Fri Nov 07 2008 - 12:15:48 PST

This archive was generated by hypermail 2.3.0 : Fri Feb 16 2018 - 13:20:15 PST