On 18 Jul 2016, at 03:54, Bruce Kellett wrote:
On 16/07/2016 4:28 pm, Bruno Marchal wrote:
On 15 Jul 2016, at 02:07, Bruce Kellett wrote:
On 15/07/2016 9:42 am, Jason Resch wrote:
I printed the following "Duplicate Questionnaire" and gave one to
both John-Washington, and John-Moscow. The questionnaires each
had 8 questions:
1. What city did you last recall being in?
2. How many cities do you see now?
3. What is the name of the city you see before you?
4. True/False: You see two cities right now:
5. True/False: The prediction that you see Washington was:
6. True/False: The prediction that you see Moscow was:
7. True/False: The prediction that you see Washington and Moscow
was:
8. True/False: The prediction that you see Washington or Moscow
was:
When I gave the questionnaire to John-Washington, he filled out
the following answers (in bold):
1. What city did you last recall being in? Helsinki
2. How many cities do you see now? One
3. What is the name of the city you see before you? Washington
4. True/False: You see two cities right now: False
5. True/False: The prediction that you see Washington was: True
6. True/False: The prediction that you see Moscow was: False
7. True/False: The prediction that you see Washington and Moscow
was: False
8. True/False: The prediction that you see Washington or Moscow
was: True
When I gave the questionnaire to John-Moscow, he filled out the
following answers (in bold):
1. What city did you last recall being in? Helsinki
2. How many cities do you see now? One
3. What is the name of the city you see before you? Moscow
4. True/False: You see two cities right now: False
5. True/False: The prediction that you see Washington was: False
6. True/False: The prediction that you see Moscow was: True
7. True/False: The prediction that you see Washington and Moscow
was: False
8. True/False: The prediction that you see Washington or Moscow
was: True
Both Johns expressed deep regret over insulting people on the
Everything list, most especially Bruno. It turned out neither
John-Washington's, nor John-Moscow's prediction that they would
see both cities was true from their own first person points of
view.
But you have introduced a distinction between John-W and John-M
that is not present in the original protocol. Remember that the
criterion of personal identity you are working with is based on
person memories (verified by a personal diary if necessary). Both
copies of John have these memories and these diaries, so they both
have equal claims to be John. "John", as this duplicated person,
predicts with certainty that he will see W, and that he will see
M, so he predicts that he will see both cities.
Yes, but only one in all 1p views accessible, and the question is
on the future 1p view, not on the 3-1 views, for the same reason
that when we look at 1/sqrt(2)(up + down) in the {up, down} base we
can predict with certainty that we will see either up, OR down and
never both at once.
I have said several times that probability is a problem for the
Everettian or MWI view. This is not a problem of defining a measure
over a possible infinite number of worlds -- though that is
certainly a problem that has not really been solved -- but the main
difficulty lies in the observation that probability makes little
sense in a situation in which everything possible does happen.
The beauty and elegance of Mechanism is that it solves the problem. In
a reality where a computation can differentiate, we have 3p
determinacy, and a subjective, or first person, indeterminacy which
occurs, and admit a trivial explanation, in term of memory access or
self-reference.
So there is no workable notion of probability in the Everettian
multiverse.
Quite the contrary, accepting just the most common of all theories in
cognitive science: the brain is some machine.
Standard quantum mechanics gets around this in a fairly
straightforward way: the "other worlds" in which alternative
outcomes occur are disjoint, with no possible future interaction
with the world in which we find ourselves. Such alternative outcomes
can thus be safely ignored because they can have no possible effect
on the observer or on his future evolution. Decoherence ,and the
irreversibility of completed experimental outcomes, thus reduce QM
to an effective collapse situation -- there is no physical collapse,
but FAPP the other worlds have vanished from existence.
Then add in all steps of the UDA that we change the protocol so that
the copies never met, and the proof will remain valid, showing that
your remark is not relevant.
It is even less relevant that the copies in the arithmetical reality
cannot interact either. Duplication in one world is used as a
pedagicial tool to express a problem, motivates the definitions, and
appreciate the tiny bit of solution already given by simple known
Löbian machines.
The fact that this appears odd is that our conventional intuition
is essentially dualist -- we think that there is a central core
that is the "real me" that gives me my continuing sense of
personal identity. This intuition breaks down when you have
duplication of persons.
Then computationalism, and Everett QM breaks down. Computationalism
guaranties that whoever you will become, it is lived and felt as
one unique person in one city, or you bring some telepathy of a
kind not possible with computationalism with the protocol discussed.
Everett breaks down only if you try and imagine that the observer
actually lives in the multiverse. But he does not -- the observer
only ever observes one world. He might be uncertain as to which
world he will inhabit in the future (which experimental outcome he
will observe), but this does not introduce the paradoxes of person
duplication in one world.
Which paradox? There is no paradox, only an unavoidable first person
indeterminacy. There is nothing paradoxical, as this indeterminacy is
a simple consequence of self-duplicability. Even PA proves it exists
conditionally to self-consistency.
As you say in another post, computationalism depends on the
breakdown of transitivity for personal identity: M is the same as H;
W is the same as H; but M is not the same as W. Given this, you have
all sorts of problems with the nature of personal identity -- maybe
it is not a modal concept! I will talk more about this in reply to
your other post.
Well, the machine notion of 3p-self can be defined in arithmetic, and
all correct machine knows that her 1p-self is not. Sure it is a tricky
notion, but the non transitivity is not a problem, as the "Parfit
person series" will work transitively in all cases, except when
duplication occurs, but why would that cause any problem, you tell me.
Nothing here threats the validity of the reasoning leading to the
reversal physics/arithmetic. I think you confused non transitivity
(the failing of some transitive link) with intransitivity (the failing
of all transitive link). With self-duplication, we lost transitivity
in one case, but both surviver recover it as long as they do'nt
duplicate again, and so the old guy who stayed in Moscow remains the
same young guy who teleported at Moscow through some duplication a
long time ago. You might elaborate on your problem, as I don't see any.
Bruno
Bruce
--
You received this message because you are subscribed to the Google
Groups "Everything List" group.
To unsubscribe from this group and stop receiving emails from it,
send an email to everything-list+unsubscr...@googlegroups.com.
To post to this group, send email to everything-list@googlegroups.com.
Visit this group at https://groups.google.com/group/everything-list.
For more options, visit https://groups.google.com/d/optout.
http://iridia.ulb.ac.be/~marchal/
--
You received this message because you are subscribed to the Google Groups
"Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to everything-list+unsubscr...@googlegroups.com.
To post to this group, send email to everything-list@googlegroups.com.
Visit this group at https://groups.google.com/group/everything-list.
For more options, visit https://groups.google.com/d/optout.
