On 10 Jan 2013, at 16:37, Quentin Anciaux wrote:
2013/1/10 Bruno Marchal <marc...@ulb.ac.be>
On 09 Jan 2013, at 20:02, Quentin Anciaux wrote:
2013/1/9 Bruno Marchal <marc...@ulb.ac.be>
On 09 Jan 2013, at 12:10, Quentin Anciaux wrote:
Hi,
let us start with the proposed QS experiment by Tegmark,
I publish this before. It made some physicists rather nervous
against me, so that I find worthy to vindicate it. I propose the
comp suicide and immortality even well before.
OK, this is only anecdote. But you can see that I made the "Tegmark
point" in my 1991 "Mechanism and Personal Identity" paper, i.e.
the point that the witnesses are increasingly astonished, and not
the experimenter, who can actually easily predict that
astonishment. I made that point to illustrate the relativity of the
points of view in the comp setting, and the fact that the HP events
(the first person white rabbits) although first person impossible,
are still possible and highly probable in the 3p view of the first
person of others. David Nyman's heuristic makes me think that they
could be zombie, but I am not sure this can work with comp. It is
not an important point, as we don't need this for the UDA.
a QS machine with a 99/100 chance of a *perfect* kill (so let's put
aside HP failure or whatever so to have either the experimenter is
killed with the given probabilities or it is not, no in between, so
in 1/100 he is not killed and perfectly well, 99/100 he is killed).
You are a witness of such experiment, and you're asked to make a
bet on the experimenter surviving (or not).
So you bet 100$, if you bet on the experimenter surviving, if he
survive, you'll get 200$, if he does not you'll lose your bet,
likewise if you bet on him die.
What you should do contrary to what seems reasonable, is to bet on
the experimenter will survive for the following reason:
If MWI is true:
1st Test: in 99/100 worlds you lose 100$ (and the bet ends here,
there is no experimenter left for a second round), in 1/100 worlds
you win 200$
2nd Test: well... you cannot play again in the 99/100 worlds where
you did lose 100$, so you start already with 200$ in your pocket
for this 2nd test, so you should do the same, no here in 99/100
worlds, you did make a draw (you put 100$ in 1st test + 100$ win on
the 1st test - 100$ you did lose now because the experimenter is
dead), in 1/100 you win again 200$, that make 300$ in your pocket.
From the 3rd test on, you can only get richer, weither the
experimenter lives from your POV or not.
In QM+collapse, if the guy luckily survive two tests, you win
money... you'll only lose money if he is killed at the first test.
So contrary to what you may think, you should bet the experimenter
should live, because in MWI, it is garanteed that you'll win money
in a lot branches after only two succeeded test, and as in QM
+collapse, only the 99/100 of the first test lose money, all the
others either make no loss or win money.
OK. But the probabilities for any amount of money that you can win
individually remains the same with MWI and collapse. MWI is just
more "fair ontologically", because all the possible winners exist,
and indeed the descendent of the two first win have got something,
but they got it with the same probability with the collapse, at
each state of the procedure. They just don't exist, in the "non
lucky" collapse scenario.
You give only a reason to prefer more, or to fear more (if you
think to the bad rare events), the MWI than collapse.
What would you say to someone telling you that he prefers collapse,
as with collapse, you have 1/100 to win some dollars, and 99/100 to
lose, but there will be only one winner possible and only one
loser. And in the MWI, there is always one winner and 99 losers!
(times infinity!). So if the question is in making more people
happy and less people unhappy, may be collapse is preferable at the
start (with that kind of reasoning).
For the witnesses, your bet is more "socially fair", but not in way
making possible for them to test MWI or ~MWI.
I still stand on "repeated improbable outcome" implies either MWI
or QM false.
If it's not the case then a 1/10⁶ probability outcome doesn't mean
anything... if you notice 10⁹ validated outcome of a prior
probability of 1/10⁶ I would say your prior probability calculus
is wrong, if it's from your theory, I would say that your theory
has been disprove. The point is in QM+collapse such outcome as
1/10⁶^10⁹ probability of occurence, it could not happen in our
current universe lifetime *without* a *very good* explanation
principle. Hence if that happened, I would say QM+collapse is
falsified. *But* in MWI, such outcome **do** happen, probability
calculus is not about happening but about distribution in MWI
(contrary to QM+collapse) so it still stand.
So if you see such event, you're left choosing between a new theory
or MWI... QM+collapse *without* a very good explanation principle
for such improbable occurence should be disproven... In MWI you
have that good explanation principle, which is in MWI it *does*
happen.
OK. But for the witness it happens with the same probability than
with the collapse. So this does not help the witness to decide.
You seem to believe that if P is low, it can no more happen with the
collapse theory,
No, I say it can no more happen in collapse theory without *a very
good* explanation principle. I'm sorry but if the theory predict it
happens with a 1/10⁹ probability of occurence and every time you
test it, it happens... I'd say your prior probability calculus is
screwed, so without a *good* explanation, your theory can be said to
be falsified. As I said, the *good* explanation with MWI is that *it
does* happen.
I would very much like to concede on the MWI appeal, for the witness,
living a quantum finite rare event. But the math tells us that the
probabilities, how small they can be, remains the same for MWI and
~MWI. So my wanting to concede is probably only a measure of my bias
for the MWI, I'm afraid. The final day, the witnesses have seen a
finite rare quantum event. The believer in the collapse among them
will just say "well, the amplitude of probability was not null".
Bruno
Quentin
but then MWI and ~MWI can be distinguished without QS, in some 3p
way, and this is highly dubious.
Bruno
Quentin
Bruno
Quentin
--
All those moments will be lost in time, like tears in rain.
--
You received this message because you are subscribed to the Google
Groups "Everything List" group.
To post to this group, send email to everything-
l...@googlegroups.com.
To unsubscribe from this group, send email to everything-list+unsubscr...@googlegroups.com
.
For more options, visit this group at http://groups.google.com/group/everything-list?hl=en
.
http://iridia.ulb.ac.be/~marchal/
--
You received this message because you are subscribed to the Google
Groups "Everything List" group.
To post to this group, send email to everything-
l...@googlegroups.com.
To unsubscribe from this group, send email to everything-list+unsubscr...@googlegroups.com
.
For more options, visit this group at http://groups.google.com/group/everything-list?hl=en
.
--
All those moments will be lost in time, like tears in rain.
--
You received this message because you are subscribed to the Google
Groups "Everything List" group.
To post to this group, send email to everything-
l...@googlegroups.com.
To unsubscribe from this group, send email to everything-list+unsubscr...@googlegroups.com
.
For more options, visit this group at http://groups.google.com/group/everything-list?hl=en
.
http://iridia.ulb.ac.be/~marchal/
--
You received this message because you are subscribed to the Google
Groups "Everything List" group.
To post to this group, send email to everything-list@googlegroups.com.
To unsubscribe from this group, send email to everything-list+unsubscr...@googlegroups.com
.
For more options, visit this group at http://groups.google.com/group/everything-list?hl=en
.
--
All those moments will be lost in time, like tears in rain.
--
You received this message because you are subscribed to the Google
Groups "Everything List" group.
To post to this group, send email to everything-list@googlegroups.com.
To unsubscribe from this group, send email to everything-list+unsubscr...@googlegroups.com
.
For more options, visit this group at http://groups.google.com/group/everything-list?hl=en
.
http://iridia.ulb.ac.be/~marchal/
--
You received this message because you are subscribed to the Google Groups
"Everything List" group.
To post to this group, send email to everything-list@googlegroups.com.
To unsubscribe from this group, send email to
everything-list+unsubscr...@googlegroups.com.
For more options, visit this group at
http://groups.google.com/group/everything-list?hl=en.