On 03 Mar 2009, at 13:40, Stathis Papaioannou wrote:
>
> 2009/3/3 Bruno Marchal <marchal.domain.name.hidden>:
>
>> I think that comp practitioners will divide, in the long run, along
>> three classes:
>>
>> A: majority. Accept teleportation but disallow overlap of
>> "individuals": annihilation first, reconstitution after. No right to
>> self-infliction. In case of accidental or exceptional self-
>> multiplication, consent is asked at any time.
>> B: a stable minority (in the long run). Accept teleportation but do
>> allow overlap of individuals. Some will fight for the right of self-
>> infliction including the consent made before the duplication, but
>> with
>> precise protocol. You know the problem of the masochist: I say no,
>> continue, I say "no no", stop!
>> C: the bandits. They violates protocols and don't ask for consents.
>> They should normally be wanted, I mean researched by all the polices
>> of the universe, or already be in jail or in asylum.
>
> I think B might work, since it is more or less like the present
> situation, where our decisions are based on a rough risk-benefit
> analysis, i.e. we decide on a course of action if as a result
> gain*Pr(gain) >= loss*Pr(loss). So we decide to smoke, for example, if
> we judge the pleasure of smoking (or the suffering caused by trying to
> give it up) to outweigh the suffering that may result from
> smoking-related illnesses. However, there are also differences if the
> copies are allowed to overlap. If I make a decision that has an
> adverse effect on my future self I may regret the decision, but it's
> not possible to ask my past self to reverse it. On the other hand, if
> I agree for one of my copies to torture the other it is always
> possible for the victim to ask the torturer to release him. Also, it
> is possible for the torturer to come to believe that he is never at
> risk himself after repeated duplications: I've done this many times
> and it's always the *other* guy who suffers, not me, so there is no
> reason for me not to repeat the process. This would be so even if the
> agreement was for 100 copies to be made and 99 of them enslaved: the
> one who does the enslaving may come to believe that he is never at
> risk, and continue creating copies 100 at a time.
You can then imagine the surprise of the copy or copies: "- I did
this often and thought there are no risk, but here I am enslaved, and
I will suffer and die".
That is why the B people made a law, for helping those who
misunderstand the probability. If you decide (before duplication) to
kill the copy, the choice of victim/torturer is still decided through
a throw of a fair coin. This makes the decision unbiased by fake
protocols based on a bad understanding of what the comp probabilities
are.
Iterating the procedure, with the throwing of the coin, could make you
believe you are incredibly lucky, but the computationalist should know
better: this is just the "usual" comp-suicide self-selection (assuming
of course we can really kill the copies, which is in itself not an
obvious proposition).
Bruno
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.domain.name.hidden
To unsubscribe from this group, send email to everything-list+unsubscribe.domain.name.hidden
For more options, visit this group at
http://groups.google.com/group/everything-list?hl=en
-~----------~----~----~----~------~----~------~--~---
Received on Tue Mar 03 2009 - 12:56:39 PST