Lee Corbin writes:
[quoting Stathis]
> > >I think that if it is given that either you or your duplicate
> > >must die, then you should willingly sacrifice yourself if it will
> > >enrich your duplicate.
> > >
> > >Either way, I think you wake up the next morning very satisfied
> > >with the outcome.
> >
> > How do you wake up the next morning if you're the one who died? Unless
>you
> > can effect some sort of mind merge just before dying, you lose all the
> > experiences that you have had since you and your duplicate diverged...
>
>Well, Stathis, for heaven's sake! You've already admitted that a
>little memory loss does not threaten your identity! Recall the Aussies
>you wrote about who customarily lose an entire evening's inebriation :-)
Yes, and I also admitted that there is an inconsistency in my position.
Having my duplicate who has already diverged live on while I die is not just
memory loss, but rather replacement of the lost memories with someone
else's, which I feel is a greater threat to my identity and which I would be
less likely to agree to. Memory loss would be more like having myself backed
up and the backup run after I have died. If the backups are frequent, I
suppose it is better than no backup at all, but I would still feel afraid of
dying. At its most basic, for me anyway, the fear of imminent death is the
fear that the person I am *now* will be wiped from the universe and never
have any more experiences. The same consideration ought to apply to memory
loss, but people don't generally think of it that way, because they know
that they'll be OK afterwards, on the basis of past experience.
>So remember: your duplicate in the next room is *exactly* in the
>same state you are in right now if you lose a little recent memory,
>and then have some new experiences that are identical to his over
>the last few minutes. So he *is* you! That's what I mean when I
>say that we must regard duplicates as selves.
>And let's go back to this crazy "transfer" that seems (from my
>viewpoint) to occupy the attention of those who believe in
>"continuers". So you expect to be the person who arrives at
>the other end if you are disintegrated here and teleported
>there. You even expect to be him if the original here is
>not disintegrated. (Here I must lash out at the bizarre
>probability calculus that ensues for many at this point:
>whether or not they *are* the remote version seems to depend
>on what happens locally here. Sheese.)
There is a transfer of information in order to effect the teleportation, but
this is just a technical detail; there is no actual "transfer" of identity,
if that is what you meant. There are only people's beliefs and memories
concerning who they are and who they were, coupled with the fact that human
minds can only experience being one person at a time. As for how what
happens here can affect whether or not the person entering the teleporter
finds himself over there: if you destructively teleport an apple there,
there is a 100% chance that the apple is over there; whereas if you
non-destructively teleport an apple there, choosing an apple at random
because you are only allowed one apple at a time, there is a 50% chance it
will be the one here and a 50% chance it will be the one there. Where's the
problem?
>So if all the 1000 Stathis's in the various rooms are to die
>but one, then that one "continues" all the others. Now we
>play a trick. The 1000 don't actually die but are placed in
>instantaneous suspended animation.
>
>Oops! The big Bean-Counter Upstairs who keeps track of where
>the serial numbers go is confused! He had better have all
>1000 Stathis's continue in the one, just in case something
>goes wrong with the suspended animation machinery. But then...
>what to do when nothing goes wrong? How to send all the souls
>back into the original bodies??? (Of course, here in this paragraph
>I am just attempting to ridicule a point of view in which I do not
>believe. The truth is, of course, is that no "transfers" take place,
>and the whole idea of a "continuer" is wrong.)
>
>If you believe that the 1000 will continue in the one (what, I
>wonder, with probability 1000?), then they'll "continue" in the
>one whether or not they're disintegrated.
Sorry, I don't understand what you're saying here. Do you mean that 1000
versions of me are running in parallel and all but one are stopped or
suspended? It's obvious to me that however often the number running is
changed, I won't be able to tell that there is any difference. This wouldn't
work if they had serial numbers because if each version knew his number,
they would start to diverge, and stopping one of them would then lead to the
loss of unique experience, as discussed above (you might say it doesn't
matter if it's something as trivial as a serial number, like losing a second
of memory, but the point still stands). It wouldn't work if they had souls
either, because killing some of them, even if they remained running in
parallel, would send the souls to heaven or hell or wherever souls go, so
again there would be a difference. It only works if they are running in
parallel, no serial numbers and no souls. It seems either you have
misunderstood my point or I have misunderstood your point.
>And what about this, as a further attack on the probability idea:
>We start with 1000 Stathis's in 1000 hotel rooms, all happily
>looking forward to the afternoon. Then during 1 millisecond all
>but one are killed off, and in the next millisecond the random
>one who survived is copied into the remaining 999 hotel rooms.
>I suppose---if you believe in that silly probability thing---
>that you figure your odds (as a particular person, say in room
>506) as being 1 in 1000.
>
>:-) Then let whole scenario repeat every 50th of a second,
>and pretty soon your odds of surviving are less than one in
>a billion???
I am certain to survive (from my point of view) with the experiment as
written, because the copies have no time to diverge. The problem arises when
each of the 1000 copies has time to think, "Oh no! I'm going to die!" Let's
say the cycle repeats every minute rather than in milliseconds. After
several cycles, I am almost certain to experience this feeling of impending
doom and then die. However, one version of me will always survive, and at
the end of the experiment this version will recall being afraid he was going
to die but not, of course, actually dying. As discussed above, this version
could be considered the continuer of any of the versions who died, but with
memory loss/ replacement. This example reminds me of the thought experiment
you proposed in an earlier post, where your bank balance increases by $1000
every day in return for allowing yourself to be tortured every night, and
then having your memory of the experience wiped. If you have gone through
several cycles of this you might stop worrying about it, but at night,
between the screams, you will be cursing yourself for agreeing to such a
stupid and horrible deal. In the current experiment, fear of death takes the
place of physical torture.
>No: we have to accept the simple solution I suggest: you are
>your duplicates, near and far, future and past, whether or not
>your instance is collecting memories of their own particular
>locales. A few minutes', or even a few months', worth of
>memories doesn't really matter much so far as SURVIVAL is
>concerned. This is simple, and it seems to me, obviously true.
I concede that the memory loss/ death equivalence issue gives rise to an
inconsistency on my part. I could fix that by either saying that any degree
of memory loss is as bad as dying, or else agree that a backup copy allows
you to survive death if it isn't too old. The latter solution seems more
practical, but is less elegant as it requires an arbitrary definition of
what "too old" means. I don't think it would do if a version of yourself as
an infant was the only backup copy available to you, for example.
--Stathis Papaioannou
_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today - it's FREE!
http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/
Received on Sat Jul 02 2005 - 12:46:14 PDT