[ExI] The Anticipation Dilemma (Personal Identity Paradox)

Lee Corbin lcorbin at rawbw.com
Thu Jul 19 03:55:36 UTC 2007

Stathis writes

> I've changed my view on survival through copies which are not up to
> date, mostly through discussions with you. I originally thought that
> it was obviously false, but now I see it as presenting a conundrum:
> (a) Death is bad because if I die, I can anticipate no further experiences.


> (b) However, partial memory loss through drugs like midazolam also
> results in a state of mind from which I can anticipate no further
> experiences, and that doesn't seem so bad.

Check.  (All very well put, it seems to me.)

> (c) Therefore *either* death is not so bad *or* there is something
> else about death, not present in the memory loss example, which makes
> it bad.


> (d) The difference between what we normally think of as death and the
> memory loss example is that in the latter someone with some of my past
> experiences will be able to anticipate future experiences, even though
> I-now won't.

Yes, that seems to be the difference.  So the entire question now
appears to revolve around whether the "someone with some of
my past experience" should be regarded as the same person as one.
And yes, the instance you call "I-now" will have, say, no further

> (e) Therefore, maybe death is not bad if someone with some of my past
> experiences is left behind.

Doctors who apply midazolam to their patients probably do not
look upon their intoxicated patients (as you sort of called them)
as though these people were going to perish. And to speak of
just plain old hypothetical memory erasure makes the arguments
crisper. A wife who loves her husband but realizes that her 
husband is to lose all of today's memories is probably far less
concerned than she would be if he were to break an arm.
She can't help but think that "he'll be just fine" tomorrow,
and that he'll be the same person as today. We could *suppose*
that she is correct. But all this is rife with circular argumentation,
I'm aware.

Here in (e) of course you're using a meaning of "death" that is
at odds with the one I use. But from your point of view, yes,
I guess I can't argue with your logic.

> (f) But to change my original view to the above seems hardly easier
> than deciding that death with no copies left behind is not bad.

Well, *everyone* is alarmed at the thought of the patient, or
the husband, or the subject himself  not being around any long.
I guess the whole question is still, "Should one be alarmed or
at all worried if some memory of the past few minutes or past
few hours is going to be erased."

> You see, there are several consistent positions possible, and which
> one I choose depends on psychological factors, not on science or
> logic.

I suppose so, since I can't fault your logic.  It still seems to me that
there is a kind of scientific, detached, analytical, third-person view
that strongly suggests that the ensemble of physical Lee Corbin's
who could awake in my bed tomorrow and still be me is very
large. A zillion things, from the gravitational attraction of passing
trucks to whether an old friend rings me up on the telephone
tonight all vastly change the physical state of the person who
awakes in my bed tomorrow.   But within a very large range,
we consider them all to be me.  But here I am saying nothing more,
I suppose, than that this detached physical viewpoint is the
"similarity" viewpoint. 

>> (And I know from his "Luckiest Man in the Universe" scenario, in which
>> Francis Bacon by sheer luck is brought back to life by a random
>> collision of molecules---and argued by Max to be just as legitimate
>> a copy of Bacon as was the original---is bolstered by quite a few
>> arguments in his thesis.  Anyway, while it may be less than a self-
>> inflicted brainwashing, it is more than just my opinion.
>> However, in the same article Ben Goertzel

[blast it, I misspelled his name last night]

>> defends what I call the
>> "path conception of identity", in which the final *state* is not
>> what is important, and is not evaluation, but the *path* by which
>> the state is reached.  A good friend of mine endorses this path-
>> view of identity.  Perhaps you'd find it appealing too.
> I think Bacon would still be the Bacon he was at the time of which the
> copy is a representation, but not the Bacon he became after that (that
> is, even if the copy was made aeons after all original Bacons were
> gone, and regardless of how the copy was made).

I would agree if he changed greatly later on in life. For example,
if he became quite old and senile, then I'd go along with him not
being the same person.  

> That's the straightforward case, and I'd go as far as saying
> that I don't see how it could logically be otherwise, without
> allowing that Bacon might have been a different person from
> day to day even before his death in 1626.

As you know, the logic I endorse is that of something gradually
changing from one fuzzy category into another. A mountain is
different from a molehill, but the latter can be slowly changed
into the former by infinitesimal degrees.

> (Well, he was strictly speaking changing from moment to
> moment, but he *felt* he was the same person, and that's
> what matters in this context.)

But then there is the case of the crackpot who believes he
is Napoleon.   But if you mean by "*felt*" that he really
did have the same thoughts, memories, and emotions that
Napoleon had, then naturally it would be isomorphic to 
the Francis Bacon case.

> Could you explain the path identity concept further, or give me a
> reference I can look up?

I will post something on it later.


More information about the extropy-chat mailing list