X-Message-Number: 14167
Date: Mon, 24 Jul 2000 22:19:13 -0700
From: Mike Perry <>
Subject: Survival

Brook Norton, #14150, says
>Mike Perry, #14141, writes
>>>> As a rough approximation, I will say that my present self survives in a
>future self if that self has memories of having been the present self and
>feels a sense of identification with the present self. This is not an
>utterly meaningless idea. <<<
>response- The use of "survival" above certainly has some meaning.  But it
>still doesn't indicate to me that some essential common denominator survives
>into the future.  One thing I like about the present-moment-self hypothesis
>of survival is that it passes all the hypothetical problems I can subject it

But in part that is because it is so very limited. It would imply that I
will not even survive to the end of this sentence! I'll brave the paradoxes
to get something better than that! I think with diligent, though possibly
hard effort, the problems with at least some of the more interesting
theories of survival are resolvable.

> Whereas, other hypotheses of survival can all be challenged with
>counter-examples that leave them in a state of paradox.  For example, in
>your above description, one could imagine that a future historian (not your
>future self) finds great interest in you and gets to know your past very
>well and even comes to feel a camaraderie with your past.  Meanwhile, your
>future self has moved on to issues of the day and rarely if ever remembers
>his distant past self.  In fact those memories may have been wiped clean
>from his brain.  Therefore, the historian, not your future self, has the
>greater link to your past.  Have you become the historian?

Let's go a little further. The historian has just uploaded my memories and
really feels he experienced my past and he is me. While what you are
referring to as "your future self" (i.e. "me") has forgotten so much he
doesn't feel he was ever the person now typing this message. In this case, I
give the nod to the historian. He is a continuer of me, i.e. he is "me" (I
have become him), while the other fellow is not that at all. I have no
problem with this, except insofar as it makes my survival look rather
precarious. (Such a hypothetical historian, who would rescue "me" in this
way, might be a great rarity.)  I hope it won't be so. By the way, I also
have no problem with two or more continuers. Both could be "me"--the "me" of
today, grown older and hopefully wiser and better all around, and be
entirely separate individuals. If many-worlds is true, we expect this to be
happening all the time anyway, though in this case the different continuers
are in different universes and unable to communicate with one another.

>  I think this
>example shows the limits of using memory as a survival criterion.

I don't.

>  I find no
>paradox in saying "We are who we are now. We will become someone >different
in the future.  Our present and future selves may share some >commonality
such as memories...or they may not."

I find no paradox in the idea that I might be shot at sunrise tomorrow or
otherwise perish. But that doesn't make me happy with the idea or
uninterested in alternatives. 

Lee Corbin, #14153, writes

>Mike Perry, #14141 writes
>>As a rough approximation, I will say that my present self survives
>>in a future self if that self has memories of having been the
>>present self and feels a sense of identification with the present
>I almost entirely agree with Mike.  But suppose that I ingested
>midazolam half an hour ago, and therefore will remember nothing
>of what I am doing at the present moment.  I think it's a mistake
>for me to suppose that I'm about to perish, because objectively
>there will arise tomorrow a physical object as identical to me as
>are the physical objects that arise every morning in my house.  So
>we who hold to the information theory of identity, as Mike Perry
>has always called it, must consider ourselves other than just
>memory supersets.
Response: Note I said "rough approximation." I consider this issue at
greater length in my book, mainly chapter 15. In one sense, assuming you
base survival strictly on the memory superset idea, if even one bit (in the
computer sense of a 0 or 1) of your memory is ever lost, "you" die. That is
too restrictive (in fact forgetting can be a good thing, depending on what
it is you forget). Roughly, I handle this difficulty by reformulating the
problem somewhat. I think about the case of the would-be immortal, and ask
what would be a reasonable requirement for there to *be* a being who
survives and is sentient forever. Among the requirements: the being in
question must accumulate memories that are never forgotten, so it becomes a
successively larger memory superset. *But* not every memory that is taken in
need be unforgettable. Many can be lost, so long as a growing core of
memories that are not lost is maintanied. This is what I call convergence to
an ideal self. The requirements for this are far less strict than for the
strict memory superset idea, but are far more practical too. I am reasonably
confident that, assuming certain technical hurdles can be overcome, "I"
could adequately survive through such convergence, though indeed some
memories may be lost on the way. So this I believe will overcome the problem
you raise. There are other, interesting ways of dealing with other problems
such as "false" memories or fusion or fissioning of individuals.
>I know it's strange to think that any such close resembler (like my
>duplicate in the next room) is really and truly and totally me, but
>that's what our best theories of how the universe works are telling

I find myself in agreement here too. See "Interchangeability," ch. 7 of my book.

Mike Perry

Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=14167