Re: Memory Merging Possible For Close Duplicates

From: Lee Corbin (
Date: Tue Mar 11 2008 - 22:42:29 MDT

Jeff writes

>> > I don't think anything in this thread is all that paradoxical.
>> Right. As regards *anticipation*, not yet. I should dig up my old
>> essay "The Anticipation Paradox" and post it. The bottom line
>> is that there are TEs in which you have to look forward (or
>> "anticipate") something that already happened to you.(!)
> Anticipating something that already happened to you is definitely not
> possible with my definition of anticipate.

Exactly! It has to be impossible. So I gave up on "anticipation",
and now regard it as this somewhat irrational module nature had
to endow us with.

> I think your problem may be that you're using a poor definition of
> anticipate. The purpose of anticipation is to prepare a current brain
> state to be able to respond appropriately for some future change in
> that brain state (when it is exposed to a new event and hence
> modified to become a new brain state).

Right. That's why in 1986 I was so disturbed by a certain TE that
I had concocted that I'd just pace the halls at work, unable to focus
on my job. I published a very condensed version of it in 1994 in
"The Venturist", a Mike Perry cryonics-related publication. I once
tried to write it up in a more user-friendly story-telling style, but
it's just too complicated. But NOT too complicated for the people
here. I'm sorely tempted to give it another shot. I don't have to
explain determinism, for example.

>> Note how totally and completely different (and impossible so far
>> as I know) for memory merging to plausibly take place between
>> two actually different people.
> I don't understand what you mean by "actually different", since in the
> example you gave, you're merging two "actually different" people.

In my example, the two duplicates (well, one original and one copy)
were extremely similar. I call these "close duplicates". You are a
close duplicate of the person you were yesterday. It would be
theoretically possible to take a copy of the you that you were
yesterday, and just "add" some memories (what happened to
you today) and subtract others (what you have forgotten), and
this is in fact what nature easily and routinely accomplishes.

In my example, the copy---who had gone to a movie and the
original---who had stopped at a bookstore meanwhile, could
relatively easily have their memories merged because, forgive
the language, "they are the same person" (with only tiny
differences). We just take the original, who naturally remembers
going to the bookstore, and "add" the memory of getting off
the freeway early, going to the movie and watching it, and getting
back on the freeway. (I will say that he is then too tired or it is
too late to satisfy an urge he might also have to visit the bookstore.)

> There's no such thing as two identical people, unless you're talking
> about right at the single moment you copy someone. One millisecond
> afterwards they are already two different people, with a different set
> of memories and behavior.

I think that that is quite wrong. You are the same person from minute
to minute, even though the changes to you are vast, vast, vast by
comparison to the "change" one millisecond later that takes place
with copying (e.g. teleporting).

> In your example, they would be very different by the time they
> were merged again and probably completely unmergable.

Well, why? We can easily imagine the outcome. So you go on:

> Memory in humans is not stored at a single memory address
> like it is in a desktop computer. When you're exposed to new
> experiences, it recongifgures whole clusters of neurons and the
> structure of your brain and the interconnections between lots of
> different things change.

I didn't say it would be easy! I said that it is theoretically possible.
We "just" adjust all those neurons. Compared to the bulk of the
neurons that constitute your memories---who you are---they're

> It's highly sensitive to the order in which you experience things,
> as new events will change the way you understand past events,

That's why the TE I presented about the copy going to the movie
had to be carefully done, so that later it wouldn't seem strange to
him that he totally forgot about the movie while he was at the
bookstore---or, since it's weeks later, whichever came first.
This depends on the fact that surely two separate, memorable
things happened to you last month, but you don't recall which
happened first.

> While I have no proof that it's impossible, I have a very hard time
> imagining any plausible way you could accomplish such a merge.
> Maybe you could do a very "dirty merge", but there would be a lot of
> ambiguity and choices to be made about all of the conflicting neural
> modifications that are present in both of the copies... it seems
> guaranteed to cause massive psychological problems, even if the person
> was similar enough that they had only been split for an hour or less.

Would it be easier to imagine if they were already uploaded? Would it
be easier to imagine if we're already running in a sim?

Anyway, I claim it isn't even *imaginable* how two truly separate
people (on the usual meaning of words), e.g. William Shakespeare
and Albert Einstein, could be merged, without, as you say, creating
"massive psychological problems", or a completely confused brand
new person.


This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:01:02 MDT