Re: Memory Merging Possible For Close Duplicates

From: Lee Corbin (
Date: Fri Mar 14 2008 - 00:15:42 MDT

Mike writes

> I find it difficult to phrase my thoughts in words because I tend to think more in
> pictures and analogy.

Me too!

> To give a more concrete (real & not rhetorical) example:
> Today on my way home I took a slightly different path than usual, yet
> after some meandering I found myself on a familiar road and
> continued from that point along my usual route. Suppose I make
> similar (yet each unique) small detours each day for a week. After
> the experiment you ask, "How did you get from work to home?"
> but you do not provide the specific context of a particular day, or a
> particular travel method. When I think of all the routes I had used,
> they are a superposition of valid solutions- and they're all
> correct solutions within the scope of the question.


> If you ask the same question to both me and my wife, then I might
> think also of those routes she used to travel from her work to
> our home. I did not subjectively experience those routes in real-time
> detail, but I am still capable of imagining them as solutions to the
> intentionally open-scoped question. I don't think these 'imagined'
> routes are difficult to include within the sphere of possibility for how
> I did (or might have) get from work to home. There might have been
> a route enter my mind's view of possible routes that I had taken months
> before I started this experiment. I believe that experience may or may
> not have happened (a third-party might have told me how to imagine it)
> but it is still worth some possibility of correctly answering the question.


> The analogy works here because the question itself does [is] not limited
> by time - so each of those serialized events are presented simultaneously
> as possible answers. I believe subjective experience of run-time in a
> computation environment where the observation of time can be
> manipulated by an external agent provides for a context switch,
> several processing threads each deposit results into shared "memory"

Ah. Here you mean not only the computer science "shared memory"
but real human type shared memory.

> and a context switch back will appear to those inhabiting the suspended
> environment that the results of those independent threads have been
> computed instantly.

That's not very clear, IMO. With ordinary raw threads or processes
running on a computer, sure, one moment the process has access to
data structures X and Y, and the next, equal access to Z. But that
entirely ignores the knotty problem of how memories are added to
people, as you say "instantly".

Normally each new experience you have is immediately compared
on some sort of salience measure to everything else that has ever
happened to you, i.e., to all your other memories. That's why you
are "reminded" of things, some of which happened a long time ago.
Now if you get enough new experiences, the new memories that
are generated are slowly integrated into all your existing ones.
The computer analogy seems a little strained here, at least with
the kinds of algorithms we have today running on our machines.

> I have some proficiency with SQL as a visualization for
> structured information / thoughts. Recent discussion of
> GLUT makes me think of database concepts. Given that
> database theory is essentially a scaled-down version of set
> theory, would it be acceptable to represent (for discussion)
> these ideas in set theory terms?

Well, I suppose so. At least keep the provisos I suggested
above in mind (thanks) and be very skeptical of every step
you take, of course, to minimize the damage done to the
original questions. But sure. After seeing Stuart Armstrong
successfully get away with modeling states of a TM (or
generations of a Life Board) with mere polynomials, and
state transitions as differentiation, I'm pretty willing to listen
to anything!

Well, almost anything :-)


> (I admit to not having the rigid formalism of Set Theory available
> as a natural language, but might it be worth the effort? Is that too
> domain-specific to have anyone make the effort?)

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