Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10 5/3/83; site sequent.UUCP
Path: utzoo!watmath!clyde!burl!ulysses!harpo!seismo!hao!hplabs!tektronix!ogcvax!sequent!merlyn
From: merlyn@sequent.UUCP
Newsgroups: net.ai
Subject: Re: A topic for discussion, phil/ai persons.
Message-ID: <503@sequent.UUCP>
Date: Wed, 16-May-84 11:21:24 EDT
Article-I.D.: sequent.503
Posted: Wed May 16 11:21:24 1984
Date-Received: Mon, 21-May-84 05:26:41 EDT
References: <277@wxlvax.UUCP>
Organization: Sequent Computer Systems, Portland
Lines: 72

> From: rlw@wxlvax.UUCP
> Message-ID: <277@wxlvax.UUCP>
> Date: Tue, 15-May-84 23:33:31 PDT
> 
> It seems that it is IMPOSSIBLE to ever build a computer that can truly
> perceive as a human being does, unless we radically change our ideas 
> about how perception is carried out.
> 
> The reason for this is that we humans have very little difficulty
> identifying objects as the same across time, even when all the features of
> that object change (including temporal and spatial ones).  Computers,
> on the other hand, are being built to identify objects by feature-sets.  But
> no set of features is ever enough to assure cross-time identification of
> objects.  

I dunno about that.  Think of the many times you have heard "is that
the same Mary Jones I went to high school with?  She's not the same
person at all!"

Now, obviously, Mary Jones is the same person (or the whole statement
is false).  Yet, by some method of *distinction* (key word!), the
person is distinguishing Mary Jones @ now from Mary Jones @ high
school.

Distinctions seem to be the key.  I suggest that what we are doing as
human beings in PERCEIVING is merely drawing a line of distinction
between things that are in the set, and things that are out of the
set.  For example, we distinguish UP from DOWN, LEFT from RIGHT, WRONG
from RIGHT (bad pun!).

So, in the case of Mary Jones, *enough* (another keyword) similar
distinctions exists for Mary Jones, even invariant of time and space,
that she gets to retain her identity (for most people).  For the person
uttering the statement above, however, *not enough* of the distinctions
are consistent, hence the statement of her not being who she was.

I think (this is guessing now, but based on observation [ARRGH... more
perceptions!]) that the act of perception is based on triggering ENOUGH
DISTINCTIONS to warrant a match with past data.  Data, by the way, is
all stored as distinctions as well.  So if eyes, face, name, and
approximate age match for Mary Jones, the inference is that all other
distinctions made for her (such as favorite music), as well as for
high-school people, girls, and so on, all fit identically.

My evidence for this MODEL (also a keyword) [that all we are are the
distinguishers of sensory input] is that I work with people to produce
breakthrough (beyond predictable and explainable) results.  I operate
with them like life is merely an interpretation, a collection of
distinctions.  People make distinctions unconciously that are counter
to their goals.  I simply have them remove distinctions that are
barriers.  Poof.  (That's a side topic by itself... I just said it to
describe the source of my evidence.)

MODEL is a keyword to demonstrate that this, AGAIN, is merely a
distinction that *I* am drawing.  As far as I can tell, everything I can
possibly recognize is based on my own interpretations and labling and
distinctions I drew in the past, but that's also a model.  It could be
just as likely that a little bird whispers in my ear to tell me that
Mary Jones is standing out there, and that her favorite music is Rock,
but that model makes it much harder to produce breakthrough results.
I"ll stick with one that works to produce my own intentions.

Sorry for the length... this area is one of my favorite topics.
I'll answer mail, but I'd prefer news for discussions of interest
to more than just you and me.

-- A particularly personal and original observation from the thought-stream of
Randal L. ("distinguished observer") Schwartz, esq. (merlyn@sequent.UUCP)
	(Official Legendary Sorcerer of the 1984 Summer Olympics)
Sequent Computer Systems, Inc. (503)626-5700 (sequent = 1/quosine)
UUCP: {decwrl,ogcvax,pur-ee,rocks34,shell,unisoft,vax135,verdix}!sequent!merlyn
Original Material (C) 1984 by Randal L. Schwartz [ALL RIGHTS RESERVED]