From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.ecf!utgpu!cs.utexas.edu!qt.cs.utexas.edu!zaphod.mps.ohio-state.edu!wupost!uunet!orca!javelin.sim.es.com!biesel Wed Feb  5 11:56:34 EST 1992
Article 3436 of comp.ai.philosophy:
Newsgroups: comp.ai.philosophy
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.ecf!utgpu!cs.utexas.edu!qt.cs.utexas.edu!zaphod.mps.ohio-state.edu!wupost!uunet!orca!javelin.sim.es.com!biesel
>From: biesel@javelin.sim.es.com (Heiner Biesel)
Subject: Re: Strong AI and Panpsychism
Message-ID: <1992Feb3.184156.4027@javelin.sim.es.com>
Keywords: panpsychism
Organization: Evans & Sutherland Computer Corporation
References: <1992Jan28.164410.9509@psych.toronto.edu> <21879@life.ai.mit.edu> <1992Jan31.163659.11670@javelin.sim.es.com> <1992Jan31.231006.7248@news.media.mit.edu>
Date: Mon, 3 Feb 1992 18:41:56 GMT

minsky@media.mit.edu (Marvin Minsky) writes:

>In article <1992Jan31.163659.11670@javelin.sim.es.com> biesel@javelin.sim.es.com (Heiner Biesel) writes:
>>minsky@transit.ai.mit.edu (Marvin Minsky) writes:
>>>[stuff deleted]   
>      But "reflective" short term memories -- records of recent mental
>>>states that can be used as uinputs to other processes -- have many
>>>other uses, and (surely) many different mechanisms with different
>>>evolutionary histories and functions.  So as far as I'm concerned, it
>>>is the use of this word, as though it represents anything important,
>>>e.g., some irreducible attribute of mind -- that has kept philosophy,
>>>since the time of Kant, from contributing important insights to
>>>psychology.
>>
>>True. However, "consciousness" is, for most human beings at least, the 
>>defining property of their own existence, and hence irremediably tied up with
>>their notion of mind. One can conceive of a "mind" capable of any of the
>>activities we are usually so proud of: logical analysis, poetry, whimsy,
>>dirty jokes, campaign rhetoric..., all without being "conscious". I, and
>>presumably Minsky, would consider an artificial construct capable of these
>>feats a "mind". Is it an interesting "mind", simply because it passes the
>>Turing test? Does it contain the experience of "I", rather than simply the 
>>concept?
>> [religious hypothesis deleted]

>Ah, that's my point.  Perhaps *you* can conceive of a "mind" capable
>of any of the activities we are usually so proud of --- without being
>"conscious"  -- but I can't or, at least, I've trained myself not to.
>This is because I'm developing that hypothesis mentioned above, that
>the phenomena we call consciousness are caused by parts of the mind
>accessing partial records (that is, condensed traces) of the recent states
>of other parts.  Now, *I* can't conceive of a mechanism being able to
>do the sorts of things you mentioned -- logical analysis, poetry,
>whimsy, -- all without (where you said "being "conscious") being able
>to use recent records of its activity.  Otherwise it would only replay
>stored scripts, or get into horrible loops, or run in a completely
>"situation-driven" manner.  

Why do you make a distinction between recent records/memories and
(presumably) older ones? How does the age of a memory bear upon its
function in defining consciousness?

>The difference between our views is that you seem to be thinking of
>consciousness as a functionless adornment, whereas I'm saying that it
>comes from a complex of mechanisms that seem useful and, er, causal.

Actually, I'm quite fond of consciousness, both personally, and as an
effective means of creating novelty. However, I don't see that is
a necessary component of mind. The subjective experience of selfhood,
and the consequent introspective analysis we are all so fond of, seems
a trivial aspect of "mind" in the abstract. If one contends that the
internal monologues that we commonly take as being typical of consciousness
are somehow necessary for realizing the higher functions of "mind", one
could postulate some mechanism which does, as suggested, replay recent scripts
and fragments in various partly-correlated-partly-random fashion, albeit
without any self-awareness.

As I recall, D. Hofstadter made the case that consciousness arose somehow
spontaneously from self-referential systems of sufficient complexity. The
attempt to invoke self-awareness in a complex of mechanisms arising from
evolutionary accidents and adaptations seems to suffer from the same
difficulties that Hofstadter experienced: how does the self arise, such
that it can be conscious of itself? Hofstadter waved his hands, mumbled
the magic incantation "strange loop", and Presto! consciousness came
into being.

This time I'd like to watch real close while someone does it in slow motion.

>Returning to the "panpsychism" thread, this could be applied to
>appropriate degrees to any mechanism (or part of any larger mechanism)
>to the extent that it makes (useful) applications of memories of its
>earlier states.  This, of course, might mean that the use of the
>current term "conscious" would not be of much value until we attach to
>it a functional grammar of more specific terminologies.  Otherwise you
>get into silly questions like "are viruses alive" which come from not
>having a more sophisticated descriptive language for the variously
>possible sorts of "adaptive? metabolizing? reproductive"
>self-preserving? mechanisms."

Absolutely. However, the problem that usually arises when one tries to nail
down one of these marvelously fuzzy terms to a few useful defining properties
is that most people will complain that you've taken away all the interesting
parts, leaving only a dead skeleton. 

I suppose something analogous must have happened when scientists first
began defining the physical concept of mass. Presumably, the term had
(and still has) numerous associations, all of which had to be stripped away
before the formula F=ma could make sense. In time, the physicists mass has
taken on the quality of a simple, and its meaning in physics relates only
to the mathematical descriptions which express its relations to other
physical simples. Perhaps we are in the early stages of a similar process
with respect to mental phenomena.

Regards,
       Heiner biesel@thrall.sim.es.com


