🚀 go-pugleaf

"r norman" <rsn_@_comcast.net> wrote in message
news:vlc4rvsm5aodd5n77h59a07f3h68i7ofe7@4ax.com...
| On Wed, 12 Nov 2003 16:31:48 +1100, Matthew Kirkcaldie
| <Matthew.Kirkcaldie@removethis.newcastle.edu.au> wrote:
|
| >In article <8e53rvs4p64t9ba1njjfb2e3n7ufpnk6e2@4ax.com>,
| > r norman <rsn_@_comcast.net> wrote:
| >
| >> However, if it is true that consciousness is nothing more than
the
| >> working out of laws of physics and chemistry given a particular
albeit
| >> highly complex arrangement of molecules, organelles, and cells,
then
| >> once we understand the mechanisms it should be conceptually
possible
| >> to produce an artificial device that uses the same laws of
physics and
| >> chemistry with a rather different arrangement of non-biological
| >> components to acheive the same effect.
| >
| >So, to exaggerate that line of reasoning, it should be possible to
build
| >a hydrogen atom out of other non-elementary components and have it
| >behave the same way?  Or make a working atomic bomb out of
Meccano?
| >
| >I know these extrapolations are ridiculous but I make them to
point out
| >that the assertion that a process is created by the physical
properties
| >of its consituents does not imply that it could also be created by
| >different materials.  Right or wrong, I simply do not believe that
| >consciousness is seperable from the biological nature of the
brain.
| >
| >Most AI and connectionist approaches seem to me to be like piling
up
| >large heaps of components and hoping that they will become a 747
| >somehow.  Actually, more like piling up large heaps of sandwiches
and
| >hoping they will become a 747 somehow.
| >
|
| We are not really disagreeing, just having a lot of fun with
| disputations!
|
| You hit the nail on the head by your "assertion that a process is
| created by the physical properties oits consitutents does not
| imply..."  I originally included some form of computation or
| information processing as an essential part of consciousness.  In
my
| way of thinking, "computation" like "negative feedback regulation"
or
| "controlled release of energy" is a  "process" separate from its
| physical implementation.  Processing a broadcast electromagnetic
wave
| to produce a television picture with sound is the "same" whether
the
| machinery is vacuum tube or very large scale integrated circuitry.
I
| argue that "consciousness" is likely to be the same type of thing.
In
| computerese, it could be ported to a different platform.
|
| The failures of AI to date merely indicate the enormity of what we
| don't know.  At some point we are likely to figure it out.  I can't
| imagine that point coming even in the next few decades, though, if
| even in the next century!
|
| Where we strongly disagree is you believe that consciousness is
| inseparable from the biological nature of the brain.  If
consciousness
| lies in the physical implementation, then there must be something
| special about the structure of neurons.  But there isn't anything
| there that is not found (in one form or another) in virtually all
| cells!  Do all living cells have some form of nascent
consciousness?
| Then consciousness is an integral part of what we think of as
"life".
| I disagree.  It must be in the pattern of arrangement and
interaction
| of the nerve cells.  And patterns of interaction can be replicated
in
| different hardware.

[Sorry, I'm pushing 30+ hours without
sleeping consciousness, and am too tired
parse the lengthy prior discussion. Apol-
ogizing for the way my NG reader will
screw-up the justification.]

Cogent.

Replicate the energydynamics.

One thing that makes complete repication
of the energydynamics so hard is that the
Biology does all the infinitely-divisible
'Coulomb force' stuff so elegantly.

The other thing is that Consciousness
derives in TD E/I-minimization that occurs
within individually-unique experience -
so one cannot just replicate the energy-
dynamics. Given a system that is globally-
integrated via TD E/I-minimization, one still
has to replicate the energydynamics via
replication of experience [which is as close
to being Impossible as anything that I'm
aware of be-cause WDB2T has moved-on].

It's easier to just engineer a system that
does TD E/I-minimization [including 'drive'-
precedence-hierarchy prioritization [AoK,
\Ap5, Ap7], and not 'demand' that it replicate
a =particular= instance of biological Consci-
ousness.

Then just turn-it-on and turn-it-loose.

[And wait for it to learn about its inherent
TD E/I-minimization dynamics :-]

[Such learning must be optional. Hard-coding
it eliminates the possibility of Free Will.]

Cheers, Dr. Norman,

ken [k. p. collins]