I respect the amount of thought that when into Richard's paper
"Consciousness in Human and Machine: A Theory and Some Falsifiable
Predictions" --- but I do not think it provides a good explanation of
consciousness.
It seems to spend more time explaining the limitations on what we can
know about consciousness than explaining consciousness, itself. What little
the paper says about consciousness can be summed up roughly as follows: that
consciousness is created by a system that can analyze and seek explanations
from some, presumably experientially-learned, knowledgebase, based on
associations between nodes in that knowledgebase, and that it can determine
when it cannot describe a given node further, in terms of relations to other
nodes, but nevertheless senses the given node is real (such as the way it is
difficult for a human to explain what it is like to sense the color red).
First I disagree with the paper's allegation that "analysis" of
conscious phenomena necessarily "bottom" out more than analyses of many
other aspects of reality. Second, I disagree that conscious phenomena are
beyond any scientific explanation.
With regard to the first, I feel our minds contain substantial
memories of various conscious states, and thus there is actually substantial
experiential grounding of many aspects of consciousness recorded in our
brains. This is particularly true for the consciousness of emotional states
(for example, brain scans on very young infants indicate a high percent of
their mental activity is in emotional centers of the brain). I developed
many of my concepts of how to design an AGI based on reading brain science
and performing introspection into my own conscious and subconscious thought
processes, and I found it quite easy to draw many generalities from the
behavior of my own conscious mind. Since I view the subconscious to be at
the same time both a staging area for, and a reactive audience for,
conscious thoughts, I think one has to view the subconscious and
consciouness as part of a functioning whole.
When I think of the color red, I don't bottom out. Instead I have
many associations with my experiences of redness that provide it with deep
grounding. As with the description of any other concept, it is hard to
explain how I experience red to others, other than through experiences we
share relating to that concept. This would include things we see in common
to be red, or perhaps common emotional experiences to seeing the red of
blood that has been spilled in violence, or the way the sensation of red
seems to fill a 2 dimensional portion of an image that we perceive as a two
dimensional distribution of differently colored areas. But I can
communicate within my own mind across time what it is like to sense red,
such as in dreams when my eyes are closed. Yes, the experience of sensing
red does not decompose into parts, such as the way the sensed image of a
human body can be de-composed into the seeing of subordinate parts, but that
does not necessarily mean that my sensing of something that is a certain
color of red, is somehow more mysterious than my sensing of seeing a human
body.
With regard to the second notion, that conscious phenomena are not
subject to scientific explanation, there is extensive evidence to the
contrary. The prescient psychological writings of William James, and Dr.
Alexander Luria's famous studies of the effects of variously located bullet
wounds on the minds of Russian soldiers after World War II, both illustrate
that human consciousness can be scientifically studied. The effects of
various drugs on consciousness have been scientifically studied. Multiple
experiments have shown that the presence or absence of synchrony between
neural firings in various parts of the brain have been strongly correlated
with human subjects reporting the presence or absence, respectively, of
conscious experience of various thoughts or sensory inputs. Multiple
studies have shown that electrode stimulation to different parts of the
brain tend to make the human consciousness aware of different thoughts. Our
own personal experiences with our own individual consciousnesses, the
current scientific levels of knowledge about commonly reported conscious
experiences, and increasingly more sophisticated ways to correlate
objectively observable brain states with various reports of human conscious
experience, all indicate that consciousness already is subject to scientific
explanation. In the future, particularly with the advent of much more
sophisticated brain scanning tools, and with the development of AGI,
consciousness will be much more subject to scientific explanation.
Does this mean we will ever be able to ultimately explain what it
means to be conscious? The answer is probably no more than we will ever be
able to fully explain many of the other big existential questions of
science, such as what is time and space and existence. Just as we humans
have developed from the grounding of experience common sense notions of
time, space, and existence, we also have common sense notions of
consciousness, and various of its states and behaviors. The only
difference, is that until recently the tools necessary to objectively
measure consciousness have been much more primitive than our tools for
measuring many other aspects of physical reality. But that is starting to
change rapidly. If people like Kurzweil are right,
we will soon be able to measure brain states with amazing accuracy, and,
thus, we will soon be able to measure consciousness more completely than
many aspects of physical reality.
So what can we currently say or guess about consciousness, based on
introspection, brain science, and AGI.
First, just as there is no aspect of physical reality that can be
described that is anything other than representation and computation, there
is no aspect of consciousness that is anything other than representation and
computation.
Second, It follows from the first point that it should be possible to
create consciousness from a computer, but it is not clear exactly what type
and scale of computer would be required.
Third, there may well be different degrees of consciousness. Arguably
all computation, and thus all physical reality is conscious, but perhaps the
particular type of computations we humans describe as consciousness is an
extremely complex computation that has multiple characteristics that appear
to distinguish it from most of the computation that takes place in physical
reality. For example, it is a computation that can have many millions,
billions, or arguably trillions, of rapidly changing states, in which
various nodes in that state space can respond with relative crispness to the
states of a large number of other states, including the history of its own
state, and that of other nodes, over various time scales, and in which
computational focus can be rapidly switched pursuant to competition between
competing assemblies of activated states. Experiments on the correlation of
neural synchrony and conscious experience, indicate that conscious awareness
of a thought or sensation involves fairly large spread coordinated behavior
in the brain, which probably results in a corresponding flood of activations
related to a conscious concept that sufficiently ground that concept to give
the brain an awareness of its meaning.
I could go on listing what I believe to be the probable computational
aspects of a human consciousness, but I think those in the list that
understand some of the possible correlations between the operation of a
large (i.e, human level) Novamente-like AGI and the operation of their own
consciousness --- as derived from a study of their own subjective experience
--- already understand much of what additional things I would say.
An AGI billions of times less powerful and complex than the self aware
computation supported by the human brain could meet the definition of
consciousness used in Richard's paper. But it is doubtful that such a
miniscule computation would have much meaningful similarity to the rich,
full sense of consciousness in the human mind, and, thus, I think Richard's
paper sheds little light on the miracle that is a human consciousness.
So although I appreciate the serious, careful, respectful tone of
Richard's paper, I disagree strongly with about two thirds of its basic
conclusions.
Ed Porter
-----Original Message-----
From: Richard Loosemore [mailto:[EMAIL PROTECTED]
Sent: Friday, November 14, 2008 12:28 PM
To: agi@v2.listbox.com
Subject: [agi] A paper that actually does solve the problem of consciousness
I completed the first draft of a technical paper on consciousness the
other day. It is intended for the AGI-09 conference, and it can be
found at:
http://susaro.com/wp-content/uploads/2008/11/draft_consciousness_rpwl.pdf
The title is "Consciousness in Human and Machine: A Theory and Some
Falsifiable Predictions", and it does solve the problem, believe it or not.
But I have no illusions: it will be misunderstood, at the very least.
I expect there will be plenty of people who argue that it does not solve
the problem, but I don't really care, because I think history will
eventually show that this is indeed the right answer. It gives a
satisfying answer to all the outstanding questions and it feels right.
Oh, and it does make some testable predictions. Alas, we do not yet
have the technology to perform the tests yet, but the predictions are on
the table, anyhow.
In a longer version I would go into a lot more detail, introducing the
background material at more length, analyzing the other proposals that
have been made and fleshing out the technical aspects along several
dimensions. But the size limit for the conference was 6 pages, so that
was all I could cram in.
Richard Loosemore
-------------------------------------------
agi
Archives: https://www.listbox.com/member/archive/303/=now
RSS Feed: https://www.listbox.com/member/archive/rss/303/
Modify Your Subscription:
https://www.listbox.com/member/?&;
Powered by Listbox: http://www.listbox.com
-------------------------------------------
agi
Archives: https://www.listbox.com/member/archive/303/=now
RSS Feed: https://www.listbox.com/member/archive/rss/303/
Modify Your Subscription:
https://www.listbox.com/member/?member_id=8660244&id_secret=120640061-aded06
Powered by Listbox: http://www.listbox.com
