To: Steven Ericsson-Zenith, Ph.D.
Dear Steven,
After reviewing the information you sent, materials at your
website, and your correspondence below with Michael
Brunnbauer, it is clear you would consider any discussion of
biophysical consciousness (or of biophysical sensation,
perception, or experience) based on Turing computation to be
inadequate.
You are developing a continuous mathematical formalism to
explain biophysical sensations and responses, and believe the
biophysics cannot be Turing computable, at least with adequate
efficiency to achieve biophysical power constraints. (Viz.
SEZ-2013,p.7, and remarks at end of SEZ-2013 lecture.)
To contrast, my thesis discusses an approach toward
human-level artificial intelligence which includes artificial
consciousness and takes Turing computation as a starting
point. Yet the TalaMind approach does not preclude use of
continuous computation transcending Turing machines, per
section 4.1.2.4. While I think much can be accomplished with
Turing computation, I'm open to the possibility that quantum
computation and/or continuous computation may be needed to
achieve human-level AI.
Only a first chapter of your future work entitled 'On the
Origin of Experience' has been released to the public. When
your work is more fully published, questions and criticisms
may be expected from the biophysics and theoretical physics
communities, in their forums for discussion. However, it seems
clear that discussion of your theories would be out of scope
in this forum.
It appears your approach will be controversial, at least in
the theoretical physics community, given your claim that a
valid reformulation of general relativity theory can be
developed in which light does not move. (SEZ-2013,p.17) Your
approach seems very ambitious, since it appears you aim to
derive quantum mechanics and the Standard Model of subatomic
particle theory from Einstein's general relativity
(SEZ-2013,p.19) and you aim to develop a formalism for
biophysical consciousness by adapting Einstein's equations for
generality relativity, treating consciousness as a universal
primitive similar to gravitation.(Viz. slides 53-59, SEZ-2008;
the term 'consciousness' does not appear in SEZ-2013, which
refers to 'sense' and 'experience' instead.)
So, at this point I am doubtful about the theoretical
prospects for your research approach. Still, I wish you all
the best in your research.
However, further discussion between us is moot at this point.
It is clear we need to agree to disagree in this forum about
AI and consciousness.
Sincerely,
Philip C. Jackson, Jr., Ph.D.
References:
SEZ-2013: Ericsson-Zenith, Steven (2013) On The Origin of
Experience - The Shaping Of Sense And The Complex World.
Preview, read at Stanford University, November 13th, 2013.
https://www.createspace.com/Preview/1137409
and
http://youtu.be/zF5Bp_YsZ3M
SEZ-2008: Ericsson-Zenith, Steven (2008) A New Kind of
Positivism. A Stanford University Seminar, March 13, 2008.
http://www.iase.info/2008/03/a-new-kind-of-positivism.html
Jackson, P. C. Toward Human-Level Artificial Intelligence –
Representation and Computation of Meaning in Natural Language.
Doctoral Thesis, Tilburg University, 2014.
http://www.philjackson.prohosting.com/PCJacksonPhDThesisInformation.html
Date: Thu, 25 Dec 2014 17:13:52 -0800
From:
steven@xxxxxxx
To:
ontolog-forum@xxxxxxxxxxxxxxxx
Subject: Re: [ontolog-forum] Child architecture
Should you watch the video? I'd say that it
certainly can't hurt.
I like Aaronson's work and find much to agree with.
However, It should be clear that I believe, speaking now
as a computational theorist with much experience in
dealing with parallelism and locality, that we need new
ideas. I expect most experienced people in the field
will agree with this. The hump that I overcame a decade
ago is to accept that there is something new in
biophysics and that this is now accessible due to
biophysical research funded for other means. IOW, we
need new physics and we are informed by biophysical
research, primarily funded for medicine. In this
Aaronson probably does not agree with me, but you'd have
to ask him.
What do I mean by "non-locality that is evident in
biophysics"? I would ask you to sit at a desk with a
couple of thumb tacks, one in each hand, and play with
them awhile and to tell me if the answer is not readily
apparent, or better and more fun, to ask you to listen
to Beethoven, view magazines while you masturbate, and
eat honey :-) If the non-locality in biology is not
then readily apparent to you then there may indeed be
(philosophical) Zombies among us.
The important point is to look at the consequences of
sensation, feelings, upon behavior. There is a role for
sense in physics, a role that it plays in behavioral
outcomes. In my work this is a claim for the primacy of
structure, in particular flexible closed structure -
that covers every form of biology from bacteria to you
and I. It is time for us to put aside the inactive and
passive view of what philosophers call "qualia."
With a mathematical characterization of this effect
in flexible closed structure - I propose a couple of
routes to this, in a state of flux - we can specify a
new model of computation able to apply the non-locality
involved. Very generally, consider the following a
preliminary account of the "symbolic processing"
involved. Imagine a holomorphic function that describes
a dynamic "shape" upon the surface of such structure
(producing an apprehension) and a holomorphic function
upon the "opposite" surface that describes the shape of
a response and combine the two in a hyper-function.
Imagine a simplified bacteria with receptor formation
potentials at one end and motor function potentials at
the other. My claim is that this mechanism is the basis
of all thinking, feeling, and all "life", at all levels.
It also provides a way into new any-scale
computational mechanisms in which "recognition,"
"memory" and "large-scale decision making" comes,
essentially, for free in energy terms.
There are obviously a lot of questions that remain to
be answered and a lot more work to do.
I hope this clarifies.
Regards,
Steven