ORB Visualization
(soon)
1/25/2004 9:26 PM
Dr. Rita Colwell National Science Foundation
Dr. Colwell,
Thank you for your short note.
My colleagues and I will request this meeting with Dr. Peter Freeman.
The BCNGroup founders are meeting February 4 with a university leader, regarding our planning for a National Project to be funded directly by the Congress.
http://www.bcngroup.org/area2/KSF/nationalProject.htm
after this meeting we will contact Dr Freeman.
Peter Freeman is the Director of The Computer and Information Science and Engineering Directorate, and as such may not represent the issues that we are addressing in the development of the knowledge sciences and the knowledge technologies.
His Congressional testimony does NOT address the issues of our primary concern.
http://www.nsf.gov/od/lpa/congress/108/pf_supercomp071603.htm
In fact Peter Freeman is supporting what we regard to be a paradigm that is not as interesting as what might be developed if computer science was to take a lesser role in the development of information systems. I will explain the position, but I am sure you have heard this before.
As you know there is a deep and dividing issue regarding the nature of models of causes in complex natural systems such as social systems and ecological systems.
Sir Roger Penrose framed this issue in several popular books regarding the appropriateness of computational models of mind function. The work leads to a model that depends on non-formal models of human awareness. Memetic sciences might be better termed biosemiotics sciences. The fact is that a stratified approach, towards modeling brain function, appears inhibited by classical science and business interests.
But an analogy might be in order to show that sometimes the Powers That Be are simply not exactly correct. Stratified theory might be like Alternating Current when compared to Direct Current as a model of how to use electricity to light homes.
The stratification, I speak about, is local and produces an endophysics (causes within a complex system) and exophysics (causes related to the complex system's environment). The concepts here are not precise and tidy, as yet. But the alternative simply does not account for a large body of research derived from direct observation about the "causes" of behavior. The AI paradigm is in fact nothing but a odd myth, and most natural scientists know better. (Most detail is being left out of my argument against AI, but I am sure you know the scholarship.)
An action perception cycle is needed that is short and involves human cognitive acuity in each cycle. Such an action-perception information technology paradigm follows the scholarship of J. J. Gibson and others.
This "AC" knowledge technology will be simpler, cheaper, and more useful than the current "DC" information technology.
http://www.bcngroup.org/area2/KSF/HIP.htm
The current Intelligence community technologies are DC, and they need to be AC. But for this to occur the funding streams MUST be altered. (This is an opinion). (We are willing to gather our scholars on this issue and debate the NSF program managers; but we would hope that there is a wide enough debate within NSF on this issue.)
The Penrose/Hameroff argument is not the only one. Robert Rosen makes the argument from a different perspective, and questions the categorical correspondences between mathematics (and computer science) and natural systems.
They are not the only approaches to a fundamental criticism of computer science as currently practiced and funded by the Federal government.
Our more immediate problem is the exclusive funding by the NSF, NIST and DARPA of the alternative viewpoint, and the control of funding decisions by business processes.
http://www.bcngroup.org/python3/fortysix.htm
But, in any case; we will follow up on your suggestion.
With deepest respects and appreciation for your thoughtful conversations
Dr. Paul S. Prueitt
Founder, BCNGroup.org