[50]                               home                            [52]

ORB Visualization

(soon)

 

National Knowledge Project

Membership -> .

 

 

 

4/12/2004 7:36 AM

 

Harriet Riofrio

Office of Secretary of Defense

 

 

Harriet and all,

 

In looking over the agenda and discussion leading up to the April 14thmeeting, I am wondering if I might be allowed to make some observations regarding the correctness of Semantic Web language, for example the story regarding a machine intelligence assisting an individual in medical treatment (first part of the May 2001 Scientific American article by Tim Berners Lee and Jim Hendler). 

 

I have discussed the difficulty that alternatives have in getting a voice. 

 

http://www.bcngroup.org/python3/fortysix.htm

 

But over the past few years I have been able to develop a language that separates what is likely correct about the Semantic Web from what is clearly a type of science fiction (centered around the notion that a silicon processor understands the state transitions in a way similar to the understanding that occurs as part of human introspection.)

 

Conjecture of the intelligence failures

http://www.bcngroup.org/beadgames/conjecture/two.htm

 

One alternative technology (Anticipatory Web)

http://www.bcngroup.org/beadgames/InOrb/home.htm

 

A private evaluation of In-Q-Tel investment

http://www.bcngroup.org/beadgames/evaluation/home.htm

 

Technology as an evolutionary process

http://www.bcngroup.org/beadgames/techInnovation/home.htm

 

Graphical representation of human knowledge

http://www.bcngroup.org/beadgames/graphs/home.htm

 

The core problem that some natural scientists have, with the concept of Semantic Web, is with the viability of using long deductive chains, metadata and sets of processing rules – in cases where novelty and incompleteness of information is known to lead to false sense making (premature closure).

 

The alternative is more complex, and sets aside the use of machine-based inference in all but those cases where it is known that the natural world is in fact entirely mechanical.  My sense, from Jim Hendler and other SW advocates, is that SW advocates do not want to have an objective discussion about the limitations of deductive inference as coded in computers.  Because they are in fact fighting for recognition and funding themselves, it is often the case that they are forced into an inhibition of the alternative that I propose.

 

The alternative cannot compete with the force of history and with the type of inhibition that comes from SW advocates.  But the alternative can be placed into historical context.

 

http://www.bcngroup.org/beadgames/graphs/two.htm

 

What should be the response to my alternative suggestion (Anticipatory Web as an alternative to the Semantic Web), is a discussion where one does not have to accept the complete, religious like, authority of Jim Hendler or AI experts at MIT or Doug Lenat.  One can critically examine the MIT focused claim that RDF is arbitrarily expressive (of knowledge) or Jim Hendler’s claim that the Topic Map standard offers nothing not already available from RDF.  But the W3C folks (often) flame anyone who would suggest an alternative view. 

 

Due to self-selection by the standards process, no one in the W3C standards committee is left who will endure the single mindedness of those whose authority is based on a willingness to be an authority on a standard.  I will not speak for Steve Newcomb or Michele Biezunski as I know that this issue has been one from which great personal suffering has arisen. 

 

The discussion about a viable alternative to SW does not occur. 

 

Moreover, SW or Cyc ontology advocates shoots down my many funding proposals.

 

SW and Cyc business models do not want Anticipatory Web (AW) alternatives to be in the marketplace of ideas.  This may make some folks mad that something like this would be said, but said it is. 

 

The work that I have proposed over the past three years is representative of a school of thought that has suffered the same fate that I have.  Intellectual authorities reinforced by business models marginalize the work.  Proprietary mechanisms will not allow a true science grounded discussion to be fairly made. 

 

I have made the argument many times, and am tired of having to make it.  But what else can I do, given economic ruin that continues even in spite of the many accomplishments I have been pleased to be part of. 

 

It is hard to know where funded AW work would lead to. 

 

But anticipation depends on an encoding of invariance, and this is exactly where ontology representation would be most useful.  Anticipation also depends on top down metadata that constrains the formation of awareness (in real time).  The Anticipatory Web would provide both ontology structures about the structure of invariance in data and top down metadata that is underconstrained so that in real time a human perception is required. 

 

I discuss this concept in the context of the control of complex manufacturing processes at:

 

http://www.bcngroup.org/beadgames/InOrb/eight.htm

 

The core AW concept is that formal semantics is not a correct concept, since meaning is meaning only when the pragmatic axis is present.  This axis is only present in real time and is not experienced by silicon processors.  Humans, on the other hand, cannot but experience the present moment and thus have somehow evolved the ability to act and behavior in a fully anticipatory fashion.  

 

The AW alternative to SW proposes that the computer can only address structure and cannot assign meaning.  Once this is acknowledged, then the formative and differential ontology technical means, that I invented, will be seen as a means to allow formative processes in real time and human-centric intelligence production. 

 

Dr. Paul S. Prueitt

Chantilly Virginia

703-981-2676