September 17, 2004
Background material on why a National Project is required
Link to: Tutorial on the
Nature of a new Memetic Technology
based on categorical
abstraction and event chemistry
Steven Smith
Ron Buikema
Intellibridge Corporation
A group of scientists have been developing an "anticipatory" information science based on several deep lines of scholarship, including quantum cognitive neuroscience and semiotics (theory of sign systems).
One of our proposals, in 2002 was deemed fundable by NIMA's Novel Intelligence in Massive Databases program (The Glass Box), but not funded. The innovations that we are offering peer reviews well in certain natural science communities, but not in the IT or computer science community.
Meanwhile the group has continued to develop an information science paradigm that is predicated on rejecting machine deductive reasoning as a primary focus.
The rejection opens the door to Human-centric Information Processing (HIP), as discussed at:
http://www.bcngroup.org/beadgames/anticipatoryWeb/18.htm
Deep lines of scholarship reinforce the concept that HIP will be found to be more satisfactory in environments where human communities need to know what is going on in real time. HIP enables the use of anticipatory mechanisms available to humans, as a consequence of our natures. The enabling comes from a
1) provably optimal way, the keyless hash table, to solve the set membership problem (for example, 32 machine cycles to compute if a token is or is not in a set of 10,000,000 tokens)
2) conjecture on stratification of physical processes (and on the existence of "systems of metastable substructural ontology")
3) formative and differential ontology that extracts facts from latent semantic algorithms (algebraic and stochastic LSI).
4) mutual induction between computer states and human cognitive acuity
5) formal modeling of natural systems exhibiting points of complexity (tipping points)
Soft cognitive engineering arises quite naturally from HIP. Ecological affordance, arising in the natural world in real time, can be observed and response patterns predicted based on frame filling type computation.
In a sense this computation might be regarded as "deduction" but we are more comfortable regarding this as mere computation, leaving the notion of inference to human activity that is informed by real time aggregation of categories of structural information derived directly from processing massive, or small, streams of data. Our software visualizes these categories and patterns of category aggregation, while using a key-less hash table to support two orders of magnitude faster processing of streams of data.
The key-less hash table is an innovation beyond a 2003 patent ... (the innovation is not hard to understand and can be demonstrated).
Even in the case of stochastic induction, and similar forms of category formation and routing into categories; we feel that this is useful but is mislabeled by most information sciences as a form of "deductive inference". The inference is actually a hidden induction that occurs from the assumption, by the human, that some type of abstract structural measurement and computation has a correspondence to real world events or phenomenon. HIP separates this assumption and allows the inductions, by humans, to be explicit and open to manipulation by humans in real time.
Notational work on part of the foundation to this work is at:
http://www.bcngroup.org/area2/KSF/Notation/notation.htm
Our working material for the Oct 13 preliminary proposals to ARDA is at:
http://www.ontologystream.com/beads/nationalDebate/115.htm
I am looking for assistance in orienting our work to current IC information science needs.
I inquire about Intellibridge's interest in serving as an interface between Mitre and elements of the IC and a group that will be centered (tentative) at West Virginia University, George Washington University and perhaps two other universities.
I am able to make technical presentations on the underlying technology.
Not everyone is interested in collaboration with university scientists, particularly a group that has become known as working outside the artificial intelligence and cognitive systems "box". So I will understand if Intellibridge Corporation is not interested in working with us. If this is the case, in fairness, you might talk with others within the community to see who might champion the cause my group has been defining for some time.
We offer an alternative.
Dr. Paul S. Prueitt
Founder OntologyStream Inc
Research Professor
The George Washington Univ