ORB Visualization

(soon)

Building the case for a WF&A complaint on

DARPA funding of improper IT in the intelligence community

1/25/2004 10:49 AM

Whereas “cognitive engineering” is representative of a mainstream, the center of gravity for the knowledge sciences is "outside" the cognitive engineering paradigm.

The primary difference is in regarding the assumption of some type of reductionism, and the appropriateness of the types of reductionism that has occurred within the cognitive engineering paradigm and within the paradigm called artificial intelligence.

Our belief is that the Friends of the Intelligence Community (IC) meeting last week was pre-arranged, by funding sources, to bring an, unearned, scholarly endorsement of what can be called "the strong from of cognitive engineering". The intent is to make a business case, and successful business cases have been made on the cognitive engineering and artificial intelligence paradigms for five decades.

We suggest that cognitive engineering is not “all bad” and that this is good research when put into proper context. However we do object to claims that knowledge sharing, within the complex analytic processes involved in national intelligence, can be enhanced using a heavy imposition of structure on the communication efforts between knowledgeable humans. This is an automation process.

There is not clear evidence that this heavy imposition of structure on the communication efforts is productive. In fact, most at the Friends of the Intelligence Community (IC) meeting did in fact express reservations, in spite of the economic rewards to go along with the business model.

The evidence is, we claim, that analysts use other tools (phone, faxes, and human language) and only when results are to be presented higher up the chain of command, are the already deployed cognitive engineering tools used. One should not conclude that IF proper presentation of technology was imposed that we could not find gains in some things. However, our claim is that the types of software tools that are being developed at great costs are not capturing the full spectrum of characteristics of human knowledge sharing. The funds are being mis-spent.

Of course, the weaker claim is that some thoughtful structure in the form of shared taxonomies and ontology constructions can be useful if the imposition is fully informed about the true natures of human cognitive and human knowledge sharing within communities. But developing business cases around this scientific proposition has been impossible because of the way the procurement system is manipulated by business models.

This is simply unfortunate, since the business person really has no stake in delivering more IT product of the type that has already been rejected by analysts. But the business person is unwilling and unprepared to understand what the alternative value propositions are.

The causative problem is that the business model is required to be simple minded.

We conjecture, that incumbent contractors, to justify specific requirements that lead to a new round of funding, orchestrate “pre-arrangement” of scientific endorsement. In the specific of the Friends of the Intelligence Community meeting, the sum of money involved may be as much as 20 Million dollars (in new 2004 revenue).

Our group makes additional comments on this.

In software technology based on this "strong from of cognitive engineering", human communication is truncated in order to produce a control centered intelligence production system. This software ignores the real processes that really occur in social communication and in the production of cognition.

The majority of presentations and discussions at the Friends of the IC meeting were directly critical of this approach.

The anticipation of funding for the hard form of cognitive engineering puts into place a distortion of scientific intent, essentially influencing the science community to go along (with something they know is in-correct) or not receive funding.

It would be an easy process to poll the attendees of the Friends of the IC meeting to expose their professional opinions about how available funds should be distributed.

We observed that few attendees of the Friends of the IC meeting feel that a new round of hard cognitive science software is warranted.

We conjecture that a polling survey would confirm that most feel that the natural science needs to be better developed and that software should be recast. Most also feel that the government funding process cannot address this issue.

http://www.bcngroup.org/procurementModel/structuredCriticism.htm

In fact, most of the discussion at the meeting focused on the non-usability of software designed from this same paradigm over the past two decades.

What we suggest is that the science be deepened and that a new generation of information production technology be created - based on biosemiotics and memetics.

We therefore recommend that funding agencies be directed at developing an open source collaboration between those scientists who can make direct contributions to a new Human-centric Information Production (HIP) technology based on real and available science.

A common feeling at the Friends of the IC meeting was that the software procurement process was broken and that nothing could be done to change a cyclic process that has and will continue to produce software which is not used in the development of intelligence, but then only used in the showing of results developed from manual means (like talking on the phone or looking at raw data).

More than one of the Friends of the IC meeting formal presentations focused on a reality where in policy leaders, generals and members of Congress, are deceived into thinking that expensive delivered software is being used to develop the results that are reported using the software. The repeated and clear evidence is that these demonstrate use the software only in demonstration of the results and not in the development of the results.

The true results are most often developed manually.

Dr. Paul S. Prueitt

703-981-2676