440 Davis Court #1602
San Francisco, CA 94111-2496
415 781 5700

Date: Mon, 27 Nov 2000 20:31:54 -0800

04 00067 61 00112701

OHS DKR Project
SRI International
333 Ravenswood Avenue
Menlo Park, CA 94025

Subject:   Alliance and Partners for Development or Research??
OHS Start Narrow Design

Dear Eric,

I agree with you in your letter today that, if OHS/DKR was a normal product development project, then the care you propose to formulate a comprehensive core design for KM, is essential. However, KM is a research and discovery effort. Everybody else has failed, as reported by Lee Iverson at SRI on March 24, 2000. Since nobody else has it, there is no template to follow. Everything is a blind alley, until you understand the core design. Therefore, learning what to create requires experimenting, trial and error, if you start without SDS. I re-did the core design of SDS three (3) times starting entirely from scratch over a period of about five (5) years; and, three (3) years ago, did another major change to about 60% of the code. Thankfully it was not necessary to start over again with a new concept. Experience with each design stage builds awareness of the pieces and integration that lead to understanding a stable architecture that now seems to work. This is not a sterile exercise of simply using a tool, and every once in awhile making an improvement, or fixing a problem. Rather, it entails constant interruptions of daily work and life. Often it is not convenient to stop using the program in order to create an improvment; yet, if you wait, you lose understanding of the nuance about the issue that needs correction. So, the synergy between use, discovery and development is powerful, but not necessarily fun, nor conducive to a well balanced life. Yet, it is not clear how this discovery stage can be avoided, if you start from scratch.

Of course many feel they are not starting from scratch, in that they have been to school, and produced a lot of good software. I could be wrong, but experience seems to show that this talent can only be applied to KM, if it is either directed toward implementing the core KM design, based on an SDS-type template; or, by discovering the core design by doing KM using succeeding initial designs, so that you get the feel of what is needed. It has to become part of your life, because when you augment intelligence, that is, in many ways, the core of humanity. It is not in a book, it is not on the web, except in POIMS. Talented, successful people don't want to look at POIMS. They want to collaborate with friends and colleagues. But those people don't know how to do KM, so collaboration does not yield the answer, until you begin to collaborate on implementing the core design, per your letter on February 8, 2000.

I can tell you that, once you discover the meaning of knowledge, and the intelligence process that transforms information, you will gain the advantage of no longer being boggled by a lot of connections. Email and other other information will become a resource that grows knowledge for you and others every day, rather than a continual worry about searching through the archives, as you related on August 24, 2000. SDS makes the record a productive part of your memory that puts you in command of history, what folks on the project call a "repository." Improving memory is a big part of augmenting intelligence, which you recommended as the purpose of KM on April 23, 2000. This benefit is not without social cost. You will be cut off from people, even your friends, because everybody else will be boggled, as you worried in your letter to the team on January 25, 2000. No one will understand you, since they will not have had your experience. You will comment at meetings and send correspondence that will not be heeded, because credibility is the dominate dimension of life, and your capability will be incredible, i.e., beyond belief. Then you will feel the burden foretold by the Legend of Prometheus, as noted by Henry van Eykan in his letter on November 22, 2000.

As I said in the letter on November 21, 2000, you can build a better KM "mouse trap," and in less time than I did. I am only suggesting that you may not be able to avoid the stages of discovery that require experience with succeeding designs, such that the normal product development process does not apply. If KM could be done that way, it would already have emerged, based on all the effort over the past 10 years or so, where people have been chasing the proverbial Killer/Compelling application. KM is it, what I call Communication Metrics. It may take another 100 - 200 years for it to catch on. We are just doing the spade work, tending the garden of knowledge awaiting future generations.

Some might ask, if we cannot get it all done in our lifetime, then why try? This goes to the core of why we are here. I guess my thought would be that in exchange for the privilege of life, as we know it, each generation owes its best effort to advance civilization. In ways large and small, we do what we can.

If you were at Microsoft, Intel, HP, GE, take your pick, they would not approve trial and error in order for you to discover the meaning of "knowledge." They would want you to prove that requirements are all figured out, before approving the project, as you properly propose today for sound management of product development. It would only cost a few $M to do the trial and error research to discover the meaning of "knowledge." But, it would not be approved, because big companies do development, not research. Development requires market (not technical) research, showing that "intelligence" saves time and money, and that people are willing to pay $20 at Fry's in order to avoid continual bumbling. They would overlook the case study of civilization that shows cost savings of knowledge compared to continued reliance on information technology. They would worry that cultural resistance is too high a price to pay for improving earnings. They would say the idea that you can augment intelligence is funny and alien, just as Doug was told many years ago that linking is a pretty funny idea, but behold, it works: see telecon with Doug on October 29, 2000.

Perhaps, this letter will show in a small way, that not only does linking work, but that it can be put to work growing knowledge for people and organizations through a practice we have come to call knowledge management. The design is not complex, but pioneering discovery, without guidance, is an arduous journey.

As you say, thems my thoughts.



Rod Welch