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

Date: Sat, 13 May 2000 10:07:31 -0700

04 00067 61 00051301

Paul Fernhout
Kurtz-Fernhout Software

Subject:   OHS/DKR Meta NIC Fallback Plan

Dear Paul,

Thanks for thoughtful analysis and commitment in your letter today to move ahead on the DKR project despite lack of complete agreement on the details of open source. The tension you cite reflects underlying binary forces of competition and cooperation, between responsibility and authority. Like debits and credits, life and death, up down, 0-1, left right, dare I say man, woman, we need not wait for the perfect solution, in order for everyone to profit from cooperating, as you point out.

The big issue is whether anything can be done to significantly augment human intelligence, and if so how can that be unleashed to help everyone do a better job?

Andy Grove was interviewed by Charlie Rose the other day, and was asked what is the next wave of technology? Andy paused for a bit, and then offered up "implementing what we already have", or words to that effect. Andy did not have the vision to see a new wave forming on the horizon that not only will implement, but in fact extend existing capabilities by an order of magnitude. One big task for the open source community is to help Andy and others who know where to the get money needed for doing the work, that KM is a good description to frame the future. Andy was right not to cite KM because we have not yet given him the words to explain it, what Eric calls the "elevator" version that works good on television.

Doug has done important work on this. He recalled during a recent meeting at SRI (according to Joe Williams' letter on May 10) that a foundational issue NSF posed was to define knowledge. Accordingly, Doug has asked the current "seed" team to tackle this issue, set out earlier on January 20, 2000. So far there has not been enough time. Eric points out in a letter on May 3 that KM is difficult to grasp, and you have commented on the onerous challenge of formulating useful data structures. Jack Park noted on February 21 the bucket of snakes that is revealed when we begin to think about ontologies.

So, we have a lot of work to do, and for some, there is not a lot of time left to get it done.

Suppose there is a way to make progress on a core capability that also facilitates open source development of useful tools and processes. If we spend all of our time arguing about how to pay the bills, the path for development may pass from the scene, perhaps lie dormant to await another chance awakening by future generations.

Right now, we need leadership to focus on taking the next step.

I see a lot of leadership and wisdom in your letter today.




Rod Welch


  1. Jones, Morris E.

  2. DeHart, Bill

  3. Welch, Millie