Re: [unrev-II] I'm having a problem

From: Eric Armstrong (eric.armstrong@eng.sun.com)
Date: Fri Jan 21 2000 - 18:01:00 PST


From: Eric Armstrong <eric.armstrong@eng.sun.com>

Mike,

I don't think you're going wrong, but rather asking important
questions. My take on some of them...

Mike Taylor wrote:

> I think there is a big difference between improving the collective
> intelligence of an organization - a team
> or a business unit or whatever - and trying to solve what are
> perceived as big social problems. Yet these are treated as the same
> problem....I have no problem thinking of
> an organization as a "social organism" and using the ideas of CoDIAK
> to make it work better....But then the leap to solving perceived
> social problems is too huge for me.

The connection is this:
   * There are serious problems coming down the road
      (At the very least, this colloquium is *motivatinig*...)

    * Every organization that is attacking one of those problems has
       work to do. (The "A" work.) They need to improve their ability
       to do it ("B" work) as well as their *capacity* for improvement
       ("C" work) to stand a chance of catching up to the growing
       complexity of the problem domain.

    * We're working on the "C" point in an effort to augment
       those organizations. The process is recursive -- we also need
       improve our capacity to improve.

    * Here at the outset, the possible approaches are not constrained.
       Using technology seems promising. We have a lot of it, and we
       know it's powerful. But maybe the solution is something
       simpler, like inculcating an ethical principle by virtue of
education.
       At this point, all possible avenues of attack are open. Maybe a
       combination of ideas will work. Or maybe Doug has something
       up his sleeve that we haven't gotten to yet, in depth.

   * To develop the augmentation-ability, whatever it turns out to be,
      we will need to keep an eyeball on the entire problem domain,
      not just one, in order to keep the solution appropriately general.

      If we make it too specific to a single domain, it may never have
      applicability outside of it.

    * On the other hand, if we are ever to push past surface generality
       and wrestle with the deeper issues, we are probably going to
       have to pursue at least one problem domain in depth -- so we
       see what issues come up as the early prototypes get used.
       (While still keeping half an eye on the wider set of problems.)

> But at some point the concept of a unitary organization disappears and
> a market takes over. For example, there is no organization that
> manages the world energy supply. Therefore there is no obvious target
> for improvement.
> There is simply a market of buyers and sellers...

Interesting point. More below.

> It is argued that we are running out of fossil fuels...
> This is really an argument that the market is wrong and is
> setting the price of energy too low...But no matter how apparently
> convincing the arguments that this assertion is true, remember that
> the market price is set taking into account the collective expertise
> of all the participants.
> And remember the Club of Rome and other historical forecasts of
> shortages of commodities which have proven wildly inaccurate even
> though they seeemed convincing at the time. The market worked and
> supply appeared or demand
> moderated as appropriate....In this respect it is hard for me to
> imagine that any knowledge management process is going to be
> successful for this kind of global social issue.

In other words, a single organization is not the answer, in your view
(for this problem, at least). Your claim would be the market is either
(a) wise enough to handle the problem well or (b) at least no worse than
a single organization, and probably a lot better.

> So long as you have a market where economic advantage can be gained by
> trading on the use of private and proprietary information,
> participants will not reveal or share that information in a
> repository...

An important point. In the software domain, on the other hand, we seem
to be transitioning into more of a "gift economy" where open-source
software is available to all. Organizations who provide it, though, make
a business out of providing support, bug fixes, training, books, and by
implementing desired extensions. The "business surrounding the product"
becomes the business, rather than the product itself.

Academic experience shows that respect, credibility, and reputation are
often more powerful motivators than financial reward for knowledge
workers. So it is not totally clear that information will *not* be
shared in the future -- only that the "survival mentality" of the past
has mitigated against it.

Now, competition is good. But if one is competing for respect and pride
of place, rather than for food and shelter, one "wins" by publishing and
contributing more, rather than less. That appears to be true for
individuals, at least. Once basic human needs are met, one is free to
pursue loftier ambitions. Perhaps there is a way to foster that same
kind of attitude among corporations?

(Along those lines, I keep wondering: What happens to the world if all
corporations are required to be "non-profit"? They still compete in
order to
pay salaries, but not to amass billions. Would we be better off? Or
not?)

> If there is no market, then we have a regulated or controlled
> environment where some organization controls prices and supply. This
> has been tried....It was not
> successful. It may be argued that a CoDIAK-enhanced equivalent of
> Gosplan should be created to manage the energy supply. I'm not
> convinced that this
> has any more chance of working than Gosplan. The organization would
> never have access to the amount of knowledge that a market system
> brings to bear
> because the economic incentives to develop that knowledge would have
> disappeared. Even though the CoDIAK enhanced organization might be
> better, it simply would not encompass all the knowledge that all the
> people who participate in decision-making in the market model might
> possess or be
> incented to acquire.

This is a very strong argument that a single organization may not be the

answer. The only effective counter-argument would have to be along the
lines that a DKR-equipped organization would be better than all the
brains
who are actively addressing the issue in the world today. I'm not
prepared
to go that far, yet.

> So in economic terms, the problem comes down to a "tragedy of the
> commons" problem. This is the economic problem where a shared resource
> that nobody owns, such as the common grazing land in a village or the
> fish on the Grand
> Banks, is abused because no-one has the economic incentive to manage
> and care for it.

If I get this right, we haven't got a prayer! We've already said that a
single organization is not the answer. But the "tragedy of the commons"
says that a common resource (like oil) is quite likely to be overworked
and overused for individual financial gain! An oil company, in fact, has
no incentive to conserve oil except insofar as it drives up prices. In
every other respect, they are incentivized to sell more than the
competition, as fast as they can, right up to the last barrel!

I think there is a midground, however. The "tragedy of the commons"
delineates the *proper* role of government. Whenever there is an
advantage to all, but no potential reward for any individual -- be it a
person or organization -- then that is an area where government *should*
be active. Roads are an example: We all benefit, but no one organization
would profit from building one for free. So we have the government take
charge of building and maintaining them.

Taking oil as a great example: What will the oil supply curve look like
in the future? My mental simulation suggests that market forces drain it
at a steady
rate right up to almost the end, when there is a sharp drop off in
supply and prices skyrocket. Disagreements erupt. Economic retaliation
and wars ensue.
If not oil, such a scenario might develop over water, food, or some
other
critical resource.

The alternative, in my mind, is a forward-looking government that sees
the problem on the horizon, and begins working proactively -- raising
prices a dollar a gallon each year for 10 years, for example -- despite
the adverse economic impact that such a move would entail. The result
*might* be to flatten out the oil-availability curve so that we run out
more slowly, with more chance to adjust, and more opportunity to develop
alternatives. (Alan Greenspan appears to have a magnificent grasp on how
seemingly minor "anticipatory" adjustments can prevent major swings in
the country's economic fortunes.)

But I speak from purely a mental visualization, here. How accurate is my
model?
It's totally unquantified, but is it even qualitatively correct? I see
the DKR as
a mechanism for defining and capturing that model so that it can be
refined, improved, compared to other models, and eventually replaced by
something better.

The DKR, as I see it, stands to elevate arguments like these out of the
realm of unprovable assertions. Although a "provably correct" model is
probably an impossibility, at the very least "provably incorrect" models
can be eliminated as a result of real world feedback.

> The DKR (for a social problem) appears to depend on altruism for its
> content rather than economic advantage.
> This is fundamentally different than the situation within an
> organization, where the interests, economic or otherwise, are served
> by the DKR and the management of the organization will provide the
> incentives to support and
> utilize the DKR.

Granted that an organization has a much keener interest in developing
such a thing. Hopefully we can get General Motors to build one that
improves it's automobile-developing capabilities, while at the same time
producing the means for governments to function with a simulation of
intelligence.

--------------------------- ONElist Sponsor ----------------------------

  Hey Freelancers: Find your next project through JobSwarm!
  You can even make money in your sleep by referring friends.
<a href=" http://clickme.onelist.com/ad/jobswarm1 ">Click Here</a>

------------------------------------------------------------------------

Community email addresses:
  Post message: unrev-II@onelist.com
  Subscribe: unrev-II-subscribe@onelist.com
  Unsubscribe: unrev-II-unsubscribe@onelist.com
  List owner: unrev-II-owner@onelist.com

Shortcut URL to this page:
  http://www.onelist.com/community/unrev-II



This archive was generated by hypermail 2.0.0 : Tue Aug 21 2001 - 18:56:37 PDT