Subject: Re: near/medium future digital media economics
From: "Ben Tilly" <btilly@gmail.com>
Date: Fri, 19 May 2006 20:37:20 -0700

 Fri, 19 May 2006 20:37:20 -0700
On 5/19/06, Thomas Lord <lord@emf.net> wrote:
> Ben seems to be offering this rebuttal:

There is a subtle distinction between a rebuttal and trying to correct
misunderstandings.  This falls in the latter category.

>   a) Forget Metcalfe, like many things free software becomes
>      *exponentially* more valuable over time.  If anything,
>      think Reed, not Metcalfe (but you're missing the point that
>      this has nothing to do with networks).  You, Tom, have
>      no need to show that collaborative effects have these
>      value impacts because...

See, for instance,  The Innovator's Dilemma  for documentation of this
in several fields, as disparate as the speed of chips, how much a
hydraulic shovel can scoop, and the capacity of batteries.  The state
of our knowledge today is better than it was a decade ago.  And it
seems that whenever we find concrete ways to measure that improvement,
the improvement curve is exponential.

Without looking, I'd expect that the precision of our measurements of
the fundamental constants of nature also shows exponential improvement
over time.

>   b) That growth has insignificantly little to do with Metcalfe
>      or anything like it.  Open source processes may bump the
>      exponent of the growth in value, but that growth is just
>      what software itself does -- as do many products
>      susceptible to ready improvement.

What do network effects have to do with the examples that I offered?

>   c) Meanwhile, we are interested in the value of a network
>      *today* -- at a specific point in time -- compared to a
>      rival network.  Even the increment in the exponent of
>      future value achieved by collaborative processes in free
>      software is not important to the value of the *network*
>      because rational actors discount future improvements
>      exponentially.  As Ben puts it: "I really am not interested
>      in nor do I care about what other people are doing [in the
>      free software world]."

This is a key point.  Thinking about, "What do my users value, today,
access to my network at?" is very different from, "Here is how I think
they will benefit over the next several years of using my network."
The one that is going to drive their usage decisions is the former,
not the latter.

> On this basis he declines to give uptake to my analogy between
> free software and hyperlibraries.

No.  On this basis I am pointing out that your arguments for
Metcalfe's Law don't work.  The analogy is a complete red herring.

> Well, then.  There are a some points to refute there.
>
> First, non-consumable, non-rival products (free software,
> hyperlibrary content) do *not* become exponentially more
> valuable over time.  Their cumulative value is bound by the real
> value of surplus of consumable, rival, desirable goods.  By
> various reasonable metrics free software programs (etc) may
> become exponentially more featureful.  They may lead to
> exponential increases of efficiencies in certain industries,
> from time to time.  But non-consumable, non-rival products don't
> have a cumulative value that grows Z^T, they have a cumulative
> value that grows more like (C - 1/Z^T) where C is the total
> value of the consumable "essentials".  Unless you believe in
> magic.

No magic required.

Over time economies show exponential growth, and therefore the real
value of the surplus of consumable, rival, desirable goods also shows
exponential growth.  Which means that your bound is no barrier to
exponential growth in value over time for non-consumable, non-rival
products.

In fact one would PREDICT from what you said that that non-consumable,
non-rival products should experience exponential growth in value over
time.

If you doubt that economies show exponential growth over time, then
pick a fairly stable and well-governed country and look at its GNP
over the last century.  As long as you avoid the Cambodia's and
Nigeria's of the world, you'll find exponential growth sustained over
very long time periods.

For the USA I believe that it has come to about 10%/year for the last
several decades.

> Second, at a large enough time scale, sure, people exponentially
> discount the future potential of improvements.  Smart bet,

This is true even on short time scales.  If you don't believe that,
then I'll gladly offer you $1010 in a year for $1000 now.  You'll make
$10!  I'm good for it, why not do it?  Conversely I'm not about to
offer you that same deal with my repayment coming in a month, do you
see why?

> there.  But here on human time scale the power-law-weighted
> future dominates value in free software and, one would expect, a
> hyperlibrary.  In other words, the near future is worth a lot,
> and further future decays exponentially.  Think about how a CIO
> decides on software acquisition and deployment policy.  As long
> as I'm on a Tiemann jag let me quote him:

Um, if you exponentially discount the future, then the near future is
worth a lot and the further future decays exponentially.  You're
saying the same thing that I did, only with more words.

>     Can you imagine trying to sell a $10,000 support contract to
>     a manager of five embedded systems programmers, and getting
>     hung up on whether or not Cygnus can go public based on its
>     business model? For all that Open Source was a great way to
>     open doors into the best and most innovative software
>     development groups, it proved to be a major roadblock when
>     selling to the mainstream market. We were about to learn
>     first-hand what Geoffrey Moore meant in his book Crossing
>     the Chasm.
>
>     This challenge became absolutely clear when I visited a
>     group of developers who were building wireless
>     communications systems at a Fortune 100 company. As part of
>     their quality process, they not only evaluated their own
>     quality, but the quality of their vendors according to a
>     number of metrics. Of all the different vendors with whom
>     they did business, most ranked "Very Good to Excellent" in
>     most or all of the metrics. Their supplier of embedded
>     tools, however, placed dead last with "Poor or Unacceptable"
>     in all categories for each of the three years this quality
>     monitoring process had been in place. Yet they would not buy
>     our tools because despite our testimonials (from their
>     customers, no less!), superior technical features, and lower
>     price, management did not want to go with an unknown
>     solution.
>
>     http://www.oreilly.com/catalog/opensources/book/tiemans.html
>
>
> The (fractally scalable[1]) "immediate future" is very much the metric
> employed by rational actors evaluating sources of dynamic but
> non-consumable, non-rival content.

Rational consumers discount the future exponentially in general, and
then discount further for perceived risk.  The discounting is
sufficient that they generally wind up weighting the immediate future
very heavily, and everything else fairly lightly.

Note that not all consumers are rational - your Tiemann quote
illustrates that quite well.  (Though I fail to see what relevance it
has for the point that you were trying to make.)

> Think about the value of critical patches in response to the
> latest security vulnerability.  Or critical catch-up to a
> feature your rival has acquired.

Think about the value of buying a present for your wife before versus
after her birthday.  Sure, one can come up with lots of cases where it
matters a lot that one do something NOW rather than tomorrow.  That
doesn't change the trend that exponential is what one should expect.

> And that brings us back to networks.   The immediate future
> of free software and (hyperlibraries) matters -- co-dominates.
> And the incremental improvements on the horizon grow linearly
> with participants (according to my "foundational component"
> argument with which you express basic agreement).  Voila,
> it's a Metcalfe network.

No.

You can't turn a bad argument into a good one by coming up with cool
words (like "co-dominates") and making random assertions.  Well you
can, but you're the only one that you'll convince by so doing.

Cheers,
Ben