Subject: Re: How accurate is Metcalfe's law? (Was: Ximian software)
From: Ben_Tilly@trepp.com
Date: Fri, 4 Jan 2002 15:22:16 -0500


Ian Lance Taylor writes:
> Ben_Tilly@trepp.com writes:
[...]
> > All of the proposals I know about for switching away from capitalism
for
> > a complex society either failed miserably or were never implemented.
My
> > judgement of the ones not implemented is that they would have failed as
> > well.  Betting on a successful replacement being found does not strike
> > me as wise.
>
> Well, if you want to consider a world in which robots can replace
> humans to be a dystopia, go ahead.  Personally, I'll just kick back
> and let the robots build my house and grow my food.  My main cost will
> be energy, I suppose, but I expect we'll have cheap fusion by then (if
> you can assume smart robots, I can assume cheap fusion).  Since
> increases in living standards historically tend to reduce birth rates,
> we may even avoid severe population pressure.

You might do this, but how will most people afford said robots?

At the time robots get capable enough to do this, they cost as much as
a person's salary, so most can't buy them.  Then they start alternately
eliminating jobs and cutting salaries, so as the cost falls people
never earn enough to afford one.

You may not believe it would happen this way, but I do.  I would even
bet about it if I thought I would live long enough to see the bet
resolved one way or the other.  Given that, and the fact that this is
OT, I think we will have to agree to disagree.

Incidentally IMHO a better bet than cheap fusion is cheap solar power
collected around Mercury and transmitted to Earth.  Big advantages
over fusion are no radioactive pollution, and we have the technology
to do this.  It just isn't economically viable.  Yet.

> > It may be wildly optimistic.  Let me redo the estimate.
>
> I replied off-list, but basically I think this skips all the hard
> problems and focuses on the easy ones.

As I replied off-list, I don't believe your hard problems are relevant
barriers.  Again, if I thought I would be around in 2050, I would be
willing to bet you on this.  But I don't think that, so again we should
just agree to disagree.

> > > I think one of the things this list struggles to investigate is the
> > > economics of abundance.  Our current system of software licensing
> > > based on copyrights and patents imposes scarcity on what would
> > > otherwise be an abundant resource.  It does this because some people
> > > think that is the best way to grow the resource; capitalism, which is
> > > the best way we know to grow resources, is based on managing
scarcity.
> > > But the only true scarcity in the software world is people's time,
and
> > > software licensing is not closely tied to that resource.  If we
> > > eliminated the artificial restriction of software licensing, what
> > > would the economics look like?  That is the world that FSBs live in.
> >
> > My answer, in brief, to the FSB problem is to locate the bottlenecks
> > where potential abundance meets scarcity, plant yourself there, and
> > make yourself an efficient solver of that scarcity problem.  But if
> > you (as many naively want to) find the abundance and position yourself
> > to stand in the middle of the abundance providing more, you may have
> > fun but you are likely to starve.
>
> Well, if I described the problem correctly, then the scarcity is
> people's time, and the positioning you mention means finding an
> efficient way to translate people's time into software which people
> need.  In other words, a consulting business priced by value to the
> customer, which I believe is in fact where most FSBs make money.

Agreed.  Furthermore Karsten's figures indicate that most software
companies of all types make their money there.  Which puts FSBs and
proprietary companies in pretty direct competition.  This means that the
real question of interest for our attempts to gaze through the crystal
ball is whether this market is something that FSBs can consistently
address more efficiently than those pushing proprietary solutions.

My weakly held opinion is that the pattern of the last few decades will
continue to hold.  In emerging markets we will again and again see
proprietary solutions explore and demonstrate the viability of new
technologies.  (Proprietary solutions which may use free ones in their
toolkit of course.)  Again and again as these markets mature there will
be a replacement process where open solutions become viable, get
created, then come to dominate the new niche.

(-: Yes, I have heard Tim O'Reilly talk, why do you ask? :-)

Cheers,
Ben