Subject: Re: "incentive void" (was Re: A different patent covenant...)
From: <stephen@xemacs.org>
Date: Thu, 5 Oct 2006 15:07:13 +0900

Thomas Lord writes:

 > Stephen, I think you misunderstand the patent system on two points:

I do not claim to understand the patent system in full, but I do
understand the points you cite.

 > you've left out the question of obviousness.

I have.  Do I have to discuss the implications in light of all of 35
USC every time I make a point?

 > My view is that if you *really* need 20 years of protection
 > and you *are not* willing to put a price tag on putting the
 > invention into the public domain that "reasonable time"
 > and "reasonable public" (or, if you prefer, the degree of
 > directness of suggestions in earlier teaching) should be
 > a very high bar.

This is sufficiently incoherent that I'm really not sure what you're
saying.

I guess that what you're saying is that innovation that is hard work
should be rewarded on a cost-plus basis ("need long protection"
implies that profit per period is small), but only if it's really
great ("very high bar").  Innovation that is based on a bright idea
and almost no work, on the other hand, should be given rewards where
the sky is the limit on ROI (because presumably there's a minimum
acceptable term, say three years, and your licensees, not you, will be
doing the hard work).

To coin a phrase, are you trying to invent a market for patronage?

IMO, this is a bad idea.  The conventional take is that you can give
incentives for work this way, but not for brightness.  Brightness must
be selected, rather than given incentive.[1]

Abolition is a much better idea.

 > Wouldn't it be nice if an inventor could say to the PTO
 > that, "No, this isn't like inventing AC power generation and
 > transmission and it's not like inventing the telephone but
 > it was hard, valuable work and still deserves *some* protection."

No, it would not.  Patents are not about rewarding inventors; they're
about rewarding innovators.  That's why they're *property* (unlike
author's moral rights), so they can be sold by someone who knows how
to get them to someone who knows what to do with them.  Nor is it the
innovator's place to decide the value; that is the role of the market.
That is why the term is given *before* you see the innovation, so that
the deal is subject to the relevant market uncertainty but as little
as possible to regulatory uncertainty.

If you don't like that theory, avoid patents, because they're the
wrong tool for the job you want done.  Patents are very well-tuned to
that theory, despite the inherent defect of the monopoly franchise and
the practical defect of transactions costs.  So much so that I really
doubt that "evolving" patents will result in a usable new kind of IP.

I recommend that you start from first principles.  What activities do
you wish to reserve to the rightsholder, and in what contexts?  What
activities should be explicitly permitted to the public?  How do you
define these in terms that would be verifiable in court?  What
incentives do you expect those terms to create?

 > Also:
 > > [....]  That is the biggest breakage in
 > > the current patent system: the fact that only patents are considered
 > > prior art.  
 > 
 > That's simply false.   I think you mean that earlier patents are given
 > what you think is disproportionate weight in the review of applications
 > by the PTO but that isn't the same thing.

Stipulated.  (What was that you said about "agreeing in a way to make
it look as much like disagreement as possible"?)

 > Voluntarism to the rescue?  The patent review initiative that
 > International Characters plans to use (and that some big companies
 > have also signed up for) doesn't require new legislation -- only
 > voluntary, early disclosure.

Well, yes.  So you guys have mostly just thrown mud at the volunteers.

Even the calm and reasonable folks like Brian Behlendorf have not
helped much to answer the questions that Larry and Rob have posed, at
least not that I've understood:

    1.  Can FSBs do anything profitable with IP covered by such a
        public covenant?

    2.  Are there further tweaks that are "cheap" (better, no-cost)
        for patent-holders that can improve the deal for FSBs?

I realize you want a lot more than the IC covenant offers, and I think
the direction you're heading is fundamentally good (despite my many
criticisms of the details).  But at the voluntary level, that's going
to require hard work by Larry and Rob to decide whether and when
they're willing to make the compromises necessary to do "better" for
FSBs than the covenant they've offered.  For them, it's not just a
matter of a couple hours BS-ing on FSB.

And if your idea is really that good, it will require even more effort
on the part of all of us to make it law.

 > There is at least some non-trivial grass-roots support.  "Too
 > little?" remains to be answered but don't go pounding the nails
 > into their coffin just yet.

Uh, I expect the volunteers such as International Characters to live
long and prosper.  It's the crabgrass-roots of the free software
movement that I see digging holes, building boxes, throwing them into
the holes, jumping in, and trying to pull us in after.

Look.  Go ahead, count the number of words of unsolicited praise for
IC's voluntarism posted to this list recently vs. the number in
complaints that they're patenting at all, and proposals that they
sacrifice large portions of their legally obtainable revenue.

The thing about IC's voluntarism that nobody here seems to appreciate
is that they have created social value out of NOTHING.  There are no
compromises for them; they are giving up NOTHING, yet manage to give
free software (as opposed to FSB) ONE HECK OF A LOT.

They come here to ask: "Can you do something with this?"  And "Are
there further cheap tweaks we can make to improve the deal for you?"

I suggest they deserve a huge round of applause and serious
consideration of the questions they asked, because the status quo
really really sucks compared with what's on offer.  Once those
questions have been dealt with, *then* get to work on compromises that
benefit free software at positive cost to the "volunteers".

You could try for a quid-pro-quo.  "Hey, Larry, my time is
valuable.  I'll review your covenant for FSB aspects if you'll review
mine for legal."


Footnotes: 
[1]  However, the patronage idea is a little novel.  Conventional
selection models involve *employable* brightness.  If there's also a
kind of *random* brightness that cannot be directed at any given task,
but it requires constant care and feeding or it dies, patronage makes
more sense as compared to employment.  I don't know if it makes
*enough* sense, though! ;-)