Subject: Re: patent trolls and X-licensors
From: "Ben Tilly" <btilly@gmail.com>
Date: Wed, 7 Jun 2006 11:52:45 -0700

On 6/7/06, Stephen J. Turnbull <turnbull@sk.tsukuba.ac.jp> wrote:
> >>>>> "Ben" == Ben Tilly <btilly@gmail.com> writes:
>
>     Ben> On 6/5/06, Stephen J. Turnbull <turnbull@sk.tsukuba.ac.jp>
>     Ben> wrote:
>
>     >> >>>>> "Ben" == Ben Tilly <btilly@gmail.com> writes:
>
>     Ben> But I believe that patents are counterproductive for that
>     Ben> goal.
>
>     >> How in the world would you know?
>
>     Ben> Excuse me?
>
> I think there's good reason to believe any beliefs in this area are
> just wishful thinking.  I gave those reasons, you may disagree on
> their quality.

I would put it differently.  I would say that there are a lot of
subtle and not so subtle factors that are difficult to untangle, so
that it is not obvious what is better in the long run, and it is very
difficult to quantify results.  That means that beliefs are likely to
reflect people's biases rather than objective reality.

>     Ben> I said *I believe*.  You're not about to find a better
>     Ben> authority than me on the topic of my beliefs.
>
> Isn't that beneath you?  Surely you don't post on public lists
> believing that nobody is going to pay attention to a word you say,
> just because you write "it's a belief/opinion."

It is not beneath me.  I post on public lists expecting people to be
able to understand that I have different degrees of evidence for
different opinions, and some opinions I have less objective evidence
for than others.  I believe what I believe, and I've given some of my
reasons for my beliefs.  But I don't have objective evidence that
should convince those who disagree with me and I *know* that I don't.
So if you push me for that evidence, you won't find it.

> I, for one, take your unfounded beliefs seriously.  This time I want
> more than that as evidence, though.

I hope that one reason that you take my unfounded beliefs seriously is
that I try to be careful in formulating them.  But that same care
means that I'm going to try to avoid lying by omission and make it
seem like I have better evidence than I do.

I've read both pro and anti patent discussions, and in the end the
opinions that I came to are based on anecdotal evidence which is
mostly from personal experiences.  I wish I could give you more, but I
can't.

>     Ben> Tom has said a tremendous amount, and if I wade through it
>     Ben> I'm unlikely to draw the right conclusion about what part of
>     Ben> it you're thinking of.  Which exercise of Tom's are you
>     Ben> talking about?
>
> The root of this thread.  The proposal to reform the patent system in
> a way that (if I understand correctly) is intended to formalize the
> notion that the patent office is just a registrar to prove priority
> and publish the application.

That sounds fair.

>     Ben> As for "commercializable products", I'd rather see wavelets
>     Ben> show up in open standard and free software products.
>
> And to hell with the 99% of mankind who wishes computers were never
> invented, except for their X-Boxes?  Personally, I've got everything I
> *need* already; I'd love it if all software from now on were open
> standard and free, 'cause that would be a lot more fun.  But the guy
> in the mirror says "Now, Steve, don't you think that's a tad selfish?"

Here is the problem.  There are a lot of good ideas in the wavelet
world that get into one or two products (if that) and that is it.
People don't put them into standards. They can't get implemented in
free software.  And the result is that the ideas are *only* taken
advantage of in specialized products that really need what wavelets
provide.

Don't get me wrong, those products are often important.  I'm as happy
as the next guy that MRIs are now many times faster to take because
they use a smart wavelet transform rather than a Fourier transform.
If I'm the one who doesn't have to spend 8 hours in a metal tube I'm
not going to argue about the fact that the software is proprietary.
However those same ideas are not finding their way as quickly into,
say, digital photography.  Whenever I watch a slow jpg fill in on a
web page I can't help but think that I saw software in 1996 that
looked better at low bit rates and whose quality filled in a lot
faster as you downloaded data.  That is sad to me.

Furthermore it is my impression that research into practical
applications of wavelets is actually going *slower* than it would go
without the patent system.

A random irony, one of the most popular test images used in digital
imaging research is a small piece of a 1972 Playboy centerfold.  The
same people who were so quick to patent everything and see that their
intellectual property was protected were rather cavelier about
Playboy's copyright!  (Playboy for their part debated enforcing it and
decided not to.  So I guess there is now an implicit licence for the
image.)

See http://ndevilla.free.fr/lena/ for more.

>     Ben> The famous example [of an innovation that the organization
>     Ben> was unable to exploit] that people like to offer is Xerox and
>     Ben> the GUI.  But it is not isolated.
>
> Emacs and GNU comes to mind. :-)

You wouldn't be biased, would you? :-)

>     Ben> When you offer those organizations the protection of patents,
>     Ben> the natural tendancy is to reflexively patent every idea that
>     Ben> comes along.  Which then means that those ideas are now left
>     Ben> to rot in organizations that simply can't pursue them for one
>     Ben> reason or another.
>
> That's nowhere close to obvious.  On the one hand, either somebody
> does a patent search and buys/licenses it, or they don't and they
> reinvent it.  It's not at all clear how those effects balance out
> against the one you describe.  On the other, do you really think that
> people are going to *give away* their R&D products merely because they
> don't have patents on them?

No, it is not obvious.  In fact the reasons why some organizations are
better placed to use an idea than others is also not obvious.

However I'll note that I have yet to hear of a real case of someone
who needed to solve a technical problem choosing to solve it by doing
a patent search looking for a previous solution.  And I don't think
that it is just a matter of putting a better search engine on patents.

>     Ben> How many startups are started by people who had a good idea
>     Ben> while in the wrong organization who then create the right
>     Ben> organization to take advantage of those ideas?  Lots.  I
>     Ben> think it would be more without the patents.
>
> Again, not obvious.  How do those VCs manage to extract revenue from
> ideas whose authors rather often quickly learn to hate the hand that
> funds them?  Where would the money that funds startups would come from
> if the pea-brained capitalists (at least with regard to the
> engineering side of the business) could not acquire property rights in
> the ideas that would be of use *without* the author?  Worse, with the
> author in *direct competition* with them?

What you're saying about the practical role of patents in startups
doesn't agree with my understanding of how either works in practice.
Perhaps my opinions on that are biased by
http://www.paulgraham.com/softwarepatents.html, but I know that I
could tell you all of the potentially patentable ideas that my
employers business has, and if you implemented all of them and tried
to compete with us head on you'd lose money hand over fist for a very
long time.

>     Ben> Why does it matter how straightforward a path it is to
>     Ben> economic value?
>
> Because I'm *assuming* that motivation for *part* of the government's
> policy.  Did I not just write "as long as it's at least amusing" as
> sufficient justification for outright grants to science?  But it may
> also make sense to formulate a policy that encourages picking off
> low-hanging fruit.

Basic science historically has a lot more value in the long run than
the short run.  When we try to make the path to value more
straightforward we run a big risk of improving short term gains at the
cost of long term gains.  That's not a win.

>     >> I don't believe those incentives are really there, in general
>     >> anyway, without some form of intellectual property.  I can tell
>     >> you that in
>
>     Ben> While it isn't in a legal sense, in a very real sense being
>     Ben> able to put your name on something as its discoverer is a
>     Ben> form of intellectual property.  And it is a valuable enough
>     Ben> form of property to be able to drive research.
>
> Research, schmesearch.  The question I'm asking is does it actually
> affect the lives of lots of people, to the extent that they're willing
> to put in their 40hrs a week to buy it?  You know what?  I asked rms
> for a Japanese OCR program 8 years ago.  As of 8 months ago, I
> couldn't find one that reads *English* well enough to be more
> effective than hand typing it.

Ironically wavelet technology is useful in both OCR and voice
recognition.  (More for voice recognition than OCR, but it shows up in
both.)  Which ties back to my impression that patents have slowed
research and adoption of wavelets.

> Fortunately, it doesn't matter.  I have 12(!!) different proprietary
> OCR programs now, just counting the CDs in my scanner/printer driver
> box.  I paid $350 for the first, 9 years ago, $35 for the second, 7
> years ago, and now I get two every time I buy a scanner (as far as I
> can tell the bundled Windows and Mac programs are actually different
> programs in most cases).  Nobody who cared about his reputation would
> be willing to be associated with any of this crap, I assure you---but
> it all beats hand-typing hands down, and any one would be worth $500
> to me.

Did you try http://jocr.sourceforge.net/?  Never tried it, but I just
looked for open source OCR and that is what I found.

Cheers,
Ben