Subject: Re: patent trolls and X-licensors
From: simo <>
Date: Mon, 05 Jun 2006 16:35:48 -0400

On Mon, 2006-06-05 at 13:20 -0700, Thomas Lord wrote:
> Don Marti wrote:
>  >> [re RSA]
>  > Would they have not published without the ability to
>  > get a patent?
> The promise of a patent gave them greater incentive to work
> within the academy and on those particular problems.
> Without that system of incentives they might have worked on
> something else entirely.  Or they might have accepted an offer
> to do their work under trade-secret or similar protection.

Not in the Academy Thomas,
in the Academy your reputation is built upon the papers you publish.
In the academy the patents are not really an incentive on making good
science, on the contrary they distract you from good science, and led
you to study only what might drive revenues (not counting all the time
wasted on preparing patent applications and keeping your stuff secret
before the patent application).
Science is essentially sharing, the patent system, as it stand is not.

>  > Did the quality or quantity of math and CS
>  > publications go up when software patents became
>  > available?
> It is hard to be certain but the RSA result suggests that in
> some cases the answer is "yes".

Please give me some data on what make you say yes.
Do you have clear evidence?

Do you have any comparative data that shows that labs that were awarded
patents produced more science than labs that were not so awarded.

To you have a score of that department before and after patents started
to be issued in that field, do you have any clear analysis that should
that production of research was indeed boosted?

During the EU case I've seen a couple of empirical studies that shows
the contrary, they shows that companies that produced more patents,
actually increased the investments in legal departments and decreased
the investment in their own R&D department!

>  > If we're Society and we have Limited Resources, paying
>  > people a bonus to do what they would have done anyway
>  > seems like a waste.
> Compare it to giving out a stock option at a start-up.
> Because of the opportunity costs, relative to spending cash, a
> start-up can more easily afford to give out a stock option.
> And that's doubly good because since the employee wants to
> collect the benefit of a stock option, he now has added
> incentive to help the company succeed.

And when the company fails the employee ends up with just having worked
underpaid for some time ...

> As a society of limited means, we can't afford to give out
> million dollar prizes every time someone successfully completes
> their NSF grant -- we'd have trouble competing with private
> interests.  But some of the grant programs are aimed at growing
> the economy by creating new types of businesses.  We can
> afford to give out the patent and that is doubly good: it gives
> the researcher incentive to help make sure that successful
> businesses get established.

It really depends on the goals.
Good research is open and is uncertain in the outcome. If you are _not_
bound to obtain a patent at the end of the process, then you can do good
science, because you don't fear to miss your goal, you can just explore
a matter in full, and in the end you may end up finding something new
and great that you could not imagine when you started.
If your goal is a patent you will not research anything that does not
clearly led you to something immediately valuable. Does the state need
to compete with private companies in producing applied science? I don't
think so, a state need to foster good science, because that is the kind
of science private enterprises are not interested in.

> I object when you say "to do what they would have done anyway".
> Without the patent incentive, research at the academies would be
> very different.  Rivest, Shamir, and Adelman may have done some
> very high quality work -- on a very different topic.  As it was,
> though, as a matter of public policy, we were able to give them
> incentive to work on this problem.

You don't need that in public institutions, private enterprises already
have that incentive. They need to compete in the market, they need to
find out practical solutions when a problem arise.
If you already have competition why should you put a patent on top? You
already have incentives. You should put (the right) incentives where you
don't have them instead.

>  > How does sitting on an algorithm and charging
>  > admission constitute "offering nifty crypto tech"?
> As nearly as I can tell, RSA beat back inferior alternatives,
> including some that would have become "standard hardware",
> because at a critical point in history RSA Data Security
> Inc. got out there and sold the thing to a lot of critical
> firms, especially firms in financial businesses.
> That selling effort -- the education, persuasion, and reduction
> to actual practice -- takes a lot of effort and money.
> RSA Data Security, Inc. wasn't "just sitting on an algorithm".
> They were building a business around the practice of deploying
> that algorithm and, as a natural side effect, helping to ensure
> that their algorithm had real-world relevance.
> RSA Data Security, Inc. was also, I think, gracious and smart
> and managing their patent lifetime.  They permitted
> non-commercial use.  They engaged in public education.  They
> handed off to the public domain quite well.

Do you think that without a patent, financial institution wouldn't have
looked for the best algorithm they could use and firms would have tried
to provide the best they could to serve financial institutions?