Linus on drugs [was: Open Source Development Labs -> Open Source Legal Labs?]

[Alexander Terekhov]

-------
Name: Linus Torvalds (torvalds@osdl.org) 12/3/06

Chung Leong (chernyshevsky@hotmail.com) on 12/3/06 wrote:
>
>That's a strange way to argue it. You're speaking as
>though there's a "normal" market in which new drugs
>without patent protection would be developed. Let say
>we remove this market distortion. Can you explain how
>the economics of developing a drug that could be
>immediately copied by others?

That's a rather idiotic argument. It's sadly a very common
one.

How do you think most drugs got invented historically?

Do you think penicillin was invented because Alexander
Fleming or St Mary's Hospital (where he was working) was
looking to make a huge profit?

What kind of sad and dark world do you live in, that you
believe that people only do things because they want to
help "big pharma" make billions and billions of dollars?

In other words, your whole argument is not only totally
stupid and scary, it's also provably irrelevant.
Most of the drugs people use today are actually not even
protected by patents, and were invented totally without
any of those protections.

In fact, even a lot of the patented stuff was invented
not because of patents, but despite them.

Btw, penicillin is again a great example: not only was
it discovered without any patent push what-so-ever, but:

- what pushed it to be developed was actually World War
   II. Arguably, wars are a hell of a lot better
   at pushing medical technology than any IP protection
   racket has ever been. That doesn't make people think
   that wars are good, though. But why do you blindly
   believe that very strong IP protection is good, if you
   don't believe wars are good? The facts and reality do
   not agree with you.

- The problem with penicillin was production, not
   the drug itself. That's arguably an area where patents
   have been more successful, and a lot of people who do
   not like patents on software, business methods, or on
   drugs and biological elements are much more likely to
   support patents on things like factory methods.

- But even there, in the production part, where
   patents were actually used, there is soem argument that
   the patents actually made penicilling production
   less effective. It wasn't actually a drug company,
   but the USDA (yes, the government agency) that actually
   came up with the best way to big production.

In other words, one of the most important drugs of modern
times totally lays to waste your idiotic and unrealistic
argument.

It's sad how people seem to believe - despite all evidence
to the contrary - that somehow patents are "required" to
make people even want to develop drugs.

So. Try to back up your opinions with facts instead
of trying to make the inane (and unsupportable) argument
that patent protection rackets are the only way to make
progress.

Money doesn't actually make the world go round. It revolves
around the sun quite well even without us having to pay it
to do so, and the same is true for technical innovation.
There is basically zero support for the notion that
technical advances (in any area) depend on strong IP laws,
and there are lots of examples where the biggest advances
were done in the absense of strong IP rules.

Yet people continue to blindly blather about how you "have
to" have patents. With zero actual fact to back it up. And
somehow it's gotten such a common belief that you don't
even get questioned most of the time.

It's called a "myth". Being widely believed does not make
it true.

                Linus


Name: Linus Torvalds (torvalds@osdl.org) 12/3/06

Dean Kent (dkent@realworldtech.com) on 12/3/06 wrote:
>
>What you are now getting into is something we might call
>'scalability', and we might even look to what we call a
>'pyramid club'. How long can the current process
>continue to work? There are already signs that it cannot
>hold up, so to continue to defend it by using past results
>may not be very smart.

Good point. One large reason you can't really compare
data from different points in time is that the technology
itself tends to change how things are done. Extrapolating
that doesn't work very well, and what worked a hundred
years ago may not work that well today or a hundred years
from today.

As a trivial example - not that long ago an average
"product" could basically be designed by a single person
and wouldn't step on many patents not just because there
weren't many patents, but also because things just
were fundamentally simpler. In that situation, maybe the
whole notion of patents seemed like a better idea, and just
worked better too.

Btw, I don't think IP is a bad thing per se. The bad
thing about IP laws these days are not that there are IP
laws (including patent laws), but the fact that they don't
work, and that people have forgotten why they exist.

A lot of people seem to think that patents exists to make
money for people who somehow "earned it" for coming up with
the patent (never mind that they may not be the people who
actually did the work, and that it's often very unfair and
not at all as clear-cut as the "they deserve it" crowd seems
to think).

Not so. The only reason patents exist is to make
for better public knowledge. I'm sure that a lot of
people on this particular group realize that, but I'm also
equally sure that a lot of people in general never saw that
point, and never realized that patents are not there to
reward the patent inventors, but solely to make sure the
benefit to the public.

That was the original intent. Not "these guys are smart and
should be rewarded". But "these guys told everybody else
how to do it, and should be rewarded".

And once you approach it from that angle, the whole patent
troll "toll bridge" thing is obviously totally bogus. A
patent troll is the antithesis of what patents were
supposed to be about. It's not about making money, it's
about encouraging progress and making new technology more
widely available.

So I'd not necessarily scrap patents per se. But I would
seriously suggest that you might want to

- disallow "holding companies" entirely, aka a "use it or
   lose it" approach (which, btw, is not new to IP law. It's
   how trademarks work, for example).

- mandatory licensing. Patents are not there to make the
   patent holders happy - they are there to help the public
   yet give incentive to innovators. However, "incentive"
   is the secondary part, and the "public good" is the
   primary one, and that means that patents that are bad
   for competition are bad patents. You must not make
   patents so expensive to license as to hurt competition:
   competition is more important for the public good than
   any "incentive".

   Again, mandatory licensing is not a new thing in IP law.
   It even exists for patents. But it should be something
   that all patents have to have, and you shouldn't
   be able to not license your patents (or ask for insane
   amounts of money, which amounts to the same thing in
   practice). Example: mandatory radio licensing with a
   fixed fee (obviously the fee still generates discussion,
   but that's at least a step in the right direction)

- no "algorithmic" patents (software or business
   models etc).

- get rid fo the insane and broken patent requirements
   that just allow any crap to make it. Let's hope the
   current case in the supreme court can at least go part
   of the way on this one. There's clearly too much crap.

The first two are problems regardless of whether a
patent is "good" or not. The last two are commentary on
how bad some patents are, and how broken the whole
system about getting them is.

And the above is just suggestions. And it's not like the
copyright industry couldn't clean up its act too. 95 years?
I'm sorry, but if you can't make a better mickey mouse in
fifty years, you simply don't deserve to make money
any more.

            Linus

Name: Linus Torvalds (torvalds@osdl.org) 12/4/06

slim (x@y.z) on 12/4/06 wrote:
>
>Other effects of no IP protection include:
>
>The only other tool would be trade secret.

No.

That's another fallacious argument ("If we didn't have
patents, people would just do things in secret").

There's a lot of reasons why that's fallacious, but let
me just state the two most obvious ones:

- People who argue against patents often do not
   argue for "no IP protection at all".

   Most of the arguments against patents are that they are
   simply broken. The USPTO is simply not able to do a good
   job on them, and a large portion of why the USPTO
   can't do it is not necessarily because they are not
   competent, but simply because the whole system is geared
   against it. There are no incentives to try to make the
   USPTO patent examiners actually do a good job, and there
   are lots of incentives against it (ie, a patent
   that gets rejected isa lot more work for a USPTO person
   than just allowing it - so of course crap gets
   allowed).

- What do you think a "patent" is? It's just a law, saying
   "if you tell others, you get these rights". However,
   there is nothing that says that "these rights"
   has to be a "monopoly". Nothing at all. Granting a
   monopoly is actually a really bad idea, because if the
   idea is to help the public good (and it is), then a
   monopoly by design is probably the worst way to
   do so. Because it inhibits competition, and competition
   tends to be better at driving technology than any
   patent monopolies have ever been shown to be.

In other words, the people who argue that patents are "good"
are invariably totally ignoring real and undeniable facts.

There really aren't all that many signs that say "patent
protection improves innovation". Really. There are several
studies in different areas that say exactly the reverse!

In other words, "strong IP protection" is not at all
guaranteed to drive innovation, but I don't think anybody
here really doubts that competition doesn't drive
innovation.

If you look at the CPU space, for example, look at the
CPU's that have the least IP protection. Look at
AMD and Intel, that have certain cross-licenses because
IBM forced it on them. Then tell me that strong IP
protection actually helps innovation. BULLSHIT.

So, instead of "monopoly" (which is bad for very real
reasons) powers, the "encourage people to tell what they
do" could work on a lot of other levels.

For example, it could be very simple: instead of a patent
law, you could have a very basic commerce law: if you sell
a product publicly, you have to tell people what it
contains, and what it does.

It could really be that simple. The "reward" for telling
people what something is, is simply the right to sell it
at all. Trust me, you want something like this anyway,
since it would not be a good idea to allow drug companies
to sell drugs that they don't tell people what they contain.

So really, the whole "people would use trade secrets"
argument is totally bogus. There's no way in hell that
a drug company must ever be allowed to trade secret the
contents of a drug. They could use trade secrets about
the manufacture, but that, I feel, is actually much
more fair. Go wild. Who cares?

The nice thing about trade secrets is that the burden of
maintaining the secret is solidly on the company
itself. So the costs are there too. If you want to have
a secret, the cost is that you have to keep it secret, and
obviously cannot use standard (and much cheaper) ways of
doing things.

So trade secrets in many ways are much fairer than patents.
In particular, if somebody else figures it out, they have
free hands to do anything they damn well please - no
artificial monopoly of any kind.

                   Linus
-------

regards,
alexander.