4 ms·

A patent is a deal the public makes with an inventor: publish your invention and we'll grant you a time-limited monopoly on it. In order for this to be a good

A patent is a deal the public makes with an inventor: publish your invention and we'll grant you a time-limited monopoly on it. In order for this to be a good trade, the invention has to make a real intellectual contribution to the art. The problem we have is that no one is unambiguously empowered to ensure that it does so. The PTO does some of this, but has neither the bandwidth to do a good job nor a clear mandate to be hard-ass about it. And then once a patent is granted, the courts tend to defer to the PTO on its validity. I haven't read the patents in question here, but I nonetheless feel safe in saying that "technologies for tracking vehicles and providing users with electronic updates" cannot possibly make a contribution to the art significant enough to counterbalance the societal cost of the monopoly the patent holder has been granted. I'm sure there are thousands of engineers here on HN that could design and build such a system using well-known techniques. In some domains, the determination of how much of an intellectual contribution an invention makes is fairly straightforward. If a novel molecule is synthesized -- the chemical and pharmaceutical industries keep a record, I gather, of all molecules whose structures have been worked out, so it's possible to know when a new one appears -- and then proves useful for some medical purpose, it's pretty clear that a contribution has been made, and what that contribution consists of. In software, it's much harder. The field is massive and sprawling, and no one can keep up with all of it. That's why I'm convinced the only solution is to amend the patent laws to require objective evidence of nonobviousness -- objective evidence that the invention makes a real intellectual contribution -- in order for a patent holder to successfully sue for infringement. Such evidence could come, for example, in the form of publication in a peer-reviewed journal or conference proceedings. Absent such evidence, the law should direct the court to invalidate the patent. An advantage of this approach is that it requires no special-case rule for software patents, which run into the difficulty of defining the term "software patent" precisely. Rather, in domains other than software, the required evidence is generally not hard to come by. So the rule is domain-independent in definition -- and defensible as such, I believe -- but would have by far the greatest impact in the software domain.

Grovara123 13y ago
Everything you have here is debatable... obviously. But interesting perspective.
kinsho 13y ago
While you have good intentions, I have a feeling your idea would backfire some how. Judging the objectivity of a patent through its inclusion within a peer-reviewed journal or a conference is only going to lead to more loopholes that can be easily exploited.
- ScottBurson 13y ago
  It wouldn't be automatic. Courts consider many kinds of evidence of varying strengths, and the probative value of evidence can be challenged by the opposing side. If it appeared that a journal had sprung up specifically to legitimize patents, evidence supporting that view could be introduced as well. Courts exist to weigh such arguments. In fact, "objective evidence of nonobviousness" is already a well-established concept in patent law [0]. The change I am proposing is, where currently patents are presumed valid until challenged, under my proposal they would be presumed invalid until evidence of nonobviousness were presented that the court found satisfactory. I don't think we're going to get rid of patents altogether. No patent system is perfect, but I think my proposal would make for a massive improvement. [0] https://en.wikipedia.org/wiki/Inventive_step_and_non-obviousness#Graham_factors https://en.wikipedia.org/wiki/Inventive_step_and_non-obvious...
dalke 13y ago
"If a novel molecule is synthesized ... and then proves useful for some medical purpose, it's pretty clear that a contribution has been made, and what that contribution consists of." That's not the case. There might be one person who thought of the scaffold, another three involved with trying different variations of the side-group, a fifth who had to solve problems when synthesizing it, and so on. Even on the far end, after it's in trials, there's the chance that it's useful for some other effect altogether. That's how Viagra's most notable use was discovered. See http://www.forbes.com/sites/johnlamattina/2013/07/17/will-payments-to-scientists-for-new-drugs-improve-gsks-rd-productivity/ http://www.forbes.com/sites/johnlamattina/2013/07/17/will-pa... for a recent commentary on the difficulty of figuring out who made "the" contribution.
- ScottBurson 13y ago
  Oh, I'm well aware that it can be hard to figure out who made the contribution. But I didn't say anything to contradict that. Reread the sentence you quoted: all I said was that it was clear that a contribution had been made: a previously unknown molecule had been synthesized, and a medical use found for it. This is all that matters for the question of whether a patent should be granted. The question of to whom it should be granted is not one I'm considering here.
  - dalke 13y ago
    I did misread your statement. I interpreted "a contribution has been made, and what that contribution consists of" as "a contribution by specific people" not "a contribution to humanity." I apologize. However, that does not change the overall intent of my reply, which is that drug discovery is not so clear-cut as you make it out to be. Take "A previously unknown molecule had been synthesized". Part of drug development includes finding previously known molecules which have new and unexpected actions. Again I point out sildenafil, which was synthesized for use in hypertension and angina pectoris. It was a nurse, during phase I trials, who discovered that was an effective against erectile dysfunction. As another example, insulin is found in the body. It was first synthesized in the 1960s, but diabetics had to use insulin from animal sources, which was expensive and hard to acquire. Genentech, in 1982, started selling insulin made using E. coli genetically modified with human DNA. Now insulin is relatively inexpensive. Insulin was not new. The medical use of the chemical was not new. But the method of creating the chemical was new, and for that Genentech received a patent, and a lot of money. This case does not fit your description of how the drug discovery market. Then there's combination therapies, where the use of two different and already existing drugs has a synergistic effect. Here's a list of some combination therapies which were patented in 2011: http://sciclips.wordpress.com/2011/09/19/recent-combination-therapy-patents/ http://sciclips.wordpress.com/2011/09/19/recent-combination-... . Again, these are known chemical compounds, while you've twice asserted that the chemicals has to be "previously unknown" in order to be patented. To summarize: getting a drug patent does not mean that the chemical has never been seen before. Then again, your thesis doesn't require that it be unique. You point to drug patents because you believe they must have an "objective evidence of nonobviousness", and you want patents to have the same requirement. However, that non-obviousness criterion is already required to get a patent. Quoting from Wikipedia: "In modern usage, the term patent usually refers to the right granted to anyone who invents any new, useful, and non-obvious process, machine, article of manufacture, or composition of matter." It looks like you want patents to have an additional burden of proof of non-obviousness. For example, "publication in a peer-reviewed journal or conference proceedings" may count. However, you don't realize how easy that is to achieve. Peer review is a relatively low bar. There are any number of so-called 'predatory journals' which are advertised as peer-review, and which take your money and publish, but which don't have an meaningful peer-review system. Someone who pays $10,000 to get a patent will not find it difficult to spend another $1,000 to publish in one of these journals. It's trivial to start a new academic journal, and have it staffed by friends of yours. In that way, everyone can published peer-reviewed literature. And "conference proceedings"? That's an almost trivial bar. Your proposal is so easily gamed as to be worthless. How much do you know about the patent system? How much do you know about drug development?

Grovara123 13y ago

Everything you have here is debatable... obviously. But interesting perspective.

kinsho 13y ago

While you have good intentions, I have a feeling your idea would backfire some how. Judging the objectivity of a patent through its inclusion within a peer-reviewed journal or a conference is only going to lead to more loopholes that can be easily exploited.

ScottBurson 13y ago
It wouldn't be automatic. Courts consider many kinds of evidence of varying strengths, and the probative value of evidence can be challenged by the opposing side. If it appeared that a journal had sprung up specifically to legitimize patents, evidence supporting that view could be introduced as well. Courts exist to weigh such arguments. In fact, "objective evidence of nonobviousness" is already a well-established concept in patent law [0]. The change I am proposing is, where currently patents are presumed valid until challenged, under my proposal they would be presumed invalid until evidence of nonobviousness were presented that the court found satisfactory. I don't think we're going to get rid of patents altogether. No patent system is perfect, but I think my proposal would make for a massive improvement. [0] https://en.wikipedia.org/wiki/Inventive_step_and_non-obviousness#Graham_factors https://en.wikipedia.org/wiki/Inventive_step_and_non-obvious...

dalke 13y ago

"If a novel molecule is synthesized ... and then proves useful for some medical purpose, it's pretty clear that a contribution has been made, and what that contribution consists of." That's not the case. There might be one person who thought of the scaffold, another three involved with trying different variations of the side-group, a fifth who had to solve problems when synthesizing it, and so on. Even on the far end, after it's in trials, there's the chance that it's useful for some other effect altogether. That's how Viagra's most notable use was discovered. See http://www.forbes.com/sites/johnlamattina/2013/07/17/will-payments-to-scientists-for-new-drugs-improve-gsks-rd-productivity/ http://www.forbes.com/sites/johnlamattina/2013/07/17/will-pa... for a recent commentary on the difficulty of figuring out who made "the" contribution.

ScottBurson 13y ago
Oh, I'm well aware that it can be hard to figure out who made the contribution. But I didn't say anything to contradict that. Reread the sentence you quoted: all I said was that it was clear that a contribution had been made: a previously unknown molecule had been synthesized, and a medical use found for it. This is all that matters for the question of whether a patent should be granted. The question of to whom it should be granted is not one I'm considering here.
- dalke 13y ago
  I did misread your statement. I interpreted "a contribution has been made, and what that contribution consists of" as "a contribution by specific people" not "a contribution to humanity." I apologize. However, that does not change the overall intent of my reply, which is that drug discovery is not so clear-cut as you make it out to be. Take "A previously unknown molecule had been synthesized". Part of drug development includes finding previously known molecules which have new and unexpected actions. Again I point out sildenafil, which was synthesized for use in hypertension and angina pectoris. It was a nurse, during phase I trials, who discovered that was an effective against erectile dysfunction. As another example, insulin is found in the body. It was first synthesized in the 1960s, but diabetics had to use insulin from animal sources, which was expensive and hard to acquire. Genentech, in 1982, started selling insulin made using E. coli genetically modified with human DNA. Now insulin is relatively inexpensive. Insulin was not new. The medical use of the chemical was not new. But the method of creating the chemical was new, and for that Genentech received a patent, and a lot of money. This case does not fit your description of how the drug discovery market. Then there's combination therapies, where the use of two different and already existing drugs has a synergistic effect. Here's a list of some combination therapies which were patented in 2011: http://sciclips.wordpress.com/2011/09/19/recent-combination-therapy-patents/ http://sciclips.wordpress.com/2011/09/19/recent-combination-... . Again, these are known chemical compounds, while you've twice asserted that the chemicals has to be "previously unknown" in order to be patented. To summarize: getting a drug patent does not mean that the chemical has never been seen before. Then again, your thesis doesn't require that it be unique. You point to drug patents because you believe they must have an "objective evidence of nonobviousness", and you want patents to have the same requirement. However, that non-obviousness criterion is already required to get a patent. Quoting from Wikipedia: "In modern usage, the term patent usually refers to the right granted to anyone who invents any new, useful, and non-obvious process, machine, article of manufacture, or composition of matter." It looks like you want patents to have an additional burden of proof of non-obviousness. For example, "publication in a peer-reviewed journal or conference proceedings" may count. However, you don't realize how easy that is to achieve. Peer review is a relatively low bar. There are any number of so-called 'predatory journals' which are advertised as peer-review, and which take your money and publish, but which don't have an meaningful peer-review system. Someone who pays $10,000 to get a patent will not find it difficult to spend another $1,000 to publish in one of these journals. It's trivial to start a new academic journal, and have it staffed by friends of yours. In that way, everyone can published peer-reviewed literature. And "conference proceedings"? That's an almost trivial bar. Your proposal is so easily gamed as to be worthless. How much do you know about the patent system? How much do you know about drug development?