How would you disrupt the patent system?(blog.iandavis.com)
blog.iandavis.com
How would you disrupt the patent system?
http://blog.iandavis.com/2011/08/05/how-would-you-disrupt-the-patent-system/
9 comments
And sadly, the Obama guy responsible for the current "patent reform" wants the Patent Office to make the filing and acceptance of patents even faster.
They're completely going in the wrong direction, full speed ahead.
They're completely going in the wrong direction, full speed ahead.
I think the problem here is that there is no universally agreed upon definition of trivial. Many things are trivial in hindsight, there is prior art for almost anything, and your trivial is my innovation where there was none.
The change has got to come from legislation, not litigation.
The change has got to come from legislation, not litigation.
There's probably problems with this, but...
Patents should not be an exclusive license. Instead they should be a limited time income tax benefit. In other words, when you are granted a patent then you get some small percentage break in your income taxes for some number of years. So the only way to benefit from granted patents is to have some actual income! Once an idea is patented, you cannot charge a license fee for the idea as the idea is now public. You can, however, still license implementations of the idea - just like you can with implementations of proprietary ideas covered by trade secret.
The granting of patents would work the same way they do now - first one to file wins. However since you do not get the benefit of owning an idea for the next 20 years and collecting licensing fees on it, the benefit of filing for a patent on new ideas immediately might be a lot lower than it would be to just keep it secret under trade secret and use it yourself as long as possible. The idea here would be to balance the value of an idea in trade with the value of it as an income tax benefit. If you have a bunch of good ideas you've been using for years but are no longer really "secret" you could attempt to patent those old ideas so that the rest of the world can share them and you can get some tax breaks on perhaps the down swing of a product's life cycle. If someone else has already patented the idea ahead of you, well, so be it. The idea is out there already so society hasn't lost anything.
My thinking is that this could be balanced so that the tax benefits are such that companies would want to keep things secret for awhile but then attempt to patent older ideas once products near end of life. It's a race and balance there between keeping something secret and trying to get the tax benefit out of old ideas that might not be very profitable on their own anymore. Any tax benefit would be better than none, so it would seem that old ideas that are no longer used but were being hoarded by the company would suddenly be "valuable" to them, so they'd get the idea out into the public and everybody wins in the long run.
Patents should not be an exclusive license. Instead they should be a limited time income tax benefit. In other words, when you are granted a patent then you get some small percentage break in your income taxes for some number of years. So the only way to benefit from granted patents is to have some actual income! Once an idea is patented, you cannot charge a license fee for the idea as the idea is now public. You can, however, still license implementations of the idea - just like you can with implementations of proprietary ideas covered by trade secret.
The granting of patents would work the same way they do now - first one to file wins. However since you do not get the benefit of owning an idea for the next 20 years and collecting licensing fees on it, the benefit of filing for a patent on new ideas immediately might be a lot lower than it would be to just keep it secret under trade secret and use it yourself as long as possible. The idea here would be to balance the value of an idea in trade with the value of it as an income tax benefit. If you have a bunch of good ideas you've been using for years but are no longer really "secret" you could attempt to patent those old ideas so that the rest of the world can share them and you can get some tax breaks on perhaps the down swing of a product's life cycle. If someone else has already patented the idea ahead of you, well, so be it. The idea is out there already so society hasn't lost anything.
My thinking is that this could be balanced so that the tax benefits are such that companies would want to keep things secret for awhile but then attempt to patent older ideas once products near end of life. It's a race and balance there between keeping something secret and trying to get the tax benefit out of old ideas that might not be very profitable on their own anymore. Any tax benefit would be better than none, so it would seem that old ideas that are no longer used but were being hoarded by the company would suddenly be "valuable" to them, so they'd get the idea out into the public and everybody wins in the long run.
It's really easy to come up with patentable ideas at the current standard of patentability. If you paid people for patents as you suggest, you'd just get flooded with patents for novel but useless contraptions. So then you'd need to get the government deciding whether something is useful and paying people create allegedly useful inventions, rather than the marketplace.
No one would be getting paid (by way of tax discount) for a patent unless they're already doing something the market values and therefore generating an income to tax in the first place. It doesn't matter if the income is from use of the idea in the patents or not. Any income generation is a sign the market is rewarding them for something. The government doesn't need to decide anything differently than they do now.
The patent system is already flooded with useless and novel things. The reason is that people are hoping that over the next 20 years (or whatever) that the idea might suddenly become valuable and they can profit off it. With my idea, at least, you cannot profit from a patent unless you're already profiting in the market.
The patent system is already flooded with useless and novel things. The reason is that people are hoping that over the next 20 years (or whatever) that the idea might suddenly become valuable and they can profit off it. With my idea, at least, you cannot profit from a patent unless you're already profiting in the market.
You're proposing that Joe the candy store owner be given money by the government for filing a patent on kite-powered lawn mowers.
You claim that since Joe is evidently providing something the market values, we should cut his taxes.
But that's what everybody with income does. So why not cut everybody's taxes? We could do that without a sideshow of filing useless patents.
This idea
1. Does not tie rewards in any way to actual innovation.
2. Explicitly suggests that people keep things as trade secrets for a while, which negates the actual purpose of the patent system!
You claim that since Joe is evidently providing something the market values, we should cut his taxes.
But that's what everybody with income does. So why not cut everybody's taxes? We could do that without a sideshow of filing useless patents.
This idea
1. Does not tie rewards in any way to actual innovation.
2. Explicitly suggests that people keep things as trade secrets for a while, which negates the actual purpose of the patent system!
If you know the perfect solution, I'm sure we'd all love to hear it. All you're doing it tearing me down. You're not offering any suggestions to fix these apparently obvious glaring problems in my idea or the current system.
I think patent pools are the quickest way to solve this issue in the short term. It won't solve the patent troll problem, but it will slow it down because everyone in the pool will have access to a vast amount of patents, so it will be harder for the troll to reach any company that participates in the pool, or at least there won't be as many trolls going after them.
But it should work very well in protecting the companies from the pool against large "patent aggressors", or even against other pools from competing firms. Think of it as NATO vs Russia+China.
In the long term, if patents don't get abolished, at least I'd like to see a law that allows companies to opt-out of the patent system[1], so they don't get to sue others, even if they become big companies, but others can't sue them either. This won't kill the patent system overnight, so big companies who are against abolishing patents, because they have so many, will still be able to assert their patents against other big (or small) companies who also have a lot of patents and don't want to lose them overnight.
[1]This opting out of the patent system idea came from someone else here on HN.
But it should work very well in protecting the companies from the pool against large "patent aggressors", or even against other pools from competing firms. Think of it as NATO vs Russia+China.
In the long term, if patents don't get abolished, at least I'd like to see a law that allows companies to opt-out of the patent system[1], so they don't get to sue others, even if they become big companies, but others can't sue them either. This won't kill the patent system overnight, so big companies who are against abolishing patents, because they have so many, will still be able to assert their patents against other big (or small) companies who also have a lot of patents and don't want to lose them overnight.
[1]This opting out of the patent system idea came from someone else here on HN.
> In the long term, if patents don't get abolished, at least I'd like to see a law that allows companies to opt-out of the patent system[1], so they don't get to sue others, even if they become big companies, but others can't sue them either.
That's the most ridiculous idea for patent reform I've ever read. There are two possibilities.
1. Patents in fact do encourage innovation (at least in some industries) by allowing innovators to recoup their R&D costs before the copiers can jump in. Without this, copiers would get in too fast and reduce the price to where the innovator never makes back the R&D costs.
2. Not #1.
In case #2, we should just get rid of patents.
In case #1, which is most likely the correct case at least for industries where it takes a huge amount of R&D to get a new product to market (e.g., drugs), your suggestion would in effect encourage the formation of more copier companies, and give them free reign to copy.
That's the most ridiculous idea for patent reform I've ever read. There are two possibilities.
1. Patents in fact do encourage innovation (at least in some industries) by allowing innovators to recoup their R&D costs before the copiers can jump in. Without this, copiers would get in too fast and reduce the price to where the innovator never makes back the R&D costs.
2. Not #1.
In case #2, we should just get rid of patents.
In case #1, which is most likely the correct case at least for industries where it takes a huge amount of R&D to get a new product to market (e.g., drugs), your suggestion would in effect encourage the formation of more copier companies, and give them free reign to copy.
> [Possibility #1:] Patents in fact do encourage innovation (at least in some industries) by allowing innovators to recoup their R&D costs before the copiers can jump in.
One problem is that patent law reflects a one-size-fits-all philosophy: It takes years, and lots of money, to get a patent, but if you do you've got it for 20 years from the filing date, regardless whether you want or need it for that long. I once argued that the Patent and Trademark Office (USPTO) could change this, without congressional action, but nothing ever came of it. [1]
[1] http://www.ontechnologylaw.com/docs/JPTOS-Reenginering-the-P...
One problem is that patent law reflects a one-size-fits-all philosophy: It takes years, and lots of money, to get a patent, but if you do you've got it for 20 years from the filing date, regardless whether you want or need it for that long. I once argued that the Patent and Trademark Office (USPTO) could change this, without congressional action, but nothing ever came of it. [1]
[1] http://www.ontechnologylaw.com/docs/JPTOS-Reenginering-the-P...
And what's so wrong about copying? Do companies really start a business thinking they won't get copied? Getting copied by more companies means you'll have to innovate faster, and the end result will be higher quality products for customers, and lower prices. I don't think a little higher competition across the board will stop companies from being built.
The religious faith in free markets you seem to imply is a bit naive. In reality, startups would be crushed by the biggest player sucking up their innovations and broadcasting it faster and to a larger audience. Let's not forget that patents were originally meant to level the playing field, and that's not a terrible goal (the execution is another matter).
Mathematically, your free markets need to be clamped at some boundary conditions if you don't want them to get stuck in a local optimum (e.g., big company swallows all). Patents are one way to do this, in theory at least.
Mathematically, your free markets need to be clamped at some boundary conditions if you don't want them to get stuck in a local optimum (e.g., big company swallows all). Patents are one way to do this, in theory at least.
And your faith that small companies actually use patents to defend themselves against big ones is a bit naive, too. If anything the big companies use their vast amount of patents to crush small startups that compete against them, even if they managed to get 2-3 puny patents. But the truth is many of them can't even afford to buy a lot of patents.
Read the point I labeled #1.
I really doubt the opt out idea would work well. Companies that make cheap knock offs don't produce anything unique, and as such would always opt out. Take the drug industry for instance, all your generics manufactures never do blue sky new drug research. They would always opt out. This makes it useless for companies that do perform blue sky research to opt in.
This and plenty of other interesting ideas here: http://news.ycombinator.com/item?id=2810689 and in the replies to OP: http://news.ycombinator.com/item?id=2809544
1. Cut the patent expiration time. Cut it by 50%.
2. Cut the patent expiration time further of non-original owner patents. Cut it by 75%.
They should last anywhere from 1 year to 5 years, depending on the importance of technology. For example, if it's a web related patent, like a "one-click payment system" it should only last a year. If it's a really important technology, in a slower industry, it could last up to 5 years. In the mobile industry, which is very fast paced, it should last only 2-3 years.
How would I do it?
I would create a patent pool administered by a non-profit that only grants licenses (for free) to people or entities who do not own their own patents or own other entities which own patents. Owners of patents would not be granted licenses even for a fee. Patents would be donated to the pool as a form of charity.
No idea if that would work, but that's what I'd try. A key problem I see is that getting a patent is expensive, it's hard for individuals (from whom most contributions would undoubtedly come) to compete with companies who make a game of churning out as many patents as possible (IBM for example). If however it became large enough, and there were changes in the current system to make defensive patents less valuable, hopefully it could reach a certain critical mass where it would make sense for companies to gift all their patents to the pool in order to gain access to the others.
Startups without existing patents would of course benefit from this.
I would create a patent pool administered by a non-profit that only grants licenses (for free) to people or entities who do not own their own patents or own other entities which own patents. Owners of patents would not be granted licenses even for a fee. Patents would be donated to the pool as a form of charity.
No idea if that would work, but that's what I'd try. A key problem I see is that getting a patent is expensive, it's hard for individuals (from whom most contributions would undoubtedly come) to compete with companies who make a game of churning out as many patents as possible (IBM for example). If however it became large enough, and there were changes in the current system to make defensive patents less valuable, hopefully it could reach a certain critical mass where it would make sense for companies to gift all their patents to the pool in order to gain access to the others.
Startups without existing patents would of course benefit from this.
Government will not solve this problem. Companies should operate anonymously on the internet, and just ignore patents and copyrights. Then they can't be stopped and IP laws will become obsolete.
5 years exclusive ownership - must be available to be licensed and may not be buried. Renewal costs double each time. Don't renew and it becomes available for anyone to use.
Possibly this method might help in disrupting patents http://peerpatent.org/
The main problem today is that the patent office grants too many trivial patents. It isn't surprising that it does so, it is motivated to do that: It gets paid per patent, and the faster a reviewer agrees to a patent, the quicker the work for that patent ends.
The only way to fix that is to reverse or at least balance the incentives. Which means, a big cost for every trivial patent the patent office issues.
It would be good if lawmakers passed legislature that did that. But we can't expect them to. A class-action lawsuit can do the same, though. Everyone that has ever been sued or forced to take a license for a trivial patent has been harmed by the patent office not doing its job properly. All those people have standing to sue. For such a massive amount of people, a class-action lawsuit is the way to go.