One could argue that ARM's "practice" is producing designs. They are heavily involved in the creation and implementation of their patents, making them different from any troll that licenses "patents" they bought rather than "designs" they created.
So patents are only valid if held by their originator? That would make the IP non-transferable. In ARM's case, it would mean a takeover would invalidate its value.
This is a problem that needs to be addressed carefully in my opinion.
Right now, it's a perfectly valid decision for an inventive/creative person(or team or company) to choose to spend their time inventing new things instead of commercialising "finished ideas", since the laws are (and have been for decades) written to allow an inventor to sell their patents.
If you're the world's best battery chemist, I'd rather you spent this year looking for an even better battery, rather than needing to be involved in making and selling last years "best battery" to earn any money from it. I don't want the top cancer researchers "wasting time" being personally involved in making drugs at commercial scale and doing deals with Walgreens, I want them doing cancer research - and having a patent/legal system that allows them to earn a living doing so (with the understanding that they've got a choice to license or sell their patents, and whether they get to share in the spectacular riches that some big-pharma company will make if they've just saved millions of lives will depend on how they chose to structure that transaction).
Somehow, we need to allow inventors to continue to be "NPEs", while stopping the obvious "patent troll" techniques of buying up uncommercialised patents with the only intent being to use them to litigate.
(And note, by "inventors" I'm including "companies set up to do inventing, either as their sole business or perhaps with a few staff or division who 'invent' as part of their regular work". I know of companies like http://www.novogen.com and http://www.mesoblast.com who are pretty much teams of biotech researchers intending to work out how to cure cancer, then sell the knowledge to other companies to commercialise it. I also know people at http://www.ecoult.com who didn't invent the technology they're commercialising, but have licensed the patents from the government research organisation that did invent/patent them.)
Nothing in this bill prevents a inventor to sell his finished invention once to whoever want to commercialize it to a product. Nothing. He doesn't even need a patent to do this.
Say that the best battery chemist invents a way to make current lithium-ion battery packs 25% better. In a bidding war, the price would easy reach hundred of millions, because the practical effects of having better batteries are economic interesting for companies being restricted by ineffective batteries.
Regarding cancer research, I find it quite wrong that state funded research is being patented. Most serious medical research is being paid for NiH with the use of tax money. Society has already paid for all the step regarding the research and it due to get what it paid for. Research which has no bases on NiH funded research theories, and has no relation with any public funds what so ever is an exceptional case, and there is a lot of companies out there that produce derivative works from NiH produced research and gets a patent for it. derivative works that has been created from tax money should not allow someone to get a 20 years state enforced monopoly on it.
Most serious medical research is being paid for NiH with the use of tax money.
This is simply not true. The NIH pays for basic research into NMEs (New Molecular Entities). The gap between an NME and a drug (i.e., a molecular entity, approved manufacturing methods, approved usage, identified side effects, monitoring) is huge.
The gap is then derivative work. They basically take the core research done by NiH for free and refine it and demand in return a 20 years monopoly from the state. The biggest single actor in medical research is by a large factor NiH, paid through tax money.
Drug companies are the only market where this is happening. Why should they get this free research if we want the market to handle the research of medical innovation? You don't see the state doing the basic research of mobile phones and then giving it away to Samsung or apple. They can't just take some already taxed funded research and "refine", getting themselves a 20 years monopoly for the effort on the expensive of the public.
And what is the benefit for the public? They give their money away with tax money and get zero back in return? It is taxed paid research, and then the state gives out a additional 20 years enforced monopoly to the first one who do the last step and refine it. Maybe a bit ism'ish, but is this the core idea of capitalism; having the state pay for the core part of the medical industry without getting anything in return?
You don't see the state doing the basic research of mobile phones and then giving it away to Samsung or apple.
The state (mainly the military) did basic research in signal processing and electromagnetism. Apple and Samsung turned that basic research into consumer devices you can buy at the store.
They give their money away with tax money and get zero back in return?
If private entities didn't turn molecular entities into drugs, then yes, the public would get nothing for their money.
The fact is you can't buy arbitrary chemicals in the pharmacy, and very likely you believe it should be illegal to do so (do you favor abolishing the FDA and all drug regulation?). Pfizer needs to turn drug ideas (what the NIH generates) into actual drugs, just as Samsung needs to turn coding schemes and relativistic corrections into actual consumer devices.
This isn't much of a problem if the inventor sells the patent outright to the company that is going to use it. Then the new owner is not a troll and could legitimately sue infringers. I agree that it's problematic if the inventor structures the deal as a license.
The problem there is that Intellectual Ventures would be allowed to remain, but they're the largest patent troll in the industry and a large part of the reason that this conversation is even taking place.
In one sense, this suggestion is as good or better than any I've heard. Let the defendant raise the possible issue with a judge - let a (presumably impartial and intelligent) judge look at it on a case-by-case basis, and let the plaintiff know in selected cases "this one's going to cost you lots if you lose". I guess there's then be a danger of a "reverse class action" thing, where someone like the scan-to-email patent holder sends out hundreds or thousands of legal threats, and some law firm decides to try and run up half million dollar costs for each and every defendant.
As long as the judge is aware when the suing party is a shell company vs. an actual small business and requires some sort of monetary escrow only for the former.
I don't know if "danger" is the word you are looking for :) I'd love to see a "reverse class action" happen to an intellectual vulture; the only way it would be better would be if the judge allowed the "defendants" to pierce the veil when it came time to collect.
The main distinguishing factor is whether or not the alleged infringement was actually independent rediscovery or not.
Patent trolls are dangerous because they hold patents on things that are easy to independently rediscover. It's too easy to infringe by accident, without ever reading troll patents.
>So patents are only valid if held by their originator? That would make the IP non-transferable.
They could sell to anyone who practices the invention, just not to a patent troll.
>In ARM's case, it would mean a takeover would invalidate its value.
This is a good point, but it's cleanly solved by exempting mergers. That would allow two ARM-like companies to merge without losing their ability to license original patents, but it still beats back the trolls because a merger only allows you to combine two complete entities, it doesn't allow you to transfer anything less than the entirety of each company's patent holdings to a third party.
Your description hardly sounds clean. In fact, given the complexity of business mergers, sales, taxes, etc., it sounds like it would result in an ever expanding nest of exemptions.
There's nothing wrong with companies that buy IP and do nothing but licence it; that's how an efficient market in IP is likely to operate. The problem is with the frivolousness/obviousness of the original patents.
>Your description hardly sounds clean. In fact, given the complexity of business mergers, sales, taxes, etc., it sounds like it would result in an ever expanding nest of exemptions.
What does it have anything to do with taxes or any of that other stuff? If you have two corporate entities that merge, the merged entity can be considered the original assignee of all of the patents that each of the merging entities was. It's not a complicated rule. But it keeps trolls from setting up "mergers" to falsify their status as an "inventor or original assignee," because it's all or nothing. You couldn't transfer any subset of the selling entity's original patents to anyone else and still call them original patents.
Here, we can make it even more simple. Forget about mergers. You can transfer a patent for which you are the inventor or original assignee to another party and have that party be considered the original assignee if and only if you transfer every such patent you own to that party.
>There's nothing wrong with companies that buy IP and do nothing but licence it; that's how an efficient market in IP is likely to operate.
So say the patent trolls. It reminds me of what the high frequency traders say about liquidity: Even if it could theoretically be true under some set of artificial constraints, in practice the cost of allowing it pretty clearly exceeds the supposed efficiency benefits.
>The problem is with the frivolousness/obviousness of the original patents.
The problem is that software patents anyone is likely to infringe are inherently frivolous/obvious, because the lack of physical constraints causes there to be a million good ways to do any given thing in software, so (putting aside standards-essential patents that constitute their own special brand of unintended consequences), the only way you get a software patent that anyone else will ever infringe is to abstract the so-called invention out so far that it covers all plausible implementations rather than only the specific one that you're supposed to have invented.
The patent system wasn't designed for software. They're not compatible with one another. The proper solution is to eliminate software patents. But if we can't do that today then putting as many sensible limits on the amount of economic damage they can cause in the meantime will have to suffice.
> If you have two corporate entities that merge, the merged entity can be considered the original assignee of all of the patents that each of the merging entities was. It's not a complicated rule.
What about when companies spin-off subdivisions as separate companies? Can they "legitimately" own a patent and then spin-off a troll with that patent as its only asset?
Or what if the original inventor own only part of the company? How much do they need to own? If no minimum, then a troll will just sell 0.0001% of the company to the original patent holder. If a minimum, then the troll will find some other way to compensate the patent holder so they effectively doesn't cost them anything to hold that requisite amount.
Seriously, I know nothing about this stuff. I challenge you to find a someone well-versed in mergers and acquisitions who thinks a simple rule would work. Anytime you give property rights to someone but then try to restrict them, there will always be away to wriggle around the restrictions when sufficient money is at stake.
>So say the patent trolls. It reminds me of what the high frequency traders say about liquidity
I don't think your opinion about HFT is universally shared.
> The problem is that software patents anyone is likely to infringe are inherently frivolous/obvious...
I strongly disagree, but it would take us too far afield to debate it. Eliminating software patents might indeed be a good, crude decisions which is better than the status quo, but the abstract nature of software does not in principle prevent novel, patent-worthy ideas from being reliably identified.
>What about when companies spin-off subdivisions as separate companies? Can they "legitimately" own a patent and then spin-off a troll with that patent as its only asset?
Spin offs aren't mergers. If you create a new entity as a spin off and transfer patents to it, it wouldn't be the original assignee anymore, so it would have to practice the invention. A spin off that doesn't practice the invention is a patent troll.
>Or what if the original inventor own only part of the company?
It doesn't matter who owns the company, it matters who owns the patent. If it's owned entirely by either the inventors listed on the patent, or the first company they assigned the patent to immediately on issuance (typically their employer), then they wouldn't need to practice the invention. If it's anyone else then they would need to practice the invention. A company's single exclusive designated successor stands in for the original company if the original company ceases to exist. You can't split up the patents in any way or else all but the designated one of the new owner(s) will have to practice the invention.
> Anytime you give property rights to someone but then try to restrict them, there will always be away to wriggle around the restrictions when sufficient money is at stake.
Any time there is money on the table someone will try to take it. That's a poor argument for not doing anything about it.
>Eliminating software patents might indeed be a good, crude decisions which is better than the status quo, but the abstract nature of software does not in principle prevent novel, patent-worthy ideas from being reliably identified.
Software itself is an abstraction. The only types of patents that software could infringe are patents on mathematics itself (e.g. compression or crypto patents) or patents on abstract ideas that can be built on top of mathematical primitives, because that is all software is. It's abstractions built on abstractions built on abstractions built on mathematics. And none of that is even supposed to be patentable. Contrary to popular belief, patents aren't supposed to be granted for ideas, they're granted for inventions. An idea isn't an invention, the invention is the thing you make after you have an idea. You patent a nuclear reactor design, not E=mc^2.
The "problem" with software is that there are arbitrarily many practical inventions that implement a given idea. Patenting any partial subset of the possible implementations would be pointless, because any competent programmer could come up with a hundred other different ones that implement the same idea. But patenting the idea itself goes too far. Which is what they've been allowing and what causes all of these problems. It's why in software accidental infringement is the rule rather than the exception and as a consequence everything infringes everything.
People have intuitively understood this forever. It's for exactly this reason that literary works aren't patentable, because it would cause the exact same thing. But software makes peoples' brains explode because we created virtual worlds where you can go simulate human interaction and even buy and sell fake stuff for real money, and at some point we went from calling those interactive works of fiction names like World of Warcraft and Second Life and to calling them names like YouTube and Facebook and the App Store.
At some point we lost track of the line that distinguishes a patentable device that physically holds your CD collection from an unpatentable work of interactive fiction that allows you to see a picture of such a device on your screen and pretend to interact with it, and we started issuing patents for simulated "inventions" that exist solely as abstract ideas in software literary works. But it was never supposed to be this way because of exactly the mess that allowing that has caused.
> Spin offs aren't mergers. If you create a new entity as a spin off and transfer patents to it, it wouldn't be the original assignee anymore, so it would have to practice the invention. A spin off that doesn't practice the invention is a patent troll.
Which would make ARM a patent troll - they're a spin-off of Acorn originally and they license designs to others rather than practicing their inventions themselves. I'm not sure you've entirely thought this through. (Probably not such a problem for ARM since all their Acorn-era patents have probably expired anyway, but the next similar company could really get screwed.)
> It reminds me of what the high frequency traders say about liquidity: Even if it could theoretically be true under some set of artificial constraints, in practice the cost of allowing it pretty clearly exceeds the supposed efficiency benefits.
I don't get the ire against HFT. The only ones who should be worried about HFT besides regulators are the human traders who are going to be put out of work by HFT shops.
>I don't get the ire against HFT. The only ones who should be worried about HFT besides regulators are the human traders who are going to be put out of work by HFT shops.
I'm not convinced that HFT is anything particularly insidious, but the fig leaf of providing liquidity is pure rubbish and it seems like the sort of activity that shouldn't be rewarded with billions in profits. It reminds me of the scam from Office Space. Take a fraction of a penny from a large number of people and nobody really cares, but that doesn't mean you deserve the money. And it's causing an enormous amount of resources and brain power to go into beating the runner up by a fraction of a millisecond that could be going to something more productive.
Why is the liquidity justification pure rubbish? That's what human traders do as well. HFT doesn't "take" money from anyone. Its all voluntary sales and purchases. They don't "deserve" it any more or less than any other middle man.
>Why is the liquidity justification pure rubbish? That's what human traders do as well.
Human traders hold securities for more than fractions of a second. If I want to sell my shares of some low volume stock today and some other human investor wants to buy them next month, a third party who is willing to buy them from me today and sell them to the other buyer in a month is providing a useful service -- I get paid today, he gets the shares at a slight discount and makes a profit a month from now. That clearly doesn't apply when a high frequency trader buys shares and turns them around in fractions of a second.
>They don't "deserve" it any more or less than any other middle man.
A middle man deserves the profit he earns by providing value. What value is provided by a middle man engaged in naked arbitrage by taking advantage of an ephemeral information asymmetry?
There needs to be a name for the fallacy where people discuss trades as if prices could only move in one direction. That, obviously, is the only way you can frame a discussion of trading in terms of "discounts". In reality, an entity that takes a position in an instrument also bears the downside risk for that instrument. Averaged out, it is exactly as likely that an HFT shields an unwitting potential buyer from a loss as it is that they snipe away the profits from that sale.
The value provided by "middle men", be they robots or crooked traders in pits, is liquidity and price discovery. Trading allows people holding an instrument to unload it more quickly rather than waiting to convince themselves to take a bigger leap on pricing; it's what, in a simplifier Bohr-model kind of way of looking at stock trading, allows you to buy shares in a stock at a market price you can look up online as opposed to waiting for the stars to align. The faster the trading, the lower the spread, the smaller the leap sellers and buyers have to take.
It's worth noting again as always that before automated and program trading, spreads were higher, and the space in the spread between the bid and the ask was basically a license to steal.
Here's a good comment from awhile ago, including a cite to an accessible paper:
>Averaged out, it is exactly as likely that an HFT shields an unwitting potential buyer from a loss as it is that they snipe away the profits from that sale.
How can this be true consistent with HFT being profitable?
> The faster the trading, the lower the spread, the smaller the leap sellers and buyers have to take.
That's the theory. The issue is that when you get into sub-millisecond trades, you're so far into diminishing returns that the profits the HFTers are taking exceed the benefit of any theoretical increase in liquidity.
Also, the paper link in that post is truncated, but I found it on Google and (at least from the abstract) it doesn't seem to support what you're saying. The paper says HFT has caused spreads to go down because in order to mitigate its effects, third parties are breaking larger trades into smaller ones. But that only reduces the spread on paper by increasing the number of transactions.
HFT market makers profit from the spread the same way human market makers do. The difference is that HFT market makers narrow the spread sharply; they are fighting over a much smaller slice of the pie than humans do.
I'm not sure what you mean by your second point. Sellers and buyers are always breaking large blocks into small blocks. Moving large blocks of tradable instruments at the best price is one of the fundamental technical problems of trading. The spread captures what at any moment you would need to give to middlemen to transact on the market. Smaller is better no matter what the reason.
The article mentions that one of the conditions that excludes a company from being a non-practicing entity is "the plaintiff is an inventor or original assignee."
I'm no expert in the area but I would guess that the primary thing they are licensing is a particular design or instruction set, legally protected by copyright, not patents.
I'm sure some patent licensing has to occur as part of any licensing for the protection of the licensee but I don't think the patents are the primary 'thing' being licensed, its the copyrighted material.
I didn't say anything about patent validity, only that it's arguable ARM does not meet some possible definitions of NPE. I admit that the reference to trolls buying patents near the end of my second sentence slightly muddies that intended meaning.
A patent troll, then, could argue its "practise" is "developing" undervalued patents by buying them off investors, who receive the value up-front, and then licensing them.
ARM is, for all intents and purposes, a non-practising entity. It's just not a patent troll. The folly is in assuming all NPEs are patent trolls. Carving the ARMs out of the NPEs will take more thought that lobbing exceptions at a general idea. Minimum viable products don't work in law because switching costs are high.
> ARM is, for all intents and purposes, a non-practising entity.
I'm fairly certain ARM employs more technologists than lawyers. They also make actual products, and even if the patent system disappeared tomorrow, designing processors would still be a viable business. In this way they differ from trolls, who make nothing of value but are merely parasitic on the productive economy.
You're using trolls and NPE's interchangeably. When an NPE buys patents from the inventor, that is a contribution to the productive economy. The possibility of a buy-out drives productive activity (this is the same as startups who hope to get bought out by Google, etc).
In the end a judge will make a judgement, which means they can take actual intent of the rules of what makes a company a non-practicing entity into account. Your point about high switching-costs is valid, but the law is not a blind rules engine.
I really don't get the "ARM is an NOE" thinking. They spend a lot of time and money developing the product they license. If ARM is an NOE then so is every software company.
Maybe one could look at how many customers elect to licence the patents vs how many are litigated into being customers. Or look at the availability of the patents for license. If a company holds a lot of patents, has no technical staff to develop products, and doesn't have sales staff to try to license their IP in any markets, we might conclude the company is a non-practicing entity whose only purpose is the litigate their way to customers.
