All right, so we all agree that software patents are a problem. So what do we do?
A). Abolish software patents. Existing software patents, or parts of patents that govern software, are nullified.
B). ???
Personally, I'm fine with A, but I would like to hear a really good B. Anyone?
Bonus points: Provide a list of ~5 software patents that are "good" software patents. Something were granting the patent led to innovation, or where the technology would not have been developed without the existence of software patents. Actually, I would be happy with just one patent, so I know what it looks like.
Intellectual "Property Tax". Have everyone declare the value of their intellectual "property" (patents, copyrights, trademarks) - each and every item, for that year, on their tax return, and have them pay 1% of the value as "IP tax".
That amount is what one pays for a compulsory license or if successfully sued, and up to 3 times that for willful infringement, per year -- and no more. (But of course, patent owner can always negotiate a lower payment)
All of a sudden, everyone has an incentive to state a reasonable value for their patent. Copyright catalogs that are not being published (old music recordings, old books, old movies) would be assigned 0 value by copyright holder, to avoid tax - which means anyone can freely make a copy. If they believe -- at the end of the year -- that someone is making a profit at their expense, they can set the value as high as they want at the end of that year, pay the tax, and sue the profiteer.
1) Patents are monopolies to begin with. A monopoly can generally pass its costs on to the consumer. So the companies now being harassed by intellectual property leeches would just feel a bigger bits
2) The government getting revenue from patents would be an incentive for the government to expand patents.
3) It would legitimize the patents even further.
4) It would be a quick way for a company to claim a huge value. Even if it cost the company some taxes, showing a return that said "10 million dollars in intellectual property" would be a fabulous way to convince a naive investor you really had something worth 10 million dollars - that they could buy for a cheap only 1 million dollars.
...
I could probably go on. File under "the seductive lure of perverse incentives..."
From the time-of-day you have replied, I have to assume you are in Europe or the UK, whereas this discussion is about the US (if you are in the US, I would have to assume you were misinformed or trolling), so ...
1) Monopolies can pass costs only as so far as they have become essential monopolies - like the power company, Microsoft, US health care or phone companies. But 99.9999% of patents are monopolies on _inessential_ things, which the customer can just avoid if not competitively priced, so they CANNOT pass it on. New cheese making patent may give monopoly on a new cheese kind, but if not competitively priced would keep customers with their old Camembert.
2) that has already happened, that's why in the US, the patent office is already a rubber-stamp, patent examiners have 8 hours to review a patent, and are encouraged to accept the patent rather than reject it.
3) Patents are at the top of the legitimacy scale already in the US. Unfortunately. Which is what Cuban (the original article) is talking about.
4) Well, patents don't do that any differently than anything else. A company could just as well buy a house for $100K, claim it is worth $100M, pay property taxes to the local council, and claim "huge value". If someone is going to be this fraudulent, it doesn't matter if they do it with "house", "patent", "copyright", "trade secret", or any other "property".
You could probably go on, but nothing you mentioned is relevant to the discussion of patents in the US (which is what the original article and my proposal is about), and very little (3, maybe) has any bearing on that discussion in Europe.
You begin making a big about you (false) assumption concerning my location. That might be clue you could take a look at the world outside the one construct in your head. In any case, considering that your solution won't happening in any industrialized country, I most looked at this abstract proposal as a chance demonstrate how the intentions of would-be tax-creators often go awry. As to your arguments, the fact that we are having this debate shows that legitimizing patents is still a challenge in the US and it is Econ 101 that monopolies pass-on price increases.
Your initial false assumption might be a clue to your excessive in your own mental constructions.
1 - The point is that software are specifically on things that on can't easily work because they are so. And they sue the companies after their software is implemented.
2-3 - The fact that we're having this debate shows that the question is still in play
4 -
....
Considering your solution will happen in any case, my point were mostly concerning
Ok, so you say you're not in Europe, and you say you're not a troll. So I would have to assume that you are misinformed (the other alternative based on the coherence of _this_ post, is that you are drunk).
> the fact that we are having this debate shows that legitimizing patents is still a challenge in the US
No. Hackernews and reddit are two places that don't consider patents legitimate. Outside of those two, there are fewer things in the US at large that are considered MORE legitimate than patents. (If we had a discussion here discussing how evil Zebras are, would that indicate Zebra innocence is still a challenge in the US?)
> it is Econ 101 that monopolies pass-on price increases.
No. Apparently, you've failed Econ 101. There's a huge difference between essential monopolies and inessential monopolies - if I open a banana stand, I have a monopoly over selling bananas in that stand. Does that mean I can pass on any price increases? Yes, if everyone needs a banana AND there is no alternative supplier close enough (making me an essential monopoly). Otherwise, no. Patents are mostly for very limited and inessential monopolies, and they therefore CANNOT do that.
> The point is that software are specifically on things that on can't easily work because they are so. And they sue the companies after their software is implemented.
I have no idea what you were trying to say here.
The important thing to note is that the baseline is not a patent-free world. The baseline is a world in which patents are legitimate, cheap, and let the patent owner set the value (including infinite value, stopping others from using that patent) without any increase to that cheap fixed value of the patent (beyond the costs of litigation)
It sounds like a good idea, I'm not sure, however, how it would work across international borders. A physical property is clearly somewhere and that's where it gets taxed. What would stop people from "moving" their intellectual properties to "IP tax heavens"?
Exactly. Also, no customs & border enforcement against trademark / copyright / whatever infringement. Louis Vitton didn't pay IP tax ? They don't get to stop knock offs from being imported.
Align value with cost - that's the way to properly allocate resources.
So, assuming that such a system was adopted worldwide, you would have to pay a tax in every country of the world if you wanted your IP to be enforceable in all territories? Because if that's the case, owning IP will be so prohibitively expensive that only large corporations will be able to afford it. I'm not sure if that would be a good thing.
If you get value from all those countries, then why not? Nobody is forcing you to get patent in each and every one of those country, just like you don't have to right now, either.
Well, it's already prohibitively expensive to get patents all over the world, and -- People keep forgetting that -- a patent is an offensive measure - it gives you the right to sue and litigate, but not much else! If you don't have the money to litigate, your patent is already worthless. In some countries, litigation is cheap; but in the US, you need $200k-$500k available to have a chance to enforce your patents.
Applying for a multiple jurisdiction patent (in e.g. US, Europe and japan - which for example still doesn't let you enforce in Russia, Mexico and Switzerland) already costs upward of $50k
This might be true for patents, but don't forget that copyright is (mostly) free. If any band, indie developer, unsuccessful writer and amateur photographer would have to pay such large fees for simply owning their works, what would happen?
They simply declare them worth zero for the year, and pay nothing. They still own the copyright. And if 5 years down the line, they realize someone else is making a profit of it, they declare a reasonable value at the end of that year, pay the tax, and collect their fees (through court or otherwise).
I love this idea. I can't own a house or a car without paying tax on it every year. If they insist that its intellectual "property", then they'd best start treating it like property.
You don't pay any tax in relation to the fact that a house or a car is property, it's based on (depending where you live) stuff like local services, roads, etc.
If you buy a £50k rolex do you have to pay tax on it every year?
This very much depends on jurisdiction. Certainly in California more expensive cars, which use the same roads, pay more tax. Similarly with property, where a more expensive property (perhaps with one by the harbour with the view, which has the same amount of garbage etc. pays more property taxes.
Certainly in say Sydney, property tax (goes to state government revenue, only paid by large land owners) and council rates (pays for garbage collection, paid by everyone) are separate.
Here in the UK council tax is tied directly to the value of the property too - it's not because it needs more services, just an extension of the logic behind income tax, that people who can afford to pay more, pay more.
My point remains that you can't tie a link between "if it is property, it gets taxed".
But much like local services, IP needs government services to function -- it's the customs service that stop cheap knockoff imports, it's the legal system that makes others stop using your patent or levy the rent you as a patent/copyright owner seek. (And eventually, it's the police force and prison system that underwrites the legal system's power).
So it's really no different than other "taxes for services" arrangements.
> regardless of whether IP should be taxed, the argument for taxing it is not that all property is taxed - because it isn't.
That's true. But hard-to-hide expensive stuff (houses, cars, land) IS taxed because it is hard-to-hide and expensive, and patents could equally well have the same fate.
It is no worse than the existing system. If you are not aware, filing a patent costs a few thousand dollars, and after receiving a patent, you have to file "maintenance fees" which easily total tens of thousands through the lifetime of a patent.
Same incentive exists today to grant patents (government makes much less money from rejected patent than from granted patent), and to prevent them from expiring (collect more maintenance fees).
Hell, the Sonny Bono Copyright Term Extension Act (affectionately known as "The Mickey Mouse Protection Act") which extended copyright retroactively from 70 to 95 years showed that congress is all too willing to do that to help their friends in the media industry, without even requesting payments; it is definitely NOT the lack of "intellectual property tax" that stops congress from increasing the term of patents (or copyrights).
Existing maintenance and filing fees clearly have not been enough to prevent large numbers of frivolous patents. Any IP tax would have to be quite a bit larger than fees existing now to be effective.
I read somewhere that although Russia has serious problems with alcoholism, public programs trying to fight it have been handicapped because a large part of govt. revenues come from a liqour tax. I think it is at least worth considering this outcome before creating any kind of 'sin tax'.
> it gives the government an incentive to grant patents and to prevent them from expiring.
as a downside to IP tax, when said downside already exists with said system, and is not made worse by my proposal. Then you say:
> Existing maintenance and filing fees clearly have not been enough to prevent large numbers of frivolous patents.
But this does not support your earlier assertion - the existing fees actually _support_ a large number of frivolous patents, because the value of a patent is (potentially) infinite, whereas the cost is known, and not prohibitive. The existing fees were NOT designed to curb patents -- they were designed to make it profitable for the government.
> Any IP tax would have to be quite a bit larger than fees existing now to be effective."
My proposal addresses this perfectly. It puts the "cost of carry" of a patent in direct proportion to its value. It is not a "sin tax", it is a "use tax" - if you use the legal system (courts, customs) to protect your profits, you pay according to these profits.
The end result is likely to be much fewer well written patents which are non-trivial and (relatively) easy to defend in court, which will be strategically selected by the grantees. Furthermore, it puts older and newer patents on equal grounds, unlike other suggestions of "from now on ..."
It is probably possible to balance the filing fees (because of fewer patents) and the IP tax so that it doesn't reduce the government intake; This might not actually be at 1%/year, but rather at 2%/year or 0.5%/year (or some other number), so the liquor thing might not be relevant -- although I agree, in general, that the government is itself a "fee / tax" junkee and therefore cannot be trusted to do the right thing.
I like the idea. One problem would be spiking fees. Company A licenses a patent from company B for the yearly fee of $1 per instance. Company A builds a successful product with 100k units sold. The next year, Company B decides to declare a 10x value on the patent. Company A, committed to the technology, either has to pay it or start over.
I guess this could be solved by long term lock ins at specific fees.
Exactly. If you build on someone else's (intellectual) property, you better have an agreement with them. However, the ip tax would make it harder to claim an idea is worth $400,000,000 (eolas, i4i) unless it makes sense, because maintaining that claim would cost $4M a year.
I think a reasonable "B" solution would be to require that any company engaging in patent litigation would have to show that they had actually been using the patent in a product; that their product had a record of sales and use in commerce; and that the allegedly infringing party was directly affecting their business.
Of course, this alternative would have to be specified in much greater detail to be effective, and have safeguards against further gaming of the system.
But it's a reasonable start, particularly as the patent trolls are buying simply to sue, not to produce a useful invention for sale.
I like this idea. We've grown too attached to the notion of "intellectual property", thinking of ideas as things you can own. We should get back to the purpose of IP: subsidizing the creation of new information, which by nature belongs to no one but benefits everyone, with a limited monopoly on its reproduction. If you aren't using your monopoly, then no one is benefiting, and you shouldn't have it.
I keep thinking about the case where someone has a valid idea for an invention that's worth a patent, but doesn't have the funds to build such a thing.
Let's say I invent a new kind of power generation that requires construction of a large-scale power plant. I shouldn't have to wait until the power plant is built and operational before I get protection for the invention. I should be able to get a patent on the idea and sell the rights to implement it to someone who can.
That's a great concern, but I don't think we're talking about requiring an exemplar to patent so much as a "use it or lose it" system.
And "use" doesn't have to mean implement. In your case, you'd just tell the truth: "I've been making a good-faith effort to license this generator technology to power companies, but they keep turning me down because they're waiting for my patent to expire instead so they get it for free." If you can document that effort, there's no reason we couldn't consider that a valid use and sustain the patent— and so the power companies have no incentive not to license your technology, and so the generator gets built and everyone wins.
Remember also that the only case this comes up in is if a power company sues you to get your technology for free on the grounds that you aren't using it. If you offered to sell it to them first, that's a pretty open and shut case.
With respect, I think this is exactly the case people have legitimate worries about, but also the case that we must clearly refute. Why should this idea be protected? Does it benefit society? I think the answer is yes only if one person having the idea makes it get implemented significantly more quickly than it otherwise would have. But I doubt that is usually the case. It's far mor likely that either a) the idea isn't really very novel and would have been produced soon after anyway by someone else or b) someone having a patent on the idea makes those with the ability to execute on it (or the capital) do something else. Or both. Patents should only exist to benefit society, since they have the force of law. I think most of us struggle to find even single instances where that has happened.
That really doesn't address the problem at all. Legitimate technology companies would still try to litigate each other out of the market with bogus patents. They would just have to make use of the bogus patent in one of their own products, which would be trivial. In fact, even the patent trolls could trivially productize their patents, or make a deal with a real tech company to do it.
The intended purpose of patents is largely to protect your idea while you develop it into a product. By the time the product is on the market, it's often too late to patent it. For example, drug companies often do 7-8 years of research between patenting a drug and releasing it.
B) Would probably best be a handful of small incremental changes. My list: shortening the lifetime of patent exclusivity, raising the patent application fees and using that money to staff the patent office to the point where they don't grant bullshit patents, require disclosure of who has been sued/threatened so that defendants can combine legal efforts, require more disclosure about who owns what patents, so you're not sued from an empty office in Texas, allow for public oversight to contest patents on the grounds of prior art.
These (small) changes have a much better chance of being made than abolishing software patents altogether, and would help reduce some of the abuses that are out there today.
Something I would add to this list: require that patents be actively enforced in order to remain valid, similar to copyright and trademarks. This would prevent submarine patents.
You're mistaken. There is no certainly such "abandonment" clause under international copyright law (which binds most countries), and I doubt any individual country as such a law either.
There are countless examples where copyright has only been defended in specific cases long after the copyright infringement was known of. Notable examples include the song "Happy Birthday" and the Amiga Kickstart ROM.
Feel free to provide any evidence for your opinion if you believe otherwise though.
You certainly don't lose the copyright, but you may lose the ability to sue in situations similar to the one you ignored - because by not suing you gave implied consent to what was going on (at least in the US and I think Britain).
Really, we should be free to invent. Option B should be that if an inventor can show a clean history of evolving an idea, it should be allowed to freely enter the market.
Ideas evolve over time, and the originate from something. Perhaps if inventors were willing to register their invention documents through a set of steps mapping to something like the Technology Readiness Level (http://en.wikipedia.org/wiki/Technology_readiness_level), then they could be eligible for immunity against patent based litigation.
But lets say this certified patent wants to protect its IP on the market through litigation. In this case, the inventor could first request certification of the invention in question before any further legal action occurred.
If we could create a certification process that is less than $2000 or so (about the cost of a provisional patent including attorney fees), it seems feasible. This system would seek to encourage a regulated invention process that filters new inventions from even being eligible for litigation. This allows for inventor insurance in a sense - keeping a capped cost for protecting certifiably original ideas while encouraging original invention and discouraging the incentive behind patent lawsuits.
1) What stops me from "cleanly inventing" something that is already patented? There's no way to prove that I derived my work from the patent.
2) How do I know which things need to be certified as inventions? Software patents are so vague that even simple applications can infringe hundreds of them. iOS apps are being sued just for having in-app upgrades.
3) Anything that requires more beurocracy or litigation is only going to make things worse because small players can't even afford to make use of the system, especially when they are fighting against dedicated full-time lawyers. The only alternative to solving that underlying problem is eliminating the ways it can be exploited.
1) Keep logs of any and all materials read, communications, etc. Disclose all sources of information involved during the various phases of invention - specifically reviewing hard drives, mobile and other storage. The inventor should take care to keep logs of all research. In original invention, the trail from research to the idea should be evident if it meets the right patterns - I'm not sure exactly what these patterns would be, but I do think there is a definition that is meaningful enough to considerably aid the situation.
2) Eligible inventions would include anything that a customer pays for - as in the exchange of value designates the scope of invention from a patent enforcement perspective. From a patent defense perspective, as long as the invention logs show no use of another patent (when the attacking inventor requests certification at their own expense). This would also focus invention on creation of value instead of creation of legal IP.
3) I disagree, this proposal is an attempt to let people opt out of a broken system by certifying invention through at least one adopted methodology of original design. Again, think of it like buying insurance - every complex industry requires some sort of overhead to deal with subjective enforcement of laws. The capping of costs reduces the average overhead so much vs the current system where anyone who succeeds in the market pays a big, unpredictable legal tax.
I've worked with computer forensics investigators before - you'd be surprised how much data you can get from your harddrive and mobile devices.. If inventors took a bit of care, and good logging / storage software was available, I think the certification process could be as straightforward as applying for a provisional patent.
This is exactly the mindset that leads to the patent system we have today. You are thinking of intellectual property as an idea - its not. Meaningful intellectual property is a functioning method or product that delivers value to people (is this all IP? no, but this the scope of IP I am suggesting protecting). It takes, research, communication, testing, experimentation, failure, and repetition to truly build something that works. And yes, forensics will show the hard drive state and web history logs before, and after you were in the shower and then all the way along the long road to you building something that people actually want.
You hear of an existing product? Who cares? If you can invent something similar without referencing related documentation or code during the actual process of invention keep it out of the legal system.
What leads to the patent system we have today is greed and lawyers. If the system can be exploited to legally steal buckets of cash from successful companies, then it will happen. The mindset of the industry seems to be overwhelmingly opposed to this, but that doesn't matter as long as the law permits it.
A new kind of certification for complex and novel inventions will not solve anything, because it's the simple and obvious inventions that everyone depends on that expose us to patent attacks.
Well, I started with the idea of 'We should be free to invent' - thanks for taking the time to poke holes in one proposal towards this end. I tend to agree that it would be difficult to build a practical certification system, at least with any proven system we've seen today.
Our government provides us many freedoms in the constitution, perhaps an amendment is a better proposal. That way the courts can interpret a more meaningful definition of Freedom to Invent as a way of shaping the current patent system.
While you're right that there are some practical issues with this sort of system, one should also recognize that the patent system doesn't current scale well either. You have to check the claims against every publication in existence. Since this isn't possible, this ultimately has to be fought out in court -- yet a granted patent still carries with it a presumption that the patent rights should be upheld.
One could, however, have a system where if two separate teams file patents before either are published, they should not have to license it.
Scientists do something similar, in that they have to log absolutely everything they do and publish the important parts. Biologists and psychologists are even more on the hook, since they have to demonstrate compliance with experimental ethics regulations. It's a bureaucratic hassle, but it works pretty well.
It actually doesn't work, since the only thing required for a patent is the ability for it to be executed. The patent applicant doesn't need to be the one doing it or need to be working with it. All the applicant needs is the knowledge that an expert in the industry believes it can be implemented.
My Plan B: Only one parameter is changed, everything else stays the same.
The parameter that is changed is how long the patent lasts. And that is determined by the pace of innovation ("number of patents / yr") being awarded in that category. The more patents in software the shorter the duration of patents in this category.
B) Shorten the term of patents for software. IIRC, the current term in the US is 20 years - basically the lifespan of the internet to date. Software innovation turns much faster than that. A patent term of 5 years would allow serious patent filers to capitalize on the R&D, while providing too short of a term for patent trolls to get a return on investment when acquiring mass amounts of patents.
(A) is my ideal scenario, but (B) might be to shorten patent terms for software. Maybe 5 years or so, instead of 20. Technology moves so fast that 20 years is an eternity. For instance, MP3 is old hat, but it still contains several patented technologies.
A more realistic first step might be "no new software patents". Otherwise it could be a pretty fierce fight with existing stakeholders, who would view abolition similarly to a nullification of a stock portfolio.
This is why there should be an opt-out of the patent system, or they can simply say all new companies can't be affected by older patents. It would be a pretty good transition - not perfect, but about as good as it gets.
I agree, although I'd make it broader, because SW patents is too tough to define and isn't the only place this problem exists.
No one really has a great definition for SW patent, especially since most SW patents are embodied as HW. Is the x86 architecture patentable considering one can implement it all in SW? Is signal multiplexing technology a SW patent? What about robotics in general?
I frankly don't think it can be done easily. I think it would be easier to say, "no patents, except drugs". And then carefully define drugs. I honestly don't think patents in the large are useful anymore, except in the case of drugs.
Given how past legislation has affected the language of software patents ("method, system, and computer program product"), I can now suddenly picture a patent claim defining software as a mechanism for producing alterations in mental states and optic nerve impulses.
I don't think the big tech companies would mind the nullification, would get a lot more objection from patent trolls and small companies that happen to hold a few key patents.
Was the patent really necessary, though? Wiki even mentions that it was only patentable in the US because of prior art, and likely would have been unpatentable if all extant prior art were known.
If anything, this seems like an example that argues the insufficiency of patents. If you invent something as important as RSA, you should be rewarded with a kingdom— and the information should be freely available to everyone. I'm not sure our compromise is doing the trick.
(I'm actually opposed to most software patents, so take this outlier as I do: a strange case that bears further consideration.)
This patent is definitely the kind of novel, amazing discovery that we want to reward.
Yes, the authors filed for a patent after their work was originally published [0], so they lost their international patents on a technicality. Yes, another scientist may [1] have secretly discovered the same thing, filing it away in classified documents, but by definition, "prior art" has been "made available to the public" [2].
Nonetheless, here are some important facts: first, if it weren't for this work, ecommerce would have been delayed for an unknowable amount of time. Second, if it weren't for this kind of fantastically profitable outcome, MIT (the original patent holder) would be hundreds of millions of dollars poorer.
[0] The wiki is misleading. It says "Since a paper describing the algorithm had been published in August 1977..." but it turns out that the referenced paper was on RSA itself, written by a pop science author. From the reference: "RSA was first presented to the public by Martin Gardner, in an article in Scientific American in August 1977."
[1] The wiki cites no source for this fact, so I can't compare the algorithms.
B) Remove all intellectual property laws period. Not just patents for software, but all patents, and copyrights, and trademarks.
You might think at first this would lead to chaos. I don't think it would. I think it would make the world radically better. I can show this would work if you're curious.
I think a lot of trademark law could just as well be replaced with existing law against fraud and defamation. The fake Apple stores in China which sell inferior knock-offs are defrauding their customers who bought thinking they were buying a quality Apple product but instead ended up with a product merely masquerading as an Apple product. So the customers would have a fraud action against the fake Apple stores. Also, by selling inferior "Apple" products, the fake stores are harming Apple's reputation. This would give Apple a defamation case against the stores.
Such a scheme would have the added benefit of encouraging quality products to the point that companies producing poor-quality goods could lose their trademarks to quality companies. So if Dell began producing their own line of "Apple" branded computers, Apple could only sue Dell if the Dell "Apple" computers were causing harm to Apple's reputation. If Dell's computers were actually better that Apple's computers, Apple would have no cause of action. In time, if Dell could show that Apple's computers were actually causing harm to Dell's reputation, Dell could force Apple to pay damages for using the Apple mark.
A lot of people wouldn't like Coca-cola scammers. They'd want the real Coca-cola. And these people would be willing to pay a small premium for it. So stores would have an incentive to sell the real thing. A lot of stores would probably refuse the sell the scam Coca-cola.
But a lot of other people wouldn't care, and they'd happily drink the scam Coca-cola which tastes just like the real thing, but is made by somebody else, and is cheaper.
If software patents were abolished, companies could still have protection for their software on the basis of copyright law. Copyright owners have the right to control distribution and adaptation, which would cover the legitimate intellectual property rights that companies should have over their software, and could eliminate the frivolous patent suits that are such a waste of resources. This of course would depend on what early cases would establish as a basis for what counts as copyright infringement, but hopefully it could be interpreted in a narrow enough fashion to provide protection for people's software while eliminating the threat of baseless suits.
B) Set up specialised patent courts. A central one for large complex cases. Local ones for smaller cases with limits on liabilities. Those representing clients in the central court must be qualified in patent law. Judges make out cases and set liability. The barrier to fanciful claims would be set higher.
Agreed. It is basically all government regulations that are the problem. Very few regulations exist for ethical reasons; most are there to entrench powers or are just mistakes. The way forward may be to create anonymous or pseudonymous companies, operating over the internet, and paid for in bitcoin, that cannot be stopped (at least not easily) because the real-world identities of the owners are not known. Thus, government would simply be made obsolete.
B) Create an alternative parallel system to the copyright / patent system, where companies, individuals, and other legal entities can choose to pay a fixed percentage (say 10% or 20%, perhaps depending on the industry) of their profit / income after expenses to a government fund, granting them total immunity from claims of infringement of copyright and patent law.
Creators of works and ideas then file a return of how much work went into the ideas they are selling, and a good faith description of what other works they used to produce their works. This goes into a formula to distribute the money from the government fund - taking into account both the utility of ideas and the effort required to create them.
I remember reading an article by a developer working on audio compression software, where he discussed how hard it was to write compression software without hitting one patent or another. I think without patent protection, music compression would be better that it currently is.
I really wish he had been a little clearer on what these patent trolls are and why they're bad. I'm sure most people here on HN understand what's going on, but this article is little more than preaching to the choir without that extra clarification.
As it stands right now, this article has sort of a whining feel to it. "All my companies are getting sued because we're infringing patents, and I'm sick of paying for it! The only way to stop them is to similarly arm yourself with lots of patents so you can sue someone just as hard as they can sue you!"
I'd like to see more detail from him about how most of these patents are obvious and non-innovative, and that the idea of patent troll companies who do nothing with the patents but sue for infringement of them. I think that's what he was getting at with these lines:
> I’m not talking about a new company that had an idea that someone beat us to. No sir. I’m talking about companies that have been doing business the same way for years that are getting hit by patent trolls.
...but that could also be read as "we've been infringing these patents for years, and I'm pissed off that someone noticed." We all know that's not what he meant, but there's a lot to be said for clarity, and clarity wasn't a major component of this article.
I don't know, I thought he cut right to the chase:
> Google just spent $900mm buying a patent collection. Other big companies are spending the same way. That is money that for many companies would have gone to job creation.
That's exactly right. That money is going into the pockets of $500/hour lawyers instead of creating jobs for people that produce products. Just the public patent deals of the last few months amounted to billions. And there is a lot more we don't know about.
He could have explained more about how the vast majority of software patents are dubious at best, etc., but that takes more explaining, it's complicated. Focusing on the direct link to job creation is more straightforward.
He's arguing the impact of it, but I'm saying he failed to clarify why they shouldn't have to pay for the $500/hour lawyers and the multi-million dollar patent packages.
> He could have explained more about how the vast majority of software patents are dubious at best, etc.
That's all I'm saying. He needs to explain why he shouldn't have to defend himself against patents, rather than just complaining that it's too expensive. Defending yourself against any lawsuit is expensive, but if you've done something wrong then it has to be done. Cuban just didn't take the time to explain that his companies hadn't done anything objectively wrong.
I wasn't disagreeing with him at all - I agree that there needs to be patent reform - but I just thought that he didn't adequately state his case. He complained about how expensive it was to have to deal with the lawsuits, but never made a compelling argument as to why those lawsuits were unjust.
I'm pretty sure that the number of patents you could get revoked with $2.5 billion in legal effort is much smaller than the total number of potentially problematic patents. Unfortunatly fighting bad patents is a classic public good [1] which means its hard to coordinate people on making them.
If your feeling altruistic, however, the EFF is working to help get some particularly egregious examples revoked. [2] I've donated to them and I encourage others to do the same.
"The term patent usually refers to an exclusive right granted to anyone who invents any new, useful, and non-obvious process, machine, article of manufacture, or composition of matter, or any new and useful improvement thereof, and claims that right in a formal patent application. "
In my opinion, the key part is the non-obvious part of "new, useful, and non-obvious"; the invention of a new medicine is usually non-obvious, but the invention of Amazon 1-click shopping is obvious (in hindsight).
What is "non-obvious" ? If a single individual can come up with a similar solution (without access to the underlying research/data involved in the patent) in a short amount of time (a week, or at least in a DAY in the 1-click example) it does not represent a significant investment.
There should also be a requirement that the idea is implemented to be able to sue others. You may stop others' implementation if you have a patent, but you should not be able to charge them for anything until you have implemented a working solution yourself (to avoid patent trolls).
Or simply use the KISS principle and remove all software patents once and for all !
This is more of a thought on Google's awkward situation rather than patent laws in general:
What if the DOJ scrutiny around the Nortel patent bundle went well for Google and Apple/Microsoft etc were forced to put essential patents in a pool and license it to other vendors under fair conditions? As Google's CLO said, it still does not give Google the weapon to negate Oracle or Microsoft or Apple's IP lawsuits. What if DOJ smites down the same way everytime an important patent auction is happening? As long as Microsoft/Apple etc is willing to throw money at it, the best outcome will always be everybody gets its fair share. Again, Google's patent WMD remains a dream.
Unless a patent reform retroactively grands Google all the immunity in the world I really do not understand how this would end well for Google without sorting everything out in court just as it is.
* Of course theoretically Google could buy up MMI or other companies with a giant patent trove in mobile space to balance things out. Then again all public companies' board have fiduciary duty to maximize shareholder value and as long as Apple/Microsoft are willing to pay more, Google has slim chance of exclusivity.
It already went well ..according to Google Lawyers that has already happen with MS..namely they were forced by DOJ to divest of those patents bought through that deal and the Patent holding company was forced to license to open source at reasonable rates
That were Novell patents, which were very entangled in the whole SCO/IBM/Novell/Red Hat/Microsoft legal soap opera regarding Linux/Unix licensing. Basically Microsoft was caught hand in the cookie jar, hence DOJ's interest.
Auction patent extensions in blocks. First, you start with a free, though much shorter than present, initial period. After that, however, you must purchase the next extension at auction. At period2, there will be X2% as many extensions as there are expiring patents. At period3, you must buy another extension out of that tranche, which has X3% as many extensions as period2. And so on.
Extensions could be sold to any party before their "apply on" date, they are not specific to a particular patent until it has been attached to it. Once a patent passes its expiration date, it can never be recovered and is now public domain forever.
I think this would work best if the sum of the extensions was less than currently available. The basic idea is that a company who felt a patent really was central to their business could pay a potentially very hefty price for the privilege, so the number of extensions for sale should be kept quite small.
The only point I would see for this scheme is to make money for the patent office.
> The basic idea is that a company who felt a patent really was central to their business could pay a potentially very hefty price for the privilege, so the number of extensions for sale should be kept quite small.
It won't be. There's too much money to be made and too much lobbying involved.
At this point I would be fine with starting with the baby set of killing any patent that implements a business process. It would be a good start, simple to explain, and kill off a goodly chunk of the trolls.
Sorry, I should have been more specific. My point is more that the problem is systemic. You also can't patent something that has prior art, but it happens all the time.
Is the tech industry really that influential for job growth in America? Say everything Cuban says is true, wouldn't there only be a marginal difference between today and his tomorrow? Aren't the majority of unemployed non-tech specialists?
It isn't clear to me why money lying around necessarily entails job growth in the tech industry, especially for larger software firms but even smaller ones. But again, I feel this would only be marginal contribution to the job situation were it the case.
So he makes an excellent point, especially for the long run, but I must admit the tech sector is not currently facing serious job issues. Of course it would probably be more beneficial to the economy and to society to have Google hiring employees instead of buying patents, but that's not going to save Detroit.
I think that just having requirement that a patent actually has to be in "real production/read product" by the holder is the only modification to the law we need. (there is challenge here how to ensure that production in not fake, but it could be possible to ensure that)
good points raised by cuban, but i have to say, he's offered zero solutions to the problem, just gripes. i don't think anyone's come up with a good solution yet for patents probably because there are an over-abundance of them (millions, possibly billions pending), you can't just do away with the system suddenly, hell i think i have 1 or 2 myself.
but let's just play devil's advocate and say that the government refused to grant anymore new patents and grandfathered all the current patents, you would still need something else to take its place, perhaps say the latest trend of open source licenses, though some of the issues with some of these licenses seem to be on similar bad footing of corporate abuse.
i don't have any solutions myself, but i think china's ip laws are relatively lax and a good example of what would happen if there weren't patents or at least very poor enforcement. basically you'd have a bunch of companies competing against each other possibly reverse engineering or out right stealing things, at the end of the day, you, the consumer, would possibly have a better set of choices and these companies would be innovating to compete, not necessarily strangling you financially with patent litigations.
my major gripe with patents are the same as cuban, these big corporations are using it as a new line of business/revenue stream and are throwing their weight around with the protection of another big corporation (a.k.a. the government), this a major loophole in the system, what was meant to quite possibly protect the little guy, the small company, has now backfired and become a major liability. e.g. nobody in their right mind would file a suit against ibm, they would surely find some infringement in their arsenal of patents and make you hurt for a good long time. you're suing ibm for using your algorithm, they'd turn right around and sue you for something along the lines of using the bathroom on an airplane (http://news.cnet.com/2100-1017-961803.html).
the problem is truly systemic. when governments/companies/countries become too powerful, it really becomes easy to abuse/manipulate the system, absolute power corrupts absolutely. that's why there needs to be more refinement of government to handle these things, to make sure that the system works.
Many standards bodies require reasonable and non-discriminatory licensing. One reform would be to require all patent holders to offer reasonable and non-discriminatory licensing.
The problem with the way software patents are being used is that they are stifling innovation. It is not just that lawyers are being given work at the expense of other workers in a 1-to-1 fashion. Every individual lawyer working for a patent troll who is suing a technology company is probably preventing two, three, maybe a dozen new jobs from being created. A patent lawyer looses his job, and a couple of programmers, a sales person, and maybe a corporate lawyer gain employment.
I've heard that argument and I just don't know how to respond to it. Why would you ever want to _create_ jobs. People don't need jobs. They need food and cars and what not. To my way of thinking jobs should be actively destroyed.
Ideally, what we want to create is opportunities for people to create wealth via their own labour. This is the good sort of "job creation".
The bad sort of "job creation" is where you create non-wealth-creating jobs. For instance, if the Government were to declare tomorrow that every truck driver needs a supervisor in the passenger seat to make sure he doesn't fall asleep, that would create a crapload of jobs but no wealth. Or, if you pass any new law you'll probably create a need for a bunch of new lawyers.
The problem is that "job" is synonymous with "entitlement to food, cars, and whatnot". There would be considerably less uproar over job creation if the jobless were guaranteed a house in the suburbs, an automobile, and food for a spouse and two kids.
Given the choice between
1. Being satisfied that society is moving forward in some abstract sense because your job has been rendered unnecessary.
and
2. Being reasonably sure that there will be a roof over your head and food in your children's bellies in a year and a day.
What I was asking was: if you change Cuban's argument from "If you want to see more jobs created – change patent laws" to "If you want to see more prosperity created – change patent laws", would you agree?
I'm sorry, but this is completely absurd and startlingly naive as a solution to the unemployment problem.
The limiting factor on the number of people with programming jobs is the number of people capable of filling those roles. There simply are not competent developers who are having trouble finding work right now.
If you want to talk about creating high-paying jobs that will allow the US to be more competitive and prosperous in the future, that's fine, but then you still have to address the greater problem of creating a qualified work force.
The problem isn't that the workforce can't be programmers, or other technology related discipline. Entry-level programming skills can be picked up in days by virtually anyone. The problem is that, unlike the jobs of the past, there is no force to drive companies to hire anyone but the very best.
If every single person in America knew how to program, they still wouldn't be working, because the companies would hold the position open until someone who is better at programming is available.
"I'm sorry, but this is completely absurd and startlingly naive as a solution to the unemployment problem."
"More jobs created" != total "solution to the unemployment problem". The logic as presented is sound; take away one source of needless drain, more jobs are a very likely result.
Also, the limiting factor on businesses hiring is not (just) the cash they have lying around, but the demand there exists for their goods and services.
"Another words"... Really, Cuban? At least we know this is really him speaking, because that's pretty damn embarrassing to put out to the whole world. It's a great post though. I agree 100%.
Yes, the patent syste is broken, but here is what I think is at the crux of it. What happens is that the patent office is completely overwhelmed by, not only the volume but the complexity of the patents being filed. So what they do is basically punt it to the courts. They take a cursory look at a filing and if it is not too obviously a bad patent they grant it and then let the interested parties fight it out in court. So that is what is driving the patent troll problem.
To do it right you really need the patent office to a real in depth analysis so that when the patent is finally granted it can really mean something.
Politically, that is simply not possible because it would require an expansion of the Federal Government which we all know now is somehow a horrifying thought.
the really insidious effect of this patent trolling is that it robs capital from true inovation. Instead of companies like Google and Apple spending money on hiring people to inovate and create new products, they are spending billions on defending what they already have. Very sad.
For one, patents are not only serving technology companies. Drug companies live ad die by their patents. A new drug cost that much to develop if without a patent protecting it, it would have not make any business sense to even develop new drugs.
While we are patenting every line of every software program ever written, we should also patent colors, smells, tastes, sounds, sensations, thoughts, individual words, photographs, individual syllables, and visual/body language gestures.
How come lines of code fall under the category of 'invention', But English sayings, phrases and quotes do not?
This is interesting because software is a composition in a computer language. So the OP has an interesting point there. Code and algorithms shouldn't be awarded patents just like you can't patent a musical composition.
Software itself isn't patentable -- software licenses work on the basis of the copyright on the source code, not a patent. Software patents generally describe algorithms rather than their specific implementations, just like ordinary patents describe a method of solving a problem.
Perhaps these patent deals that are putting a price on inventions (particularly at the billion dollar valuation level) will encourage companies to spend _more_ on R&D - hiring more highly skilled workers - because they can create relevant patents more cheaply than they can buy them.
The suggestion that abolishing patents will cause companies to spend that money hiring people instead is not intuitively obvious to me. It sounds a lot like trickle-down economics, which may be intellectually appealing but doesn't seem to have worked out as originally expected.
A). Abolish software patents. Existing software patents, or parts of patents that govern software, are nullified.
B). ???
Personally, I'm fine with A, but I would like to hear a really good B. Anyone?
Bonus points: Provide a list of ~5 software patents that are "good" software patents. Something were granting the patent led to innovation, or where the technology would not have been developed without the existence of software patents. Actually, I would be happy with just one patent, so I know what it looks like.