Hacker News new | comments | ask | show | jobs | submit login
Supreme Court Should Say No to Patents That Take Old Ideas Away from the Public (eff.org)
310 points by otp124 5 months ago | hide | past | web | favorite | 68 comments

> EFF and R Street’s brief emphasizes the need for the Supreme Court to confirm what should be uncontroversial: to be patentable, inventions must be new.

New is not good enough. In theory patents must be novel and this should be applied to practice as well. The difference between new and novel is the degree of originality. New can mean taking an existing idea and putting a fresh spin on it. Novel means the idea never existed before and the current invention is original opposed to a modification or extension. Normally novelty is documented against an inception date, but inception dates are hard to qualify in court.

Also, while on this subject patents should be limited to inventions deliberately excluding discoveries. An invention is something a person made or formed with some degree of deliberation. A discovery is something that already existed and that a person merely found for the first time. This distinction would invalidate attempts to protect genetic sequences and organic chemicals present in nature. It would continue to allow protection of drugs that are more than merely chemical isolations.

So this new version of patent trolling is to find old valuable patents that can be invalidated in some way, and patent them again, then sue everyone in the world that "infringes" because they thought the patent was long expired? Wow, that is a new level of scumbaggery.

That’s not at all what Ariosa stands for.

Not new. Turbotax has long lobbied the government to make the tax law convoluted so you have to use Turbotax (or another similarsoftware) so you can avoid fines for submitting the wrong amount of taxes.

I don't feel like thats quite the same thing but I also hate it.

That has nothing to do with patents and/or patent trolling. You don't need to pay patent royalties to file your taxes.. (for example)

You're correct that it doesn't have to do with patents. But the point is about abusing the legal system to extract value from others. What Turbotax does is worse. So it is not a "new level" of abuse of the legal system.

Both are arguably forms of rent-seeking.

I think the EFF's note is a bit confusing, because it glosses over a key nuance in the evaluation.

Here's a bit more information about the case: https://patentlyo.com/patent/2018/07/diagnostics-provisional...

Briefly: Granted patents count as prior art from their filing date. The question is what happens to provisional filings. The court has split the treatment of those documents depending on whether or not the prior art is claimed or not.

The Federal court's position is interesting, as it is a compromise position. It recognizes that the documents are not published to the public, and accordingly afford the opportunity to create so-called submarine rights, and limits the scope of these rights to the specific 'stuff' of the patent in question, while preventing the non-public remainder of the filings from popping up and cutting prospective patents down.

What's the right course of action here? Well, that heavily depends on the behavior of filers across an industry.

It is, however, clear that the federal court's position is not borne out of a consistent, principled approach. For some, that's enough for them to claim the baby needs to be tossed out without trying to address the threat of the submarines directly.

Is EFF mischaracterizing the case (which it often does when blogging about on patent cases)?

Is it true that provisional applications are not disclosed to the public? It seems fair that if the invention is never publicly disclosed (only mentioned in private filings), that the invention isn't covered by patent law. That's standard -- only public information is part of the patent system.

When a patent or patent application is published, all the applications it claims priority to (e.g., provisional applications) become open for public inspection.

However, patent applications can act as prior art as of the day they are filed, which is often well before they are published. The justification for this is that, even though the information is not yet public, the applicant has established that they know that information and that it will be disclosed to the public when the patent publishes.

The question in Ariosa Diagnostics v. Illumina is whether the provisional application can also act as prior art as of its filing date. The court's decision was that only the material from the provisional application that was later claimed in the published patent counts as prior art as of the filing date. Everything else in the provisional application only counts as of the publication date.

I recently watched the documentary The Patent Scam with the XPlane guy... it was disturbing to say the least. My question is this... With "method" patents being an absolute joke, what would stop me from patenting, "A method of acquiring patents for the sole purpose of litigating infringement without actually creating a licensable alternative"? I'm serious, these patents seems so stupid, I don't see why you couldn't just get that patent and sue all of these people?

Then again, if it is patentable, I'm sure some troll owns that one too.

Good documentary, though. It's currently on Hulu.

> what would stop me from patenting, "A method of acquiring patents for the sole purpose of litigating infringement without actually creating a licensable alternative"?

Absolutely nothing other than your willingness to pursue it and pay the fees.

As an exercise, I once applied for a patent on a device that would violate the laws of physics (specifically allowing faster-than-light communications) to test the theory that you can in fact patent anything. The result was receiving U.S. patent number 7126691.


That's fascinating because in my talk with a patent lawyer, they have a very special set of rules for "perpetual motion machines" that make them quite easy to reject.

Yes, there are a few crackpot ideas for which there are special rules, including perpetual motion. But anything else is fair game, and there is a fairly foolproof process by which you can get a patent on anything. It goes more or less like this:

1. Draft the patent application. You don't even need to do any research. The PTO will do that for you.

2. Wait for the application to be rejected, which it almost certainly will be the first time around. But the rejection is required to contain specific reasons for the rejection, which will almost always be citations of prior patents that already cover what you have claimed (because there is no shortage of stupid patents). In effect, the PTO does your homework for you.

3. Tweak the application to get around the specific objections the PTO cited and re-submit.

4. Iterate if necessary. But I have never had to go more than one round, and I have eight patents.

That is just... appalling.

Yes, it is.

This is pretty much how it’s done. But if you don’t tweak (or amend in patent terms of art) well you end up with a pretty worthless patent.

You also can only tweak with regard to disclosures already described in your application.

Yes, that's true. You do have to plan ahead a little bit. But other than that getting a patent is pretty much a no-brainer under the current regime.

I really wish I could remember the sci fi book I read this in, but basically there was a guy who forewent currency, instead inventing things, patenting them, then giving them for free to people. In return the oss community supported him.

But he'd also do this thing where he'd ddos via patent lawsuit evil companies, filing like hundreds of thousands of lawsuits against a single company in the span of a day.

Sounds like Accelerando by Charlie Stross.

The hero we need

Funny, I recall seeing a post on HN that such a patent did in fact exist. From the NPR [1]:

> It turns out, a couple giant companies — IBM and Halliburton — have been working for years to patent what patent trolls do.

[1] https://www.npr.org/sections/money/2012/08/01/157743897/can-...

> With "method" patents being an absolute joke, what would stop me from patenting, "A method of acquiring patents for the sole purpose of litigating infringement without actually creating a licensable alternative"?

I believe that IBM actually has a patent on this, to whip out in case a troll decides to step to Big Blue.

Prior art

Sounds like a "covered business method", not patentable.

The entire patent system needs to be redesigned. Lawyers always seem terrified at the prospect of throwing away decades of case law but in this case, considering the nature of what patents represent and how they are used has changed so much from their original purpose, it’s probably necessary.

Of course, big business incumbents don’t mind the system because they have corporate lawyer teams that understand how it works and how to play it, so I don’t have any hope that it will ever become a political issue

It may balance out over time, but big businesses spend an enormous amount of money defending themselves against patent suits and defensively filing patents. Patent trolls are the only types of businesses that purely see upside from patents.

They also benefit from the stability of operating within a pretty known system and the fact that they are able to afford legal expertise. Big businesses may get hurt by patent trolls, but I’d imagine that this is less strategically important to their business than using their patents to prevent competitors from starting up or from taking away their business.

Small businesses get hurt the most by the patent system. A patent troll can bankrupt them or make them unfundable, big business can squash them (perhaps offensively, eg a big business takes a small business to court even though they can’t win, so the small business can’t raise funding and has to pay tons in legal costs), and they still have to pay lawyers for advice on how to protect themselves even if nobody goes after them.

No doubt it hurts small business the worst, but I'd not agree with the notion that all big businesses net gain from this. Lawyer time isn't cheap, and the bigger you are, the more attention you gather. And of course, patent trolls optimize to use the most lawyer time possible, to make it less efficient to defend (sometimes, it legitimately isn't worth it.)

Basically abolishing patents would definitely receive pushback from bigger and even smaller companies, but improving case law to be more reasonable really wouldn't.

If you're big enough you have teams of lawyers on staff. It costs you literally nothing extra. If you're not big enough to have staff lawyers you're not big. Small businesses are pretty much everyone else. You can't polish a turd, and this system as well as the copyright system are a giant turd that's antithetical to the original intent of the law.

The lawyers they have on staff wouldn't otherwise be shooting pool and drinking beer, they'd be doing useful things for the company. There's a big (opportunity) cost.

Plus lawyers are not all interchangeable, just because you have people who know employment law or how to M&A didn't mean they are good at fighting patents on court. AFAIK they're usually externally firms, not in-house lawyers

You might instead want to make a looser pays rule standard for patent suits, due to their special notions for who has standing.

You're right, but several important business models stop working out when patents disappear, most notably drug researchers.

> The entire patent system needs to be redesigned.

That doesn't necessarily mean removing it altogether. Making more thing unpatentable (ex algorithms, discoveries, etc.) and shortening the lifetimes of patents could go a long way.

That only balances it for those big businesses. For new firms the balance is quite a lot against them.

forcing upstart competitors to spend potentially billions fending off patent trolls is a good barrier to entry

This post is really quite misleading. Better coverage here: https://patentlyo.com/patent/2018/07/diagnostics-provisional...

The EFF writeup makes it sound like material appearing only in the specification of an earlier published patent does not count as prior art, that it must appear in the claims. That’s not true in the general case:

> In the simple case outlined above, the courts all agree that the disclosures found in an issued patent or published application count as prior art as of the patent’s filing date.

Ariosa addresses a very different issue. Generally, patent applications are prior art even before they are publicly disclosed—they count as prior art starting from the day of filing. Ariosa addresses what happens when you have a provisional patent application:

> In this case, the Federal Circuit ruled that a published application can count as prior art as of its provisional filing date — but only as to features actually claimed in the application. According to the court, features disclosed in the provisional but not claimed in the published application will only be prior art as of their date of public disclosure.

The EFF writeup is in fact completely inaccurate, because it starts by talking about a “published” patent application. Once the application is published, all agree that it is prior art. This case is about what happens between the date if filing of a provisional, and the date of publication.

I haven't seen anything good about patents in years. Maybe the government should just get rid of them, along with copyright. The idea of granting someone a monopoly to an idea is crazy to me.

I agree about software patents. But I disagree about all patents, and I especially disagree about getting rid of copyright. I think copyright needs to be reigned in, the DMCA is horrid, but the concept itself generally protects artists and creators (and gets abused by the likes of Disney, etc.).

If I make something, I should be capable of earning money with it if I want to. Copyright protects software in that regard, as well as music and images, etc.

There's a lot of evidence that copyright doesn't actually help most creators, and significantly holds back progress.


The same arguments generally apply to patents. There are monetization models like live shows, subscriptions, Patreon, advertising, sponsorships, etc that work very well for creators.

Especially in the internet age, we shouldn't act like copyright and patents do anything to stop people. Creators would likely do better if the publishers weren't in the middle controlling access to their works.

I can say with confidence that >90% of new small-molecule pharmaceuticals would not be created in a patent-free world. Now, not every new small molecule drug is important, in fact probably the majority aren’t. But those that are, save a lot of lives.

Counterpoint - it would drastically reduce overprescription of ineffective medications that have strong marketing

Not sure why it would. Companies would still market the drugs they had.

The market for pharmaceuticals is already very far from being a free market. It's a heavily regulated industry with extremely high compliance costs. In this specific case, further market distortions in the form of patent monopolies might be necessary to make the market work. Most industries do not have this level of regulation.

The economic harm to society that comes from patents is more than enough to fund public research that could employ the same smart people to discover the same medicines.

Not saying you're wrong, but the implication that such public research would happen assumes facts not in evidence.

Eh, we're in fantasy-land already if we're talking elimination of patent law.

If you want to make money, offer it to people who consider your work valuable. Establish a patronage and paywall yourself. Those of us who believe in decency and social good will continue to create and distribute for free.

You know, those of us with sense about this need to start working together and lobby the government actively to demolish patents, copyrights, and trademarks. Ideas need to be free for society to grow.

Perhaps running for office would be more effective.

Patents for software are fairly nonsense. Patents for drugs, for example, are absolutely crucial to incentivizing complex drug development that may take billions of dollars of research to create something that can be copied easily without patent protection.

You can't get rid of those types of patents unless you also start either handing out massive research subsidies or perform research with government money.

> ... unless you also start either handing out massive research subsidies or perform research with government money.

Which might be a good idea if the goal is to develop vaccines or new, effective antibiotics - things which are desperately needed but still do not seem to produce a strong enough incentive even with patents.

The cell phone I’m typing this on is fueled by two patent-powered business models: ARM’s and cellular companies’. Also, a substantial fraction of all modern computing paradigms came from PARC, which was fueled by Xerox’s patent monopoly.

Because it is crazy. Ideas aren't patentable, nor are they copyrightable. You haven't seen anything good about patents in years? Look around you, do you like the objects you see? Many of them exist because of patents.

Not true. Patents play very little role in consumer products. Large corporations hold patents to defend themselves against other large corporations, just so they can compete against each other. Think of the Samsung-Apple fiasco.

Wouldn't an invalid patent still be considered prior art?

Not by the USPTO's peculiar definition of what constitutes prior art.


That is not at all clear to me, on two grounds. Indeed 35 U.S.C. 102(a) does say that a prior patent or non-recent application (which would naturally be the case for an issued patent, whether the patent was invalidated or not -- it clearly had an application) count as prior art. However if the patent is invalidated does it count as issued? I'm not sure under _Stalego_ that it does. As for the applications, regardless of the text of the US code, also under _Stalgeo_ cancelled matter in the application (but weirdly, not in the wrapper!) is inadmissible, and presumably the subject of the application is itself cancelled, per code, rather than abandoned.

Patent law is super arcane and I do not practice in that area specifically but I would be surprised if the people who down voted my comment (to which you are replying) are really steeped in the peculiar minutiae of the operation of the PTO.

So Invalid patents are prior art, but content in valid patents outside of the claims section is NOT prior art?

... and if in doubt you'd better call Saul!

I thought that the reason for the patent system was that otherwise people would keep their ideas secret, and society would not benefit. The very word "patent" means "open" --- like when you say "it's patently obvious" it means "it's openly obvious" (it's like really, really obvious). So a patent is meant to bring an idea into the open. But if the idea is already in the open . . . ?

This is basically the result of a poorly drafted statute. More discussion (and less polemic) here: https://patentlyo.com/patent/2018/07/diagnostics-provisional...

The Ariosa decision involves whether an invention could be said to be publicly disclosed on the date of a provisional patent application. A provisional patent application is unpublished, although, if it ultimately supports a patent claim, it is made available to public. Details of the court case are better explained by Dennis Crouch in the cited article.

Anyone have a link to the Fed Circuit's Ariosa opinion? I could only find their amicus brief at the EFF link.

Applications are open for YC Summer 2019

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact