Do patents kill innovation? The US patent office is asking | Hacker News Reader

Do patents kill innovation? The US patent office is asking

eenewseurope.com

I think its too easy to get a patent, and that leads to a great proliferation of patents that serve little purpose but to slow innovation.

I worked at a patent law firm about 15 years ago in my wild youth. I’ll never forget hearing an attorney explain that getting a patent is basically bullying the examiner, and if that doesn’t work, appealing repeatedly until it’s eventually accepted.

Are there statistics of the number of patent applications rejected? Even better, I would love to see the number of patent applications filed by the Apple computer corporation, and how many were accepted, and how many were rejected.

Edit: Autocorrect error.

Yeah, the question "do they kill innovation?" is too simplistic.

They exist for a valid reason and they also have negative side effects.

The current implementation has a lot of systemic problems. The ones you mention, patent trolling and length of patents might be tuned in order to optimize innovation.

I worry about our political system's ability to perform this tuning operation though. Political discourse seems incapable of tackling complex issues like this where there is no good/evil dichotomy, simply a productive balance to be found.

I think the case needs to be made that they're valid - If you look at the steam engine, progress only took off after the patents expired.

The idea is pretty straightforward. "Little guy" invents something truly novel, investing major time into generating that invention. "Big guy" sees this after the fact, uses existing resources to steamroll little guy.

Patents exist to prevent that.

Okay that's the idea, but do they do that today? It seems that today the more common story is: "big guy" patents everything under the sun before the fact and locks it away in a drawer, while "little guy" actually builds the thing and tries to sell it but gets crushed in the process. Or at least that's the one I hear told more often, maybe it really is the case that patents are promoting innovation. But we can't just assume it's working out like we thought it would.

That's the story we hear often with software patents, but those are relatively new. There's a long history of patents for physical things which may be less obviously detrimental.

FWIW, if I remember correctly, the reason we can even have software patents is because a hedge fund ended up trying to patent their process for categorizing or rating investments, and in their drive to patent a process as opposed to a mechanism, new law was decided. I'm being vague because I don't remember the details enough to be authoritative and don't have the time to look them up, but someone else might be prompted to.

To me, I can see a benefit to physical and software patents both, but I think they have very different effects and should be handled differently. If software patents were genuinely harder to get because they had to actually be complex and novel, and also only lasted 5-10 years, I don't think we'd really have much of a problem. I don't know how to raise the bar on their complexity in a good way though.

That's the theory, but it doesn't play out in reality and never really has.

That isn't evidence that patents are bad. You also have to know the counterfactual of whether the steam engine would have been invented then if the creator wasn't going to be able to patent it

And further, even if it had been invented five times over, would it have been known about by the right people had the patent not forced a public documentation of the design?

This is a good question to which we can never know the answer. Every time this discussion happens I can't help but think that the solution is to make patents much harder to get and/or last for a shorter time.

Theoretically, without them a big company can just copy or reverse engineer your design and sell it for cheaper than you can make it. In practice, big companies just patent a bunch of vague stuff so that no matter how innovative what you come up with is it will somehow end up covered. I've seen this happen over and over.

That makes it seem like we're far, far over some threshold where patents switch from protecting inventors to preventing inventors from doing much of anything.

That's the wrong metric to look at to asses the effectiveness of the patenting system. The justification for patents is that they facilitate innovation by granting a state-enforced monopoly for a limited window of time at the cost of publicly disclosing your innovation. Thereby creating incentive to pour R&D money into new ideas, and to share those ideas with the public. Under that mission, your example is proof of the effectiveness of patents, not the ineffectiveness. Innovation stagnating on steam engines, and then exploding after the patent expires is an example of this system working as intended.

I'm saying that justification needs strong evidence and it's not there.

Exploding after a patent expires is because the state has granted a monopoly to one person. The steam engine would have been invented without the patent protection.

Look at innovation in software - it's not because of patents. Patents hamper innovation.

I understand the point you are trying to make, I'm just pointing out that the example that you were using is actually one in favour of patents not against. I'm positive that the steam engine would have been invented with or without patents, but it's certainly possible that it's invention was accelerated by strong intellectual property protections. The vast majority of inventions are the product of profit-seeking endeavors, and patents help justify the immense capital required to invent. The idea of the lone inventor building something creating something purely for the joy of it is a romantic image, but not really how invention works (certainly not at scale).

Also, the technological explosion of the steam engine was partly due to public disclosure of the invention's function as required to obtain a patent. No patenting system, no disclosure. And maybe it takes society longer to figure out how the machine actually works, thus delaying innovation. If the patenting process did not exist, perhaps the inventors go to measures to secure their intellectual property themselves by intentionally making the product's inner workings obscured. A world with no intellectual property protection is a world where everyone attempts to hide and obfuscate their innovations from the rest of society. This is not an environment where innovation thrives either.

I'm certainly not trying to make the argument here that the patenting system is flawless or even good. Only that intellectual property protections of some kind or another are, and we shouldn't throw that baby out with the bathwater.

From my understanding, it's wildly considered that patents slowed steam engine innovation significantly. http://www.dklevine.com/papers/ip.ch.1.m1004.pdf

Part of them problem is patents are "one size fits all". the one size being 20 years.

In pharmaceuticals, where getting a drug to market takes a decade and hundreds of millions in investments, it seems to work well enough. The investments wouldn't be made if the investor couldn't get return because all their hard work could just be copied.

But fast moving areas where R&D can literally happen in a garage, take months rather than years, and each step builds upon the ideas in previous steps it's not all clear how patents help, and there pretty clear examples where patents slow things down. Sewing machines were bedevilled by patent thicket, as were aircraft where in World War I the Americans were force to use European designs. The latter was a bizarre outcome given heavier than air flight was invented in the US by the Wright brothers - but them heavily patented by them.

The sewing machine thicket was "solved" by the formation of a monopoly - Singer. The Aircraft thicket was solved by the US government demanding all aircraft patents be forfeited to them. In both cases as soon as the problem created by patents was solved, innovation resumed.

It's possibly true that shorter patents would prove to be a net benefit in fast moving industries, but IMHO the evidence seems pretty clear when your innovation is likely to be superseded by something better in 5 years in a free flowing environment, being able to stall all R&D by 20 years is not at all helpful.

Steam engines are only an example of patents working as intended if there was any value in publishing the ideas - that is, if the knowledge would have remained secret without the patent.

> Yeah, the question "do they kill innovation?" is too simplistic... They exist for a valid reason

But isn't that entire reason explicitly to "promote the Progress of Science and useful Arts" aka innovation? Shouldn't we ask whether or not patents are doing the thing we want them to do?

Having a patent ensures that I have all the data I need to later improve or implement it. I've read books where the steam engine were encased in explosives so that it couldn't be reversed engineered thus ensuring the inventor had a monopoly on them (the book had a lot of other things to ensure the monopoly as well).

A lawyer told me years ago never read patents: if you infringe and claim ignorance that is better than infringing and knowing. If we want to make patents useful, then we need to change this to ??? which ensures engineers keep up on the relevant patents in their field. I have ideas on what to change, but every one has serious flaws so I'm not going to state any - maybe you can do better.

> I've read books where the steam engine were encased in explosives so that it couldn't be reversed engineered

Stop! You’ll give the HW companies ideas.

If they try they will run into attempted murder issues.

> I've read books where the steam engine were encased in explosives so that it couldn't be reversed engineered thus ensuring the inventor had a monopoly on them

That's not a argument for patents, that's a argument for felony prosecution of interference against reverse engineering.

I think it is an argument for patents. The point is that before we had legal protection of ideas, knowledge was hoarded, because it has value. Trade secrets, guilds, possibly explosions! Secrecy basically.

Patents change that equation. So while their existence slows down innovation while they are in operation, after they expire society as a whole has gained that knowledge.

I'm not arguing that the patent system isn't horribly broken in all sorts of ways. Just that they are fulfilling a purpose in not forcing inventors to go to huge lengths to keep information secret.

Well, there’s plenty of evidence of how patents have delayed innovation until after said patent lapsed, like steam engines, e-ink displays mentioned elsewhere here.

I would like to see any evidence of patents actually helping innovation increase. It should be plainly obvious that the negative side effects ARE actually the main effect.

It’s one of the few things I think the US constitution completely whiffed on, by mentioning patents at all.

> e-ink displays mentioned elsewhere here.

where? not this one ( https://news.ycombinator.com/item?id=26143779 ) I hope because I questioned it and I'm pretty sure it is false and written by someone who isn't actually even involved in the display industry. That particular post got cited by Boing Boing and multiple blogs and then subsequent HN posts started citing those blogs and Boing Boing. Basically a self-referential citation infinite loop. Please see my comment history. I work in the display industry and I'm not aware of what patent is blocking anything in developing electrophoretic displays.

Okay, my mistake. if there’s no patent there, there’s no problem there. Kinda proves my point that the absence of patents allows innovation to grow, I have not seen any evidence to the contrary; it’s usually the “but what about the little guy” defense, not mentioning any “little guy” who was helped

I'm trying to understand what you're trying to communicate but I don't get it. Could you elaborate what you're trying to say because it sounds like you're claiming patents are bad (absence of patents = innovation growth) but we know that patents arose initially at least to protect inventors. Anyway we've gone well outside my initial question which was about which eink patent.

hm, didn't see this. Yeah, I do believe that patents tend to undermine innovation pretty much everywhere, and yes, I do acknowledge that that was not the intent. The intent was to grant a temporary monopoly on an invention, to give an inventor time to 'cash-in'. Unfortunately that means it's effectively 'closed-source': no one else can make improvements to it w/o permission, and so I think a lot of innovation happens when a patent lapses. If that hasn't happened with e-ink displays, cool, then I must've misread something. It's hard to imagine e-ink displays being patent-free however.

here's a paper on it: https://pubs.aeaweb.org/doi/pdfplus/10.1257/jep.27.1.3

> The ones you mention, patent trolling and length of patents might be tuned in order to optimize innovation.

Maybe there should be at least a differentiation between tech sectors for which the patent is applied (e.g. IT vs. agriculture vs. materials vs. <all the rest>) in relation to how long the patent would be valid?

Having a patent in IT valid for XX years is in my opinion a lot more counterproductive than the same for other sectors.

I could have tonight a great idea about software and implement&test&prove it within days but to do the same with new materials/crops/engines/whatever can take a lot more time/effort/resources => maybe IT patents should always have a shorter max life than the rest? At least the ones purely related to software.

In addition to the DURATION and 'USE' (in an actively produced or utilized real world example): (PS: these should also apply to copyright)

A maximum license fee and/or scheme. Lesser values might be negotiable. A fixed rate always paid to the government per licensee.

Short initial duration, renewal requires exponentially higher fees.

Too easy from the technical side, too difficult/expensive from the legal side. Solo inventors and startups working on genuinely new and interesting technology do not have the funding to hire fancy law firms, while companies like Microsoft and IBM file a dozen patents a day (not an exaggeration). It's the exact opposite of how a patent process should work.

And a perfect example of how regulatory capture works. I don't have a solution, but I am convinced that over time this is the end result of ANY regulation. It doesn't matter what the original intent is/was.

The answer is to not grant patents.

I look it as an evolutionary system, an arms race. Capture will happen if the regulators don't keep up.

But it's possible to keep up. US financial markets have a lot of internal regulation for traders, so they're wrangling some of the most aggressive do-anything-for-a-buck people on the planet. Nonetheless, the traders are generally kept in check. I used to do tech for derivatives traders and we knew to never, ever fuck around with a market.

That didn't always stop our traders, of course. We had some based in Frankfurt who hated living there. They decided they wanted to live in London, and so quietly shifted their trading terminals to our London office. This was strictly forbidden; at least at the time, the mid-90s, trading on the Deutche Terminborse was to take place on German soil and nowhere else.

Eventually the regulators got suspicious and so they were standing outside ringing the bell while watching our trading activity. So one of my technical colleagues had to rent a van, load everything up, drive it 10 hours back to Frankfurt, and have it all working by the next morning. Which we did, and so avoided formal sanction, but it was close. That probably kept those traders from doing something foolish for at least another year.

So, the traders did something explicitly forbidden and didn't get penalized for it; that doesn't sound like being "kept in check" to me.

Never mind those manipulations that make it through the regulatory filter, a la mortgage securitization circa 2008.

It was them getting kept in check, because they quickly corrected the behavior. And as far as I know, never repeated it. Explicit punishment is not the only way policing happens. Indeed, I think the most important parts happen well before that.

> standing outside ringing the bell while watching our trading activity

Do you mean ringing manually the bell of the end-of-the-day of the stock exchange? When they rang it the first time, didn’t they catch them not going out of the building?

I mean they rang the doorbell. They could see that the traders were active, meaning that they were supposedly in the office. But the door was locked and nobody answered it.

Ya, like a bureaucratic DDoS attack.

> regulatory capture ... is the end result of ANY regulation.

All evidence to date supports this conclusion.

Can we do better?

Regulations are two-party, can be distilled down to us vs them.

How do we make them three party? Transmute them into trilemmas? (Tripods are more stable that bipods.)

Maybe grant a seat to the parties on the receiving end, the natural antagonists to the parties being regulated. And empower them appropriately.

That isn’t what regulatory capture means. Regulatory capture has to do with industry vs public rather than big co vs little co.

  Regulatory capture is an economic theory that regulatory agencies may come to be dominated by the interests they regulate and not by the public interest. The result is that the agency instead acts in ways that benefit the interests it is supposed to be regulating.

https://www.investopedia.com/terms/r/regulatory-capture.asp

In this case, the small-co or independent inventor acts as a member of the public.

It's really about established players being able to invest in controlling the regulators that are specific to them, while new players considering entering obviously don't have this advantage. Structurally it's similar to NIMBYism.

As another commenter suggested, let's look at the data from the USPTO and see what the submission/approval ratio looks like for large cap, small cap, and independents before jumping to conclusions...

Yup. A friend worked for a now-extinct tech company that strongly encouraged employees to patent as much as possible. On his desk, he had one of those drinking bird toys. [1] When the head got sufficiently wet, water would drip onto his desk. He made a little celluloid collar for the bird so that excess water would get dumped back into the cup the bird was drinking from.

On a lark, he wrote it up as patentable innovation. His boss thought it was funny and approved it. So did the lawyers. He ended up with one of the little plaques they gave employees for patents. And maybe a cash bonus? It was long enough ago that I forget. But apparently that's what IBM does: https://news.ycombinator.com/item?id=17835839

[1] https://en.wikipedia.org/wiki/Drinking_bird

I think that's because IBM has a massive portfolio and just treats it statistically in IP cross-licensing negotiations. I had a boss at a company who said they added up theirs and IBM added up theirs, then his company wrote a check to IBM every year.

IBM also plays tricks to artificially boosts the number of patents it gets. For example, for a computer-implemented invention, a patent application will generally have a set of claims for the method that a computer performs to implement the invention and a set of claims for the computing device itself configured to perform the invention. What IBM likes to do is to split those two sets of claim into separate patent applications, so one application just claims the method and another application claims the computing device. So when most other companies would only get a single patent for a computer-implemented invention, IBM would get two patents. So this helps IBM maximize licensing revenue in the example where they're just adding up the number of patents.

Every big corp makes such list, and measures against others, no matter what patents there are.

Nearly 20 years ago, in initial bootcamp in big-corp, there was half a day course by a patent guy, and one of the warnings he gave was "be careful where you search for your new idea" - the most wellknown patent search web site then, was owned by.. IBM.

I worked at a place where they had patent plaques on a big wall. Guy showed me his one day...

It was just a design for a computer network, using run of the mill off the shelf equipment doing things I'm pretty sure is done / drawn out in tons of network design books and etc. But apparently it was a 'system' or some term like that he used.

I smiled and just moved on....

Patents aren't all that expensive. Defending your rights in court, however...

Consequently writing a PTAB review resistant patent is expensive because you really can't defend a poor drafted vanity patent after the fact.

That is just for the USA, patents are extremely expensive in EU, they got a great feudal system going on there where you cannot get a patent without spending at least 40k EUR, whereas in USA a single inventor can get it as low as 500 USD.

Sure, the USPTO might charge $500, but the patent lawyers that will file the application for you will charge $40k - $200k.

You can actually do it yourself, there are lots of book out there how to navigate the patenting system. You don’t need to hire a lawyer if you spend some personal time but on the EU you don’t have that option at all.

> I think its too easy to get a patent

I'd modify that to say it's too easy for a large corporation to get a patent. I've heard quite a few acquaintances who work for large companies mention offhand that they just were awarded a patent for something or other.

Meanwhile, at my tiny company, we have I think one patent and another that got hung up with the examiner due to completely unrelated things. I think we finally stopped protesting it.

Patents are pay to play.

I worked in patents about 15-20 years ago. What I remember in this golden era of patent proliferation was that the US patent office, if in doubt or just by default, just issued the patent and waited for people with prior art to object.

Not sure of the numbers, but I remember IBM would file by the thousands every year for every little thing. Including one for the work that the company I worked for actually did that they caught wind of and decided to beat us to the punch and file. We eventually were granted it, but it was a high legal cost to show we had the idea first.

This is the root of the problem. The examiners have to meet a throughput quota and that creates a perverse incentive to approve garbage. The USPTO is also dependent on filing fees which adds another perverse incentive to help the trolls with their patent spam.

Trolls usually don't file patents. They usually acquire them, usually from a company close to going bankrupt. Then they sue other "medium-size" company in a jurisdiction with a friendly and corrupt judge.

Even if we stopped issuing new patents today, trolls have enough fodder to survive for the next 100 years.

I thought patents expired 20 years from the file date? So you could only troll for ~20 years + anything manufactured using an infringing process. (so "you owe me royalties on the widget you made 10 years ago" is valid.)

I once worked at a company that paid $1K to employees who got patents. Most employees (that I knew) never bothered, or only when they actually invented something. But this one manager and one of his reports formed a 2-person patent brainstorming club. They'd book a conference room regularly and spend the meeting thinking up worthless patents. As an example, one that I remember was something like "spellcheck in a context that usually doesn't have spellcheck." Others were nothing other than descriptions of products we were working on, such as "A web based interface for communicating such information from such a person to another type of person."

If you looked in the meeting room, you would see these two giggling like kids, slapping the table, falling out of their chairs. They each had stacks of patent awards in their offices. Most of us had none. I am happy that they had so much fun, but that is what patents are in practice.

> too easy to get a patent

If you make it harder, startups will have difficulties obtaining them.

Patents are a way for private companies to benefit from the public enforcement. Like anything provided by the state, it can and will be taken out of its goals and used by the largest corporations to reinforce their dominance.

Removing the tool would make the playing field more fair.

I think patent applications should have a "challenge period", where anybody can come up with the same or a better solution which would invalidate the patent (and perhaps the original applicant has to pay a fee to the other party).

just wait for a copilot for patents that does everything, then just a small tweak for "novelty", then just flood the patent office.

Sixteen years ago, I had the idea for headlights aimed by GPS and started to make some prototypes. Then I read a patent that was basically "a plurality of inputs + processor -> headlight aiming" that covered everything imaginable. I knew the patent wouldn't hold up in court, but I wouldn't have the money to fight it.

More recently, I've been working on a self-driving wheelchair. It's a fairly obvious straight line from self-driving car to self-driving wheelchair, but someone's gone out and patented it. I still don't have the money to fight it.

This problem would be a lot easier to deal with if independent invention was a defense (as it is supposed to be, but in practice basically never is) against a patent infringement claim--particularly if the owner of the patent has done nothing to actually implement it.

Yep, that would help. From what I can tell, the wheelchair patent is aspirational and they haven't actually implemented the majority of standard autonomous vehicle practices.

The patent office should start demanding working models for everything. You shouldn't patent something unless it actually works.

Maybe demand a working model for the patent, but also reject a patent if anybody else has something similar in development?

So you have a brilliant idea like "lets put email on a cell phone", but the infrastructure isn't in place to support it yet. A few years later you finally get it working, but now you can't patent because there are 5 other companies who started the process at basically the same time--once the prerequisite technology was in place.

This would reduce the number of patents on "the most obvious way to solve this obvious problem that only appears because the technology has advanced enough" that patent trolls love.

So many of the "do X, but now on a computer" patents wouldn't be a thing.

Requiring a working prototype would also make it hard to issue vaguely worded and extremely broad patents, so someone can't claim ownership over the entire concept of sending push notifications for example.

one issue is that you might not want to start prototyping something (for cost reasons) until you have a defensible moat.

Incidentally, in the US from 1790 to 1880, supplying a working model was indeed a prerequisite for a patent application.

That's kind of similar to how copyright works - it's about whether something is "derivative work". A similar principle could apply for patents perhaps?

>independent invention was a defense (as it is supposed to be, but in practice basically never is)

It's actually not a defense at all and isn't supposed to be. Its sort of like a prize for inventing it first.

It's only a defense if you publicly disclosed it before patentee did. Secret use was a defense if you invented it first (and didn't abandon or hide your invention when it was ready), but that was changed with the America Invents Act.

> It's actually not a defense at all

Yes, looking at some law references I see you are correct. I also see plenty of law review articles arguing that it should be (with which I agree).

In the robotics community we've been working on self driving wheelchairs for decades. You're definitely not the first one to think of it, so no loss there: https://www.youtube.com/results?search_query=autonomous+whee...

I think the issue is not that he would have made the best/only chair and now isn't, but that it stifled any innovation that he might have provided by putting a high legal/monetary bar on any commercial endeavor. That is a loss, but we'll never know if it was vanishingly small or extremely large.

> Then I read a patent

If you're reading (non-expired) patents, you're doing it wrong. For one thing, that opens you to claims of intentional infringement and 3x damages. For another, there being a patent only matters if your thing works as you hope and the patent holder is litigious and there's actually a market for your thing and you can't find a bigger fish to buy your thing and make the litigious patent holder go away.

In the mean time, all the work to make your thing probably gives you lots of insights into other things to do if that one doesn't work out; and patents to file, if you're so interested.

Disclosure: I'm the named inventor on a patent (seems relevant).

You know something is wrong with patents when the suggestions are

- Don't read patents.

- Hope whoever 'invented' it first isn't litigious to sue.

- Do it anyway and maybe you will find more things to patent along the way.

> I knew the patent wouldn't hold up in court, but I wouldn't have the money to fight it.

You're not really going to be sued unless you have money to pay a judgment or settlement. If you have money, then you can pay to fight to invalidate the patent, or to pay a settlement.

This was similar to my experience as well; I've submitted a comment to the federal register linked from the article.

https://www.thisamericanlife.org/496/transcript

```

Alex Blumberg: Case in point--

Nick Desaulniers: About three years ago this time of the year, my father passed away from a heart attack.

Laura Sydell: Not long ago, I was at a meeting in San Francisco put on by the Electronic Frontier Foundation, an advocacy group that's been critical of the patent system. It was an open meeting. And this guy you hear talking, Nick Desaulniers, had decided on the spur of the moment to drop in. And after hearing a couple of other people tell their stories, he got up to tell his.

He's an engineer. And his father's death had spurred him to invent a low-cost heart monitor, which he believed might have saved his dad's life. He developed a prototype in a graduate school class. And he'd started dreaming a bit.

Nick Desaulniers: You know, I got really excited about it. And I was thinking, maybe I can make a business out of this. My grandfather started his own business.

My father made his own business. I need to make my own business someday. So this was something I was really thinking, like, hey, I can make a business out of this.

Laura Sydell: But then he started a patent search to see if anyone else had come up with something resembling his idea already.

Nick Desaulniers: I was horrified at how generic some of the patents were-- "a system for remotely monitoring physiological signals." What the hell does that mean? There are so many ways, so many variables.

How are you getting the data from point A to point B? There are so many ways to do that. And this patent, it covers-- I guess it covers them all.

And seeing some of the numbers that these companies are going back and forth suing each other for, I'm terrified to create a business because of the patent system. I'm horrified. And I'm scared.

And I'm not going to create a business because of it. So however many jobs I could have created or however many lives I could have saved, that's it.

Alex Blumberg: And that is the current state of things. Something as dry as the US patent system can move somebody to tears.

```

From the OP:

>> The study particularly wants to find examples of where a patent has been denied in the US but accepted in other jurisdictions, which will be a key area for European electronics companies.

The question at hand does not seem to be what the title of the OP indicates, but rather about patent applications denied in the US but approved elsewhere. Looking briefly at [1] too is hinting that the interest may rather be to relax the patent eligibility criteria so that more of patent applications in the US are approved. Note: IANAL.

---

[1] https://www.federalregister.gov/documents/2021/07/09/2021-14...

Excerpts: "Last year, after a split panel decision concluding that a method for manufacturing drive shafts was patent ineligible, ... The questions presented in the petition are: (1) What is the appropriate standard for determining whether a claim is directed to a patent-ineligible concept under step one of the Alice two-step framework?; and (2) Is patent eligibility a question of law for the court or a question of fact for the jury? ... The Senators indicated a particular interest in learning how the current jurisprudence has adversely impacted investment and innovation in critical technologies like quantum computing, artificial intelligence,[6] precision medicine, diagnostic methods, and pharmaceutical treatments."

Patents clearly kill innovation. Like, that shouldn't even be a question.

Fun fact: when the US entered WWI they had to buy warplanes because they didn't have the capacity to build them. Why? The original steering patent related to Orville and Wright was used to stifle innovation [1]. This was such a problem that a patent pool had to be created.

Likewise, e-ink devices have largely had no innovation due to the patents of the e-ink corporation being used to stifle innovation.

I'm relatively OK with pharmaceutical patents... within reason. The big problem in the US is you can't import that same drug from another country where the same company is making the exact same product. I realize they do this for price differentiation but this has gone way too far and is a big reason why US drugs are so expensive. In the very least you should be able to legally import such drugs from any developed nation.

But look at something like a smartphone and there are literally thousands of patents that apply to that product. That's... ridiculous.

Business models shouldn't be patentable. Software shouldn't be patentable. Obvious hardware innovations (eg applying more than one texture per clock cycle) shouldn't be patentable. Software can be protected by copyright.

The US PTO and the US Federal Court of Appeals are clearly in a state of regulatory capture at this point (so much so that the Supreme Court has taken to routinely slapping down the excesses of the US FCoA) so I suspect any feedback on this will be used to justify that yes patents do protect innovation.

[1]: https://en.wikipedia.org/wiki/Wright_brothers_patent_war

The e-ink patent is so frustrating.

Another good example of a patent restricting innovation was the patent on 3D Printers. That's why in the past few years, 3D printers have become so much more prevalent and accessible (you can buy one for a few hundred dollars today).

These definitely stifle innovation. I'd think a more reasonable approach could be, you have a patent for 15 years (or whatever arbitrary number is reasonable) or until the patent holder makes some multiple dollar amount from things that use whatever is patented.

That is, if it costs a million dollars to develop e-ink screens, the patent is valid until the company has had revenue of 20 million dollars on devices that have e-ink screens. Tying it to revenue instead of profit is intended since profits can be turned into revenue easily to bypass the restriction otherwise.

You could also add a minimum number of years (obviously less than the max mentioned above) so companies have some protection across various metrics.

In the EU, patents have escalating fees. They don't escalate very fast, but I think this would help with innovation a lot. IMO the first ~4 years can cost a few thousand dollars (so you can get your startup up and running), but the 20th year of a patent should escalate to several hundred thousand or million dollars. That way, building a long-lived "patent fence" becomes economically infeasible the way it is today.

Economically infeasible for NPEs, sure, but still a rounding error for Big Tech.

The issue with schemes that make it difficult for NPEs is that they hurt the small inventor who cannot afford to capitalize their invention. Owners of capital may be able to form a monopsony to drive down the price of patent licenses and essentially squeeze out small inventors.

Just as an example, IBM receives almost 10,000 patents a year. At $100,000-$1,000,000 per patent, that year of patents alone in 15-20 years time would cost $1-10 billion. That is most definitely not a rounding error and would constitute as much as 10% of their annual revenue just to keep that single years worth of patents. I think this sort of escalating pricing scheme would definitely help reduce the barriers for startups and new innovators, and prevent abuse by companies worth hundreds of billions of dollars.

Escalating fees still seems like a great way to limit the damage to small investors while turning the screws on NPEs, though. If the owners of capital can form this monopsony hurting small investors in the escalating fees case wouldn't they be able to do the same in the constant fee case? Is the proposition that they would be able to keep the monopsony together better with a bigger incentive (avoiding potentially high patent fees once they escalate)?

It seems to me that yes it doesn't do much to reduce the use of patents by Big Tech that they are actually commercializing. It would be nice to solve all the problems but it seems like it solves some of the problems that almost everyone but NPEs can agree on in a neat way without too much collateral damage.

Big Tech can afford to spend billions on patent moats for no other reason than to suffocate any small businesses that cannot raise enough capital to pull their heads above the water. An escalating fee scheme, while not wholly responsible for the problem, seems like it adds to it. It’s yet another barrier to entry that contributes to the enormous failure rate of startups.

Well suppose the valuation mechanism for determining the relative value of patents that the escalating fees were levied upon was implemented by requiring IP holders to declare the price at which they were willing to permanently release the discovery into the public domain in exchange for a one-time payment.

The result of such a scheme might be that is nearly always cheaper to pay the patent office to permanently destroy your competitor's patents and release the discovery into the public domain for everyone then it is to hire lawyers to go to court with your competitor in order to obtain a settlement and mutual licensing agreement.

So whenever large tech companies have disputes with each other they might start destroying each other's patent arsenals rather than going to court, which would level the playing field for everyone else. And if a large number of smaller startups were being harassed by a large company they could crowd source the money needed to pay the patent office to destroy the large firm's patents without spending any money on lawyers or legal fees and without having to halt development.

They might be able to afford it, but money is money. The question would be whether the cost of keeping the patent outweighs its potential profits.

The question should be how do we eliminate unnecessary transaction costs and overheads associated with license negotiations, legal fees, and courts costs, and ensure new public inventions can be immediately manufactured by a large number of domestic firms before they are knocked off by foreign manufacturers uninterested in obtaining licenses, while still compensating the original inventor?

A possible solution is to require the patent filers to declare the price at which they are willing to permanently release the invention into the public domain in exchange for a one time payment. Then if someone is sued for IP violation the defendant can crowd-source the inventor's fee and pay it through the patent office to have the the patent destroyed immediately without hiring lawyers or halt production.

The IP registrar could then levy a quarterly maintenance tax on the quit price filed by the IP holder to prevent the inventor's fee from being set infinitely high.

I like the idea but part of me thinks that would just benefit the companies doing well and hurt companies not doing well, potentially affecting competition.

Maybe combine this with a decreasing cap on royalties making patents less profitable over time?

A exponential cost curve to hold a patent would fix that. Every year the cost increases, making it less interesting to hold "dead" patterns bringing in no revenue.

That feels backwards. A patent nobody wants will be public domain. A patent that's actually stifling competition will stay protected. It'll at least cost money to stay protected, so it's not completely backwards, but I don't think this is the right approach.

Bad idea for patents. The public is vastly benefitted by patent expiration. It's also why we can have generic drugs, etc.

Now, exponential cost curve for _copyrights_ is interesting. And arguably already a thing... considering that Mickey Mouse is approaching 100 years old and is not yet in the public domain.

I could see a combo approach.

Ideally, I'd like to see tech patents see a sunset of something like 5 years. The current 20 year hold is bonkers.

However, if we keep the 20 year lifetime, then having an exponential cost of ownership would serve to ward off patent trolling companies. They could still exist to some extent but their operational expenses would be a lot higher than they are now. It'd put a number of them out of business (or at very least free up a bunch of patents).

The trick is coming up with good numbers here. Too low and you might as well not add it. Too high and you might as well eliminate patents all together as they've lost their original purpose (to protect the little inventor).

With the proposal for continuously increasing maintenance fees on ownership it is necessary to establish some public mechanism for determining the relative value of patents upon which the the maintenance fees are levied.

One possible solution is to require the IP holder to declare the quit price at which they are willing to abandon their claim and permanently release the discovery into the public domain prior to the expiration date of the patent in exchange for a one time payment.

This should reduce legal fees and court costs. If a firm is notified of patent infringement and pending litigation, instead of hiring a lawyer and going to court and halting production, they could instead crowd source the funds to pay the quit price to the patent office. The patent office would then pay the original inventor and place the discovery into the public domain for everyone so that there was no basis for continuing legal action.

When large tech companies have legal disputes, they might find that it is always cheaper to pay the patent office to immediately destroy each other's patent arsenals, which would also release all of the discoveries into the public domain and level the playing field for smaller firms as well.

If someone is willing to pay the quit price to the original inventor there should be no problem with expiring patents which have been issued for less than 1 day. 1 business day would likely work fine as the minimum duration for patents.

Counterargument: I get a patent for a special-purpose shell company MegaCorpA, license the technology to MegaCorpB for $1, concentrate all the profits in MegaCorpB and thus MegaCorpA will get a perpetual patent.

Tying patents to revenues of licensees does not work too, because any patent licensed to Samsung will instantly become void.

>> Tying patents to revenues of licensees does not work too, because any patent licensed to Samsung will instantly become void.

I don't see anything wrong with it. Other big companies will prefer to have Samsung as a partner instead as an enemy, and small companies will have very difficult to compete with samsung at scale. And if they do it is because they have done so many other things right, which it is great for consumers

Yeah, but why LG would rush into paying for the license if the patent will become free as soon as Samsung will pay to license it first?

Point is, the original idea of tying patents to revenues needs more thinking.

Because Samsung will use LG ideas too. As a company, I think, it is better to partner with others, so it is a win-win.

> The e-ink patent is so frustrating.

Hmm.. I keep seeing this on HN. I then ask which specific patent and I never get a satisfactory answer. But I'll keep trying. I work in the display industry and am not aware of any specific patent that is blocking progress in developing electrophoretic displays. Are you? If yes, please please point out which specific patent and how it is blocking your design.

I just have to point out, it was really easy to find these. Like took me 30 seconds.

https://patents.google.com/patent/US20080043317A1/en

https://patents.justia.com/assignee/e-ink-corporation

https://patents.google.com/patent/US9341917B1/en

https://stks.freshpatents.com/E-Ink-Holdings-Inc-nm1.php

And I have to point out that just googling for E Ink patents doesn't help us understand what point you are making. It is the equivalent of me claiming that Microsoft is blocking operating system progress and then justifying it by giving you a set of Microsoft patents. I hope the analogy is clear.

When a competitor appears, they usually sue with these two patents:

    https://patents.google.com/patent/US6120588A/en (expired)
    https://patents.google.com/patent/US20160187759A1/en

That said, there are a lot of inexpensive epaper panels for sale on Alibaba.

One of China's competitive advantages is that they don't give a shit about patents. This leads to extremely fast product cycles but also occasional stories about original inventors who feel they were screwed because people in China ripped off their design and improved it without their consent (or license fees).

An example: https://www.forbes.com/sites/wadeshepard/2018/01/31/meet-the...

In my industry, unfortunately, this has led to some level of bias and mistrust against employees of Chinese origin. The problem is, as you state, there have been multiple cases of individuals working for a company, getting access to key formulations, processes and then a couple of years later, the exact same product starts getting made in China.

> they usually sue with these two patents

I couldn't find a lawsuit referencing either patent. Is there a link you could help provide?

Legal threats won't necessarily result in lawsuits.

Lack of publicly accessible lawsuits doesn't input lawyers of Company X doesn't shot down every single remote competitor preemptively with threats of legal action. At that point possible competitors either won't even try bothering or they'll pay whatever asked and sign a NDA preventing then from disclosing the patents in question (as a common practice on patent litigation deals).

Just to confirm since this isn't publicly accessible, are you saying this actually happened with E Ink? as I mentioned I work in the display industry, and if something like that happened, surely we'd all be talking about it during SID or stuff like that. When did it happen? Where? I'd love to know.

I'm telling how patent litigation works in general and why you can't expect anybody to spoonfeed a random stranger on internet with a list of patents.

> I'm telling how patent litigation works in general and why you can't expect anybody to spoonfeed a random stranger on internet with a list of patents.

I'm unsure what's with the "spoonfeed" remark, it seems intended to be insulting. As I said, I'm really curious what patent and what lawsuits or similar criminal activity has been done (or from the sound of these comments is being done) by E-Ink. HN people are referring to it in many posts and because I happen to work in the display industry, I'm really surprised. I'd never heard anything like that amongst my peers or during conferences. The main place I've seen this refrain is on HN. Please have a look at my comment history as I keep asking about which patent people are talking about. It is the equivalent of a bunch of my display industry people claiming that Google is blocking search engine progress and then justifying it by giving you a set of Google patents and then shouting, see there's the clear proof that Google is a bad actor. I hope the analogy is clear and that it is also clear why I remain unconvinced about both!

https://www.eink.com/news.html?type=releasedetail&id=223&yea...

I'm not sure why you find it so hard to believe that the Eink group has created 2000 display patents to use against competitors. At the bottom of that page they say:

> E Ink intends to protect and defend its intellectual and other property worldwide by the means available in various countries.

I tend to believe them.

> I'm not sure why you find it so hard to believe that the Eink group has created 2000 display patents to use against competitors.

I guess it is because I work in the display industry and I've never heard them talked about this way within my industry, conferences, symposiums or other peer group events. I've only seen this here on HN and on Boing Boing and some blogs which referenced an HN post making similar claims. I'll try to give you an analogy. Imagine if someone told you Microsoft group has created 2000 software patents to use against competitors and that they intend to defend theier intellectual property. Would you then believe if I told you Microsoft is blocking progress in the software industry? Because that's what the prior post is claiming. Incredible claims ought to at least have some proof or evidence, shouldn't they?

I would believe Microsoft was blocking progress if only they were making operating systems or office suites or whatever it is they patent.

It isn't just a claim though - I pointed to an actual lawsuit. If there was a single eink competitor being used in products outside of China, I would agree with you. TCL claims to have something in the works, so maybe change is coming.

> If there was a single eink competitor being used in products outside of China

I'd say there are far fewer Microsoft operating system competitors than there are eink competitors. Also eink is an fpl manufacturer not a whole industry. Meaning they're a component supplier and there's tonnes of alternatives, memory-LCD is what I promote, and even within the electrophoretic space, there's ClearInk, Shenzhen Wenfeng, and several others. In any case, it is unclear to me what the main argument is. If you are trying to say eink is blocking progress, then I don't see any evidence for that, and that 2012 lawsuit in fact seems like an entirely reasonable lawsuit since based on what I understand, 2 engineers who had worked for Eink in Boston moved to China, started Guangzhou OED and then produced the exact same pigment and particle formulations and started selling knockoff displays with even the exact same TFT backplanes and resolutions.

> I'd think a more reasonable approach could be, you have a patent for 15 years (or whatever arbitrary number is reasonable) or until the patent holder makes some multiple dollar amount from things that use whatever is patented.

The problem with patents is that they are national laws. A 15 year patent on a fusion reactor can be ignored by manufacturers in other countries after waiting 1 day without obtaining a license. Then domestic manufacturers are saddled with legal fees and court costs for decades.

A better reform would be to require the patent filers to declare the quit price at which they are willing to immediately release the discovery into the public domain in exchange for a one time payment. Then if an inventor patents a fusion reactor, and other countries begin manufacturing it immediately while ignoring the patent, domestic manufacturers could crowd source the money to pay the quit price to the inventor to have the patent office destroy the patent after 1 day as well.

> You could also add a minimum number of years (obviously less than the max mentioned above) so companies have some protection across various metrics.

As long as the original inventor gets a cash reward why is a minimum time necessary? If the patent is destroyed one day after being filed the inventor still has the original copy of everything they built and cash reward, they just don't have a monopoly on the ability to make additional copies, and neither does anyone else.

> I'd think a more reasonable approach could be, you have a patent for [duration] or until the patent holder makes some multiple dollar amount from things that use whatever is patented.

I don't think this can work in practice. How much money has amazon made off the 1-click patent? If a company patents a new machine that lets them run a factory on half the workers, how much money does that make them?

Even for devices, this seems sketchy.

What percent of revenue counts towards the patent? Is the $100 kindle $100 towards the e-ink patent, or $20, or $5? If the device has 50 patents on various components, is the revenue counted 50x for them, or is the cost divided between them?

I think that enough patents are on internal processes and machines that revenue won't be directly tied to, i.e. on things that aren't consumer goods, that this sort of strategy won't help.

You're right that creative accounting games could be played to try to keep money made on the patent from being considered profit. While harder, I suspect creative accounting could be used to done to make it look like there was no revenue on e-ink screens despite millions of devices containing them had been sold.

Yes more on 3D printers:

The first FDM 3D printers went on sale in 1995 for $50k. Ten years later they were $25k. Then in 2008 the patents expired. By 2011 a halfway decent printer was $1800. By 2018 $300 3D printers were available worldwide.

I joined the community in 2010, bought my own printer in 2011, and I saw the community operating with my own two eyes. Thousands of people all over the world innovating without patents to improve the situation.

The idea that patents promote innovation is absurd. The primary function of the patent is to stop innovation. All those thousands of people who wanted to improve 3D printers were legally prevented from innovating while the patent was in effect.

I remember in 2012 people on the mailing list talking about 3D printing on to a belt that could move. Literally no one on the mailing list wanted to risk infringing on the Makerbot patent for this even though makerbot quickly stopped selling their belt printer and never sold a new one. To this day the only belt printers that are for sale is a printer with the carriage off angle from the belt to avoid infringing the makerbot patent that hasn’t been used in years.

The rate of innovation would be higher without patents. We’re hurting ourselves for the benefit of those who amassed large patent portfolios so they can charge rent on old designs instead of innovating.

Imagine the health and wealth that could be generated worldwide if we let every country copy the best MRI machines we have, and all our battery factories and chip designs. That wealth would immeasurably improve the lives of people the world over, and it would come back to us in the US as new economic trading partners with their own infrastructure. Along the way they’d find 100x cost savings or better and our medical care would be cheaper.

We are truly shouting ourselves in the foot with patents.

If there were a political candidate whose sole platform was "dismantle the US Patent Office," I could look past almost any other position in my conviction to vote for them. That's how harmful I believe the current patent system to be. Innovators should not be afraid to be sued over trivial implements by non-practicing entities before they can even get their business off the ground. Patents kill innovation, 100%.

The U.S. had a president not too long ago who tried to shake things up. Ended up giving the patent office truckloads of cash, considered it "job creation". Goal was to improve efficiency, but all it did was unthrottle submissions, and lower quality. On paper, it looked more efficient, apparently.

The U.S. also "harmonized" their patent laws with the rest of the world to reward paper-pushers who are "first to file", instead of those who are "first to invent." Should have been the other way around.

That with lowering submission fees and requirements drastically, those efforts basically turned the office into an even bigger joke.

Getting a patent means virtually nothing without enforcement and protection, and that requires litigation.

First to file has some advantages that are hard to ignore.

It incentivises early submission which means we all learn from it earlier.

It disincentivises submarine patents where one party deliberately hides an obvious invention until another market player has established a business and then proves they invented it first.

Strangely, after the patents for mechanical keyboard switches expired, keyboard prices seemed to have doubled or tripled, except for the really crappy Amazon ones that break after 6 months (I’ve been through two)

Then stop buying the crappy ones. Like with anything else, you get what you pay for.

That’s rude. 10 years ago I could buy a $75-100 keyboard that would last 6 years. Now the same money gets me something that lasts 6 months so I have to spend $270 to get one that’s decently made. Old $20 keyboards from thrift stores and the drug store used to last me 5-10 years. But yeah, everyone should spend $250+ or else they are chumps

Leopold. $109. One of the best boards you can get.

https://mechanicalkeyboards.com/shop/index.php?l=product_lis...

It's not rude to point out that you impulsively bought low end stuff and expected it to be better just because it was "mechanical." Lemons and garbage exist in every product category.

> Likewise, e-ink devices have largely had no innovation due to the patents of the e-ink corporation being used to stifle innovation.

Uhm... I work in the display industry and have never heard this. They seem to be a typical display manufacturer progressing at what is decent or maybe even above average for display industry products. Could you clarify what patent you're referring to and what exactly you mean by no innovation?

Pharmaceutical patents, operatively, are really more about latter stage of drug development (human trials, regulatory approval, etc.) than they are about invention and discovery. A limited monopoly (or alternatives, such as bounties) make sense here, but we could drop the patent designation altogether. In fact, working through a patent frame can get in the way. Drugs that aren't patentable fall outside the framework and therefore have no funding mechanism for these later frameworks. Start the clock at approval. Drop the limitations. Prior art

Business models, software & obvious/trivial inventions were never supposed to be patentable, and technically aren't. The fact that they often are in practice is an argument for reform. One-click remains the perfect example, as it falls into all three categories.

Also, look at the state of the patent "industry." Patent trolls. Admit-nothing mutual licensing agreements. Patent arsenals held for defense. The Apple-Samsung war, with no result or benefit to either.

It seems like a slightly weaker patent system where the holder only retains the right to share in the profits on a piece of technology would solve most of the problems.

Like, once you publish a patent, anyone who uses it needs to pay you a meaningful portion of profit they derive from using it. The patent holder still has most of the incentive to use it, but R&D on the patent would be fully open and incur no fee. Third parties could sell improvements back to the patent holder or file their own patents on the improvements.

>> Patents clearly kill innovation. Like, that shouldn't even be a question.

Surely there are examples where patents have stifled innovation. I am unclear how this would apply in general.

The idea of patents is to encourage inventors or organizations "invest" into research and development to create inventions and reap rewards for their investments. If there were no patent protection (and ignoring other forms of intellectual property for sake of simplicity), then inventors or organizations would either (a) not put money into any inventions that could easily be copied by others upon launch, or (b) have measures to keep the inventions hidden even upon launch. In that sense, patents mechanism are intended to lead higher investments towards developing inventions, and, are set up such that the inventions become publicly known and available after respective periods of time.

Is the suggestion that the patents should be abolished? If so, one also needs to explain how and why would inventors and organizations spend money into research work leading to good inventions?

I agree that problems do happen. There are many patents around which should not have been issued. However, the challenge is not that the law is at fault. The challenge is that execution is at fault wherein many obvious "inventions" get accepted for example.

I can accept if the argument made is that good execution of the well-intended patent laws is guaranteed to be a challenge and therefore the patent law does not make any practical sense. However, we also need to be aware of the side effects of it as reduction in innovation because of reduced research and development investments.

---

Recommended reading: https://en.wikipedia.org/wiki/Patent

Ideologically patents are suppose to promote innovation. The original idea and spirit of the law was that for a limited time monopoly you disclose all details on your invention instead of keeping it all a secret to yourself.

Obviously this is no longer the case as the problem is (1) too many patents on broad things and (2) too long of a monopoly term for todays fast moving technology.

It's been 150 years since we established the patent office and the only big things that were changed was moving from first to invent to first to file and increasing the monopoly term from 17 to 20 years. I think a 5-7 year monopoly term is much more reasonable and would maintain the original spirit of the law.

> Software can be protected by copyright.

You want software to be protected for the lifetime of the author + 70ish years? 95 years in case of corporate authorship?

The Beatles catalogue, released 60 years ago, won't pass into the public domain until 70 years _after_ the last Beatle dies. Maybe my grand kids will see that day.

Given that the law almost made the 8 note ostinato from Katy Perry's Dark Horse into a copyrightable bit of music I would weep if we tried to really apply copyright to software.

For an example take the "Loading Screen Game" patent (https://www.eff.org/deeplinks/2015/12/loading-screen-game-pa...)

> According to the law, a person isn’t entitled to a patent if the claimed invention already existed when the application was filed or would have been obvious to someone skilled in the relevant technology area.

This usually is completely overlook in patents. So, patents are ridiculous, they overlap and create a legal minefield for any small developer. Patent trolls exist because the patent system is so easily exploitable.

I am not against the concept of patents, I see the value of allowing some corporation to regain invested money. But, it needs to be for real inventions that took time and effort to develop thru costly processes. Meanwhile things like "one-click buying, is the technique of allowing customers to make purchases with the payment information needed to complete the purchase having been entered by the user previously" are accepted. Patents are a useless joke, at least in the software industry. And its only goal is to increase the power of already monopolistic corporations.

A bounty system for patent invalidation might be nice. An idea could be:

* When you get a patent, you put up a bond.

* If someone invalidates your patent, they get the bond.

* Otherwise, you get it back when your patent expires

An extra, funny, probably impractical idea is that when filing for the patent, you decide on the value of the bond. But, the value of the bond is the maximum payout you can be awarded for a given instance of infringement. So, you have to essentially 'price' the value of your patent yourself.

Ah, an even more pay to win system? Sounds like it'll work perfectly and no one will ever abuse their capital to unethically tie up patents.

If they are throwing money at patenting actually novel ideas, then they are using the patent system as intended.

If they are throwing money at the sort of ridiculous obvious patents that people here are complaining about, they are essentially funding an army of private investigators to go out and invalidate bad patents.

> But, it needs to be for real inventions that took time and effort to develop thru costly processes

Be careful with this. I have a friend who has a patent (might be expired by now) that basically came to be because I asked him a question about a battery charging design he wanted me to write some code for. It didn't make sense to me but in the process of explaining it, a lightbulb went off in his head and that was the basis of the idea. It's actually funny remembering the sudden excitement crossing his face as he realized he was on to something no one had thought of before (he had years of experience in the field).

At the time I was already an experienced EE and I can assure you that the idea was definitely "not obvious to a skilled practitioner in the field" but yet it didn't take much development (we had it working about a week later).

Found the patent here: https://patents.google.com/patent/US5894212A/en

[edit] looks like it's been cited by quite a few other patents since then! I never asked him how much he sold it for.

Absolutely. The design of the patent system needs to be significantly updated. Every single new thing seems to be violating an assortment of patents, most of them for extremely obvious or common things.

When Apple introduced the LiDAR sensor on iPhone, lots of people started building apps to map out the interior of homes to break the monopoly that Matterport seems to hold on that part of real estate technology.

I started researching that stuff for a tangentially related idea that I have, and if you want to go from a LiDAR point cloud to a set of polygons representing a room, it seemed like you would have to violate a dozen of Matterport's patents. Truly ridiculous stuff. As I dug into the documentation around it, I realized that Apple will actually do the spatial reconstruction for you if you want, which is nice because that presumably means that Apple is the one violating the patents instead of you, right? It's hard to say. I wonder if even a lawyer could give you a straight answer to that question.

Similarly, maybe I want to make a better kind of 2D barcode. QR Code patents supposedly won't be enforced against anyone implementing or using them for QR-compatible codes, if I remember correctly, but... a few years ago when I looked into this, I saw that the company that owned the QR Code patents had continued publishing new patents seemingly in order to extend their hold over QR Codes (and similar 2D barcodes) in perpetuity.

So, if you make any kind of 2D barcode that is even remotely similar to a QR Code, talk to some lawyers or... good luck?

And those are just two examples of things that I have run up against as an individual with ideas who would love to innovate, but feels restricted by the infinite number of patents that exist.

I think some serious reform of intellectual property rights could spur significant new innovations and societal benefit, while still protecting people who are actually innovating and applying those innovations.

(Keep in mind that I am not a lawyer.)

presumably means that Apple is the one violating the patents instead of you, right? It's hard to say.

Maybe Apple is even licensing the patents. It would be kinda nice to think that the system was actually working as intended, for once. Somebody invents; somebody else wants the idea; the license is made available at a reasonable price and they get paid for it.

Probably not. But a guy can dream.

The "system working as intended" would mean that blatantly obvious ideas could not be patented in the first place.

There are only so many ways to do basic things. I'm sure most engineering students could come up with an algorithm to go from a point cloud to a set of polygons, and their algorithms would likely violate a bunch of Matterport's patents, because those patents are so overbroad.[0] That's not okay, in my opinion.

It's like the infamous Mobileye patent that patented the idea of using a camera and a computer to read a speed limit sign while a vehicle was in motion, and then to react to that speed limit in real time. It is written so generically that there is literally no obvious way to avoid violating that patent if you want to have a self-driving vehicle (or vehicle with a strong ADAS) react automatically to posted speed limit signs. As human drivers, we are essentially violating that patent every time we read a speed limit sign, but somehow it's different when a human does it with their eyes (cameras) and brain (computer)?

[0]: Keeping in mind that I am, of course, not a lawyer. Maybe a lawyer would find those patents to be far more specific and complex than I did when I was doing some research, and therefore anyone can implement scene reconstruction without fear of judgment. If that were true, that would be nice.

Imagine being on your death bed with your only contribution to this world having been vaguely worded abstract patents which didn't translate to any real products. Especially knowing that had you not helped to stifle medical/tech innovation, maybe you would still have a few extra centuries to live. Yep, that must feel like being a cockroach and a failure in one.

Also, the article has a wrong title. A patent office would never undermine their raison d'être. In this case they're only asking to be compared to other countries so that they don't issue fewer patents than them.

The patent system is so bad that, at this point, the only way to fix it is to completely abolish it. Do you still believe that the patents are here to protect you, a garage inventor, against the mega-corps with unlimited free-money bank funding?

The real philosophical question here is where do your thoughts come from - did you really create them, are you really so special that given enough time nobody else would come up with that idea? Are you aware of the concept of multiple discovery?

  https://en.wikipedia.org/wiki/Multiple_discovery
  https://en.wikipedia.org/wiki/List_of_multiple_discoveries

Most IP should be scrapped completely in the age of free and cheap information. I really don't have any sympathy to big corps making money from IP, and small entities are finding more and more ways to give their IP away and find revenue elsewhere (hardware, swag, live events, Patreon, etc).

Scrapping IP would be a big step in making more, smaller, and more sustainable businesses instead of our current 'grow or get bought out' mindset.

Ideas are a dime a dozen, execution is all that matters. I would wager that most reading this website have had strangers and family alike approach them for help with with their huge ideas, which really means "you are the programmer you do it all". Most of which are utterly unoriginal, yet they want you to sign an NDA or promise to keep it secret. Patents and to some extent copyright/trademarks tilt the scale heavily away from innovative ideas toward 'derivative but licenseable'.

I am running on 4 hours of sleep, hopefully my word salad makes sense.

Ideas being cheap and execution being everything applies mostly to businesses that laypeople people can think up and start. Because they are not usually in a position to do new research, but great execution is always scarce. In pharma (which I’ve worked in) and I guess other industries like semiconductors, materials etc good ideas are everything and competitors could copy the invention that took years to decades for just single or double digit millions so patents arguably drive a lot of research there.

Tax patents. Progressively. Start from say $USD 1000 per year and double each year until the patent holder decides to not pay and gives the patent to the public domain. Feel free to add a couple of free years in the beginning or tweak some other parameters to suit needs of different industries.

I think a similar approach would work with copyright, too. Only with a grace period.

Say, your original copyright lasts for 20 years since creation of your work. If you want to prolong it, you can, but you must register your work and pay a yearly fee that starts somewhere around USD 5, but grows exponentially to reach millions after 50 years and billions after 100 years.

Not sure how such copyright tax would be interpreted in Berne Convention context. (Would happily support that as well, though)

Just renegotiate that. I'm convinced that copyrights > 25 years are not a limited time via the constitution, and thus the treaty isn't constitutional in the first place. Of course I'm not in a position to do anything about this.

The Copyright Act of 1790, just a few years after the constitution was signed, enabled a 28 year copyright grant (14 years with a 14 year extension.) So unless you think the copyright laws haven't been legal since their inception, you probably want to bump your number up at least five years.

Edit: Also, if you're interested in this argument, and IIRC the most recent Supreme Court opinion on it is Eldred v Ashcroft (2003).

The exact time is subject to debate.

I used to think this too, until I thought about it more. The biggest issue with taxing copyright is eventually only the biggest companies can afford it. Imagine if Disney could simply wait until Lucas couldn't afford the Star Wars copyright anymore, and then just acquire the copyright. So now Lucas gets nothing and the government gets billions. Everyone loses in that situation.

I think copyrights should be taxed like income. File a special return showing the income that copyright got you, and pay a percent. Audit it like the IRS audits taxes. Make it a tax on top of income tax. So for example Disney would pay their normal taxes, and then additional tax for income earned from Mickey, Donald, Darth Vader, etc. Maybe allow them to group copyrights for simplicity.

Then do patents the same way. Pay a tax based on how much you make from a patent.

And then once you do that you can get fancy. Allow anyone to use a patent and pay the tax, but have their tax get split with the original patent holder, or increase the tax and have to extra go to the original holder. Then invocation isn't stifled anymore. If you can make an e-ink reader and make it profitable enough to cover the tax, then great, everyone wins!

Expiration of copyright would have to mean instant and irreversible transition to public domain. In that case, Disney would have to pay reasonable money to Lucas, so that he does not let his copyright expire out of spite.

Also, 20-30 years of automatic or very cheap copyright mean that the original creator has a lot of time to make some money off it. Although there are works that only became "hits" after decades, this is a fairly rare situation.

For copyrights, why is this a problem? Consider the early copyright lengths of like 20 years + 20 years after the author's death. We're now at far, far, faaar beyond that.

What if we did 20 + 20 and then every year after it, we charge $X and then geometrically or exponentially increase the price.

Yes, it would be a tax to Disney; but, if Disney is willing to spend $2B in 2010 to keep Steam Boat Willie and then $2.5B in 2011, I think that's a fine tax for them to get to keep it.

I don't think most people had _any_ problem with the 20 + 20 or similar patterns. They have issues with the, what is it now, 120 years?

That is an exceptional idea!!

This is how it works. The prices are different in different countries. Most countries have renewals due every year. Many countries have increasing costs each year, and costs vary between tens of dollars and thousands of dollars. Costs are also categorized between "Micro", "Small", and "Large" entities. Different countries use different numbers of these categories. The US uses all three.

Some countries have fees per claim in the patents.

In the US, the renewals are only due 3 times, every 3.5 years after the grant of the patent. The renewals cost $1,600 for the first renewal, $3,600 for the second, and $7,400 for the third. Those are for large entities. Small entities are half the cost, and micro are 1/4th the cost.

Renewal fees, world wide, are a non-negligible part of a legal department's fees, and abandonments do happen before they otherwise would because of the cost to keep non-useful patents around.

As a commentor noted, there are already fees to maintain a patent.

The problem with the approach is that the fees are often trivial for big companies, who may not think twice about maintaining a patent "just in case" even when they are not using the patent.

Yeah. Two potential options to fix sitting on unused patents are to increase annuity fees until they hurt, or to enact a compulsory licensing scheme.

I'm pretty ambivalent either way. I do think that compulsory licensing schemes are hard to manage but if done right would be extremely valuable. Compulsory licensing of patented pharmaceuticals saves lives.

Doubling the tax each year will make the tax non-trivial for even the largest corporations in reasonably short timeframe. You may want to check what e.g. my initial proposal (1000 bucks first year, double that annually) would yield in say 30 years.

Makes sense. Sorry for my miss the first time. :-)

So what this means is that the small players would give up before the big players even though the invention otherwise took the same amount of investments or has the same potential for reward back.

The shape of the exponential curve makes the time difference between small and big players giving up relatively small (if that difference exists in the first place.) Both have reasonably long time with reasonably low patent tax to develop the commercials before the tax starts really kicking in. The tax with my initial numbers breaks one million USD p.a in 10 years and 10 millions in 14 years.

Yes, the math is already understood. :-)

Your proposal does seem to carry some weight. While I see some non-idealities in it, but then I don't have a better solution that's practical.

I'll think some more about your idea; I just need to consider various scenarios with it.

Thanks.

Taxing a companies patent portfolio with a minimum tax per patent would be salutary.

Of course patents kill innovation and the reasons are very simple and known to many.

It doesn't matter what the law says. It doesn't matter if the patent is generic or very specific. It doesn't matter if it covers a currently trendy area.

None of these things matter one bit.

Only one thing truly matters: how much budget for a lawyer does the suing side have.

That's it. Only that. A talented and highly-paid (and thus motivated) copyright / patent lawyer is literally the most dangerous adversary you will ever encounter in your civilized life (we're not including stuff like Mexican cartel hating you, of course).

It doesn't even matter what legislation grounds for suing you does the litigating side have. If they have a square inch, they'll use it and expand it to be as big as the USA.

A lot of small businesses are well-aware of the power imbalance and I've heard founders give up on ideas, citing various copyright / patent trials from the last 10+ years.

Truth is, if a big player comes after you and if the government doesn't help you have some semblance of a fighting chance (by covering your legal expenses and/or hiring a competent defender for you) then the 99.9% predictable result is: you'll run out of money and will be forced to settle if you are lucky -- and in most other cases you'll have to pay damages or royalties for years, which might very easily make a homeless person out of you.

I am glad that the US patent office is making an inquiry but at the same time slightly disappointed because what I just said above is common knowledge in many circles.

Patents don't kill innovation but they do next to nothing to promote it.

Only a very small percentage of patents are ever successfully licensed to anyone. Around 3% recoup the cost of filing the patent. Half of all patents expire prematurely when the owner declines to pay the required maintenance fees.

These are all features and should be accelerated. There shouldn't be a possibility to renew.

That 97% of the patents don't go recoup the cost is a statistic known to companies and patent lawyers. This should only serve to make them more careful about filing patents. In other words, inventions less likely to be effective would be passed upon by them themselves.

Patents were supposed to provide enough information so that someone else could build a widget, in exchange for a limited time monopoly on that widget. "You teach the world, we'll let you eat for a while on that."

What we got: A convoluted, arcane and rent-protected process which is gamed by the big players, and in fact diminishes advances by sowing fear, expense and uncertainty.

I'm a software engineer, but a while back I got a chance to read hundreds of patents (with expert help from some compan lawyers) in an attempt to carve out an "IP space" in an already crowded and contested space. Pretty eye-opening. In a couple of cases the parties we were looking at were trying to patent stuff you could have lifted from a textbook. The screening process for "is this a patentable idea?" seemed utterly broken.

To this day I can't keep a straight face when a lawyer says, "... and here, the '843 patent teaches that....".

/Teaches/, yeah right.

Conceptually, it should help protect innovation. In reality, it inhibits it. Small guy has a brilliant idea. He files a patent, and starts shipping. Some other guy copies the idea. Small guy has 3 options. 1) invest in better products and out innovate copycat. 2) invest in marketing to create brand appeal. 3) get lawyers involved.

#3 is the most expensive for the return on investment. #1 or #2 are likely the better path, but patent protection helps neither of these paths.

On the dark side of patents is patent trolls. They only seek to extract value for infringers, almost always via legal system intimidation (see #3). Again, the system intended to help innovation is killing it.

Questions like this will always founder if you equate software patents with real inventions. As Joel Spolsky said, there are maybe two ideas in software per year that are really worth patenting (actually, he didn't say "per year" so I was being generous).

I wrote an article [1] about this seven years ago! It was cited in an amicus brief to SCOTUS in the CLS Bank case. Nothing has changed since then.

We should realize that our companies' top management and the professional societies are not on our side on this. Engineers themselves will have to get it done.

[1] https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2399580

When I worked at Apple, we used this (in my opinion terrible) framework called "WebObjects", which used to be publicly available but not open source.

After digging around, I found out that the reason they didn't open source it was because their own patent lawyers wouldn't let them, something about having dot-properties for configuration variables... Something that nearly anyone who frequents hacker news could reinvent themselves in an afternoon.

If that were the case, china should be IP wise blooming. It does not.

The more stiffing factor was and always has been, monopolies and oligopolies. Huge power concentrations murder the competition were they can find it. Which is why the original fractured nature of Europe was such a breeding reactor for innovations. Local leader does not like your project xyz? Just pack it in, go a few miles, to the next local leader in opposition who will welcome you with open arms.

Situation wise, a fractured, heavily competitive economy, with some "safe havens" (academia) for involuntarily happening knowledge exchange is the optimum for innovation.

PostScriptum: To zoom in further, on the lab level, to promote innovation, is quite distant from the pursuit of excellence everyone agreed on as the optimum today. The ability to learn,apply and reproduce past gathered knowledge does not make you a innovator, quite contrary. Mimicry while flattering, is not able to reproduce true geniusgenesis.

One needs a gear-shifting approach, a combination of incremental researches (to follow a path through to the optimums it reveals and explore the terrain methodically) and one needs for lack of a better description, schizophrenia affected individuals, who even though their grades suck, are able to constantly recombine the currently existing ideas into new "conspiracies" and filter out the useful ones from the avalanche.

Needless to say, that guiding such a lab with the constant conflicts and drama, is a task more fitting to a theater-director, then the classic academic bureaucrat.

P.S.S: I'm not responsible for whatever hiring catastrophe this advice may produce. You may jam strange devices made from people into the leviathan at your own risk.

When I worked at Amazon, I learned that search field auto-complete on the web was patented by Amazon. When I wanted to provide a better user experience in some Amazon-owned open source software, I couldn’t because they didn’t want the Apache 2.0 license to grant patent rights to users. (If the age limit is still 20 years, I _think_ the patent expired in 2019 or 2020).

Conversely, there was that time that Creative patented Apple’s iPod interface (or scroll wheel or something), then they promptly sued Apple over it. I would assert that deeper discovery be done on prior art before granting a patent.

Lastly, the patent trolls who don’t create anything, but own the patent for it (e.g., Eolas suing Microsoft over then-standard NPAPI browser plug-ins). I would assert that granting any patent requires an actual implementation.

I think patents have a place, but (a) see necessary changes above, and (b) I think that software (specifically) should have a shorter patent lifespan because it moves so damn fast.

I think firearms related patents are an instructive example. It has long been standard practice to patent new firearm designs. For the most part it works well, the patents prevent companies from immediately churning out cheap clones of successful designs.

However, every once-in-a-while a patent becomes problematic, like the infamous Rollin White revolver patent. White patented a revolver design using then new metallic cartridges. The design itself was not commercially successful, but the patent was construed to cover any revolver design where the chambers were bored completely through the cylinder. It turns out this is the only reasonable way to make a revolver using metallic cartridges so it lead to many years of legal battles and bizarre designs attempting to avoid the patent.

I'm not sure how to reliably identify Rollin White type patents, but there certainly seems to be a distinction to be made between beneficial and detrimental patents.

I agree that it's nuanced. I think the question should be rephrased to "When do patents kill innovation". And the answer is that it happens when a new field opens up and the long hanging fruit is locked away for longer than the innovation merits. The reward should be, if not in perfect proportion, at least in rough proportion to the r&d work done.

Yes. But it's an easy fix:

* Anything that is copyrightable cannot be patentable. That should be obvious. "inventors" must be forced to choose one protection

* Software should not be patentable in every circumstance: It's well established that math cannot be patented. It's well established that business processes cannot be patented.

* Anything that is copyrightable cannot be patentable. That should be obvious. "inventors" must be forced to choose one protection

That's the case today. Algorithms can be patented. Computer code can be copyrighted. These are different things.

Well... I'm in embedded systems. There, software can be part of the machine, performing concrete, physical actions. Would the same thing be patentable if you implemented it as an FPGA? How about as discrete logic? If so, then it should be patentable if you implement it as software.

But this is very far from a software-only patent, or a "doing X, but with a computer" patent. I guess the difference might be that, if it's part of the functioning of an embedded system, then it's not math (or at least not just math) in the same sense that a mechanical machine is not just physics.

No.

If you invest a machine to dispense a weighted amount of ice cream using a microcontroller and software, that should not be patentable for the reasons listed above. This may be tough to hear, but there is nothing novel, inventive, or unique about using a general purpose device (a microcontroller) to accomplish a specific task.

Now, inventing a new breed of general purpose embeddable device, that might be a bit different.

Take anti-lock brakes, for instance. Was that patentable? Yes, it was, and rightly so. But it used some software. Did that make it not patentable?

Or take your ice cream dispenser. Let's say that I invent an ice cream dispenser with a camera. It's going to estimate the user's BMI, and based on that, it's going to decide whether to dispense regular or fat-free ice cream. (Sorry, it's a horrible example, it's the only one I could come up with for an ice cream dispenser.) Well, you know there's going to have to be a CPU in there running some algorithm to try to derive the BMI from the image from the camera. Does that make it unpatentable? (Yeah, I know, it should be unpatentable because of the overall insanity of the idea. That's not the point.)

As our physical devices become smarter, more and more of the new behaviors are implemented in software. Does that make them unpatentable? As I asked before, if I implemented the same logic in an FPGA, would that make it patentable? If so, why?

There are plenty of parts of anti-lock brakes that are novel mechanical devices, like fast acting solenoids or pumps. The software running on a general purpose device is not novel or unique. Patenting "anti-lock brakes" is a great example of an overreaching patent, as you can't lay claim to every implementation of "anti lock brakes". Patenting individual parts that are novel and unique is applicable.

This practice didn't even come into play until the mid-late 1990s when lawyers realized they could put "may use cd-rom" or "electronic" in their patents to cover a wide swath of things they didn't invent.

> As I asked before, if I implemented the same logic in an FPGA, would that make it patentable?

No. Like software, anything which you can implement in an FPGA is pretty obviously math. It's literally a mathematical (boolean logic) equation. Possibly a recurrent one when you consider registers and state machines, but still 100% math. An FPGA is no different from a general-purpose computer in this regard. And FWIW, the same would be true of logic implemented as an ASIC or discrete logic circuits.

What is not just software is the overall process, particularly the physical sensors and actuators and the relationships between them. The ice cream dispenser example is getting a bit ridiculous, so let's talk anti-lock brakes instead. The individual components to sense the car's motion, braking force, and the rotation of the tires are obviously not software. The same goes for the actuators which interface with the braking system and create the pulsing effect to un-lock the tires. The software itself which describes how these parts are related ought not be patentable, and an equivalent implementation outside the context of an actual ABS system (in a simulator, for example) ought not infringe on any patent even if the code is functionally identical to the code in the ABS unit. The ABS system as a whole, in the context of a vehicle, is not merely software, so a patent on a system which detects certain conditions via physical sensors in a vehicle and brings it to a safe stop by physically interfacing with the brake controls would not be a software patent, even if some software is used to implement the claims. Accomplishing the same effect with an FPGA or discrete logic, or even mechanical linkages, would be covered by the same patent, since the software is not part of the claims. (On the flip side, merely replacing discrete logic, ASICs, or FPGAs with software having the same effect—or vice-versa—would not qualify for a new patent for exactly the same reason.)

Going back to the ice cream example… IMHO that would be a really hard sell, mostly because all the "value" (using the term loosely) comes from the ML. It's not really much of an improvement over connecting the fluid switch to a mechanical scale, which could easily be accomplished without any software. You didn't invent the camera or the switch or even the ML system (ignoring for a moment the fact that the ML system on its own is pure math). Maybe you could claim something about how all these weighted inputs from the pixels in the camera combine to control the actuation of the switch. That seems like a really complicated patent application due to the input of inputs, and since it's ML-based and self-trained you can only explain how it works, not why. I'd be inclined to reject it simply based on the fact that such a disclosure really wouldn't benefit the public at all, if by some miracle patent clerks were allowed to represent the interests of the public in that manner. But assuming the patent were granted, it would only cover the use of this logic in the context of an ice cream dispenser. Applying the same ML system to classify images according to BMI in another context would not infringe on the patent, since it was not the ML system (the math or its evaluation in software) which was the subject of the patent, but rather its application in controlling the dispenser.

Yeah, I pretty much agree with all of that.

After State Street (which if I understand correctly only held that you couldn't reject a patent simply because it contained some software), there were a flood of patents that were "X, but on a computer". Here's hoping we don't get a flood of "X, but using ML". (Uh, that is, if we haven't already had that flood...)

Combining transferrability and making patent dirt-cheap to acquire means that most patents will end up with companies that won't put it in productive use (most recent example, see Lecun's account for convnet patent: https://twitter.com/ylecun/status/1412549942229487620). The hoarding behavior is killing innovation.

Two ways to solve this: 1). limit transferrability, for example: patents can only be owned by natural persons and cannot be transferred to other persons or legal entities; 2). making it very expensive to keep the patents.

> Two ways to solve this: 1). limit transferrability, for example: patents can only be owned by natural persons and cannot be transferred to other persons or legal entities; 2). making it very expensive to keep the patents.

Or require active productive use after a certain time period to maintain the patent, at least for certain classes of owners [1].

[1] ...because small-time inventors will probably have more trouble putting their inventions to productive use, and you don't want to disadvantage them w.r.t. some big corporation (e.g. if all patents expire in five years without use, the corp may just decide to wait instead of paying). This probably could be phrased as an exception to a general rule for entities that only control a small number of patents.

This link from the article explains how to submit your comments:

https://www.federalregister.gov/documents/2021/07/09/2021-14...

The USPTO will not be reading our comments here, so if you feel strongly about this issue, please do consider taking the time to submit them formally as well. There will very likely be astroturfing from supporters of restrictive laws, and while they’re good at detecting that, it still makes a difference!

Not in principle; but the current practice of accepting highly dubious patents with little or no scrutiny is only serving the least innovative companies.

The fix would be to make it more expensive to defend a patent and cheaper to challenge one instead of the current practice of flooding the system with dubious patents knowing full well that challenging them is prohibitively expensive and the scrutiny from patent offices is minimal. When companies exists whose sole purpose is accumulating and exploiting such patents for profit with absolutely 0$ budget for R&D, something is deeply wrong.

The fact that some judges seem to be unwilling to actually lift a finger to do anything about such companies is not helping either. E.g. Texas seems to be a preferred venue for patent extortion where patent trolls and judges pretty much openly work together. Some call this justice, I call it institutionalized corruption. Plain and simple. Maybe no money changes hands but it seems corporations are actively putting their fingers on the scales when it comes to electing judges into very lucrative jobs and everybody understands there's a quid pro quo. Just putting a stop to that would be helpful. Something is deeply wrong when shady companies can just buy "justice" like that.

This would incentivize companies to file better patents and only file those patents that they are actually willing to defend as opposed to filing numerous patents opportunistically in the hope that smaller competitors will settle out of court and pay a license fee rather than risking bankruptcy through the courts. This also would free up patent offices to actually do their job properly as opposed to dealing with a never ending flood of largely BS patents. Rejection rates ought to be much higher than they currently are.

Set a cap on the number of patents issued each year. Like 500, max, across all industries. Or set a limit of 10,000 active patents, with new ones only accepted as old ones expire. If you want a chance at a patent you file with the PTO (under NDA, in case the patent isn't granted) for a nominal processing fee and at the end of the year the best 500 are selected from the perspective of "most useful disclosure of information", "least likely to be reverse-engineered or independently reinvented", and "lowest cost to the public in being restricted from freely implementing the patented invention for the next 20 years".

Variations: Let the filer pick the duration of the patent, with a shorter duration improving the odds of being selected (lower cost to the public). Or replace the monopoly with a one-time cash payout to make the invention freely available for the public to implement immediately.

Yes and no. There are countless innovations that have been lost over time (I.E. Ancient Roman's had stainless steel but it was lost), some were rediscovered some weren't, patents hope to prevent that by giving some incentive. This hasn't always worked though, for example Philo Farnsworth basically invented the modern TV but the large companies just waited for his patent to expire before moving on thus making it so he never really got paid anything but to contrast that the expiration on patents for 3D printers is what's causing so much innovation in that space today.

The main problem is when stupid ones (like the same thing as before but now with software) are granted it drags everything down, and the very concept of patent trolls is a destructive force not contributing to humanity at all.

Combine that with how most patents are forced through by submitting it until it sticks and it's a pretty broken system tbh.

I can't speak for every industry but in the software industry, they are 100% counterproductive and harm innovation. It's good that many countries do not recognize software patents.

I see little evidence of that latest claim.

China doesn't recognize software patents and it's rife with cheap copies and clones of existing products and produces very little innovation.

The US recognizes patents and is the most innovative country on many fronts (not just software).

Australia and Europe do not recognize them either. Australia is pretty innovative in the software industry considering its small population (which is less than 1/10th of the US).

Europe doesn't have software patents neither and you have lots of complex teleco hardware.

>The US recognizes patents and is the most innovative country on many fronts (not just software).

how?

Semiconductors are in Taiwan, South Korea and EU tooling

Is there more bleeding edge tech than Semiconductors?

Simplest fix to me would be to reduce the time of validity. We aren't in the 19th century anymore and extremely complex things can be build within months.

Rent seeking kills innovation. Patents as currently legislated, issued and resold are overwhelmingly just tools for rent seeking.

There probably is a theoretical way to do it that encourages innovation but it’s hard to see a society ruled by self serving individuals not wrecking it pretty fast.

Bounties would be much better than parents, across the board.

In a world with too much consumption and work for no good reason, where new products fight for limitted demand, parents in theory even are solving the wrong problem.

Bounties mean the goals are defined up front, and the reward for putting a round peg in a round whole is you don't even need a business plan, equity-based investment, or any of that other crap.

Take some funding to hit the bounty, everyone cashes out, move on to the next thing. All the benefits of "aquisition culture" without the problem of concentrated ownership. (The bounty giver should ideally be the state and not own the thing.)

”The study particularly wants to find examples of where a patent has been denied in the US but accepted in other jurisdictions, which will be a key area for European electronics companies. […] The study is looking at patent prosecution strategy and portfolio management as well as research and development, employment, procurement and marketing. It is also interested in the impact on the ability to obtain financing from investors or financial institutions and investment strategy as well as the impact on product development, innovation and competition.”

Sounds like they might think it’s _too hard_ to get patents in the US?

I've lived through a few instances where software patents fostered innovation.

We had implemented a certain GUI without realizing that we had copied an existing one. It was completely accidental, the kind of thing that happens without really thinking about it. We received notification from our competitors that we had to change our approach because they had a patent for it.

We did. It set us back a whole month but at the end of the day, we ended up coming up with a solution which I think we superior in all the ways.

So the software patent worked here: it forced us to not be lazy and come up with an innovative approach instead of copying an existing one.

Your being sarcastic, right?

A system like patents needs to demonstrate a clear and very large benefit to offset the fact that it removes people's freedom to implement ideas of their own inventions, to implement their own thoughts. That should be a hard thing to take away. I don't think patents demonstrate a large enough benefit to justify that.

My story is an exact contradiction to your last point.

By the way, you're*.

Your story absolutely isn’t a counterexample.

It’s blind luck that the situation you described led to innovation. Even if it did lead to innovation, it still wasted a month of time that could have been invested in more impactful innovations. You were still forced to hold out on shipping for a month entirely due to the patent system.

The alternate ending to that story could have very easily (and more likely) been “we had to scrap the feature because there was no way to do it without violating the patent.” Or the result could have been that they only alternative you came up with was worse. Or the result could have been that you needed to pay the other company a licensing fee.

Essentially you just got lucky that the competitor didn’t think of the best idea first, because if they had you’d be forced to ship something inferior.

For example, Amazon patented buying something with one click. How do you propose innovating around that? You can’t, you have to license the patent from Amazon or add another step to your checkout process. There’s only one way for one click to buy something.

It really does come down to the particulars here, but there is serious value in making UI consistent between applications. Every time a user comes to an application they have to figure out the primitives from the ground up. Where things go and why, in what order, what it does, and what the names for everything is.

You copied a competitor (unintentionally), and now when someone comes to your application from the other side, they have to work out how to work things.

Your UI has to not just be better, but better while accounting for the relearning users must do.

Imagine if there were patents on the File menu, for example (knowing US patents, I'll bet there's a menagerie of them, mostly owned by FAANG in their MAD warchests). I'd prefer devs not get creative here, unless the gains are substantial.

When you say copy, do you mean someone saw the other one and copied it, or did they independently come up with it?

Funny how they ask it now that China took the lead in patents.

Why does US only recognize a bad idea when they end up on the receiving end?

I don't think China has anything to do with it. People have been complaining about patents (and especially abuse of patents) for years/decades.

Yes, ever since the beginning. Which is at least 2 decades. And until now, it was totally in vain.

Just watched this video [1] by someone who got sued because his "android app uses a license server".

Those patent troll shell companies sue one victim after another and forward the incoming money to other companies. So there's no use in going after them directly. The victim can negotiate better terms if he signs a non-disclosure agreement. The patent trolls are protected by politicians, judges and lawyers.

[1] https://youtu.be/sG9UMMq2dz4

For those who read the comments first / only: The article is really short with no fluff or filler.

Part of the problem has been that, the way the system is currently set up, after an inventor invents something genuinely novel but with a narrow focus, their company's patent lawyers massage the claims language to make the patent much broader, sometimes effectively claiming any possible way to solve the problem instead of just claiming the specific invention. These ridiculously overbroad patents are the ones that are most valuable to trolls, and it is only the claims that matter.

General question by throwing ZFS into this discussion: I never exactly understood what's the problem related to ZFS' adoption in Linux (problem related mainly to the ARC algorithm?).

I read e.g. here ( https://www.legalmatch.com/law-library/article/what-cant-be-... ):

> You cannot patent a formula. However, you can patent an application of that formula.

So... with "application" does that mean that Oracle/Sun patented some piece of ZFS (e.g. the ARC mentioned above) in the context of "IT data caching technology" and that therefore anybody that wants to use it in IT (even by writing the logic from scratch) in any area (programs, databases, filesystems, etc...) cannot without paying fees, but if I use/implement that ARC logic to e.g. handle my inventory of "shoes to rent" (dummy, but let's say that I have such a shop which has multiple rooms used to store shoes and I keep in the main one the shoes that have been most frequently & recently rented, in the other rooms other stuff split by the same ARC-logic) then I'm OK because the "application" is different than what is mentioned in the patent?

Thx :)

> I never exactly understood what's the problem related to ZFS' adoption in Linux (problem related mainly to the ARC algorithm?).

The problem with ZFS adoption in Linux is copyright, not patents. The license used by the ZFS code and the license used by the Linux kernel are incompatible, so one cannot distribute the combination of both without violating one of the licenses. This means that the ZFS code cannot be incorporated in the Linux kernel, so it's doomed to forever stay out-of-tree.

This could in theory be worked around by a clean-room reimplementation of the ZFS code, keeping the same filesystem format, but I suppose most people who could do that work either are already working with the ZFS code, or find it more interesting to create their own filesystem. It would also be a lot of work, and take a long time until the reimplementation would be considered stable.

Aha - got it, thank you :)

I wonder if a system where keeping a patent active requires paying an exponentially increasing annual fee would help. So anyone could file a patent for say $1,000 in the first year, but then it would cost $2,000, $4,000, $8,000, etc. to keep it active in subsequent years.

That way, it would be financially infeasible for even the largest companies to hoard patents for years, but an individual inventor could start small and pay the fees as their business grows.

In my opinion, yes. It's not obvious to me that our intellectual property system encourages invention. After all, every great invention in history was "patented" (given state license to monopolize) and immediately copied just outside of its jurisdiction (and many times within the jurisdiction in complete disregard of the legality).

No company on earth would accept a state license to monopolize if it meant they were subject to price controls. Especially given how useless these licenses are at preventing global mega-corps from using them.

---

Intel and AMD famously rip each others intellectual property off but don't do anything about it because they're both guilty as sin. How does that protect innovation? How does these two companies ripping each other off inhibit innovation?

Basically every smart phone is identical. These companies occasionally fight but not meaningfully. But if I were to create a new smartphone I would be sued into oblivion. How does that protect innovation?

---

China has an incredibly agile and innovative economy. Are they the strictest IP protectors of all?

---

Intellectual property exists for the big guys to keep the little guys down. Not the opposite. Happy to be proven wrong. Every argument above was made with 0 evidence just personal observation.

The name for this is cross-licensing. If you have some patents which could affect someone else's business, and they have the same to you, you can cross-license each others' patents. Anyone not party to the agreement is fair game.

https://en.m.wikipedia.org/wiki/Cross-licensing

When I worked at Dell they were giving out cash bonuses for people to submit patent ideas with different levels of rewards depending on how far your patent idea got in the process.

The crazy thing is, some people (mostly the Indian cohort) were flat out coming up with idea's and submitting them to the internal patent committee every week/month since the amount offered for them would be the equlivent of 6 months/1 years salary.

Most innovations happen in small companies. When they happen to develop a successful product and patent it, frequently a larger company just steals the idea and develops their own product. As a small company there really is not much you can do about it. Legislation is hugely expensive. Because most large companies have a huge patent portfolio they will probably be able to show that you are infringing on their patents so you are not very likely to win the case. If you are very lucky they will take over your company or offer you some money.

The only real alternative is to stop producing and become a patent troll that sues other companies that use your invention (but that still requires a lot of money).

In theory, patents could be useful, but in the current situation there are far too many patents for completely trivial things.

Patents were created to encourage inventors to explain how their inventions worked. Instead of hiding their secrets, inventors simply published how their inventions worked and gained a monopoly to sell their use for a few years. A very noble intent.

Now, we're in the 21st century. We are no longer in the years of inventors like Graham Bell, Thomas Edison, Guglielmo Marconi, Nikola Tesla... Inventions today are discovered by large groups of researches who get their money from many sources private and public. These researchers, the real inventors, often get just a small part of the profit gained from their inventions. The largest parts ends up with companies which get giant profits and benefits from the lack of competition.

While patents were invented to guarantee that small individual inventors could be protected from giant companies attacks and encourage information sharing, patents these days do exactly the opposite of that. Apple being able to patent the "genie dock effect" is a clear example that the "patent system" is broken.

I was under the impression that Edison was actually more how you describe a 21st century shop, and was more of a CEO type, but I admit that this isn't based on deep knowledge.

I often wonder how many of our rosy perceptions of the past are due to a lack of granular information. we kind of know that those exploitative structures exist now because we have access to perspectives at all levels; an inventor or other employee can tweet about this stuff, but if I had created something for Edison and he stole it, he has access to newspaper publishers and I don't.

The virtue of the patent is that it rewards invention, and its productization, in cases where either the invention itself or its productization would not have otherwise been profitable, either because of the difficulty of inventing it, or because it isn't affordable to be the first to develop it.

This suggests a very natural rule of thumb for patent admissibility: would the patented idea reasonably have been independently invented or productized prior to the date that the patent expires? Invention and productization are very different roles here, so perhaps an invention patent could last up to, say, 15 years, whereas a productization patent would only last half that.

So a “one click shopping” invention patent would have to argue that within 15 years nobody else will have the idea, and a “one click shopping” productization patent would have to argue that within 8 years of the idea being public but not acted on, nobody would have been willing to front the cost to productize it.

This allows precisely those patents which one would expect provide value, and not the rest.

> This allows precisely those patents which one would expect provide value, and not the rest.

That's only true if we assume you have a handy oracle available to determine what actually would happen if the invention were neither published nor productized by the original applicant or their licensee. Otherwise the patent office is merely speculating without any evidence on which to base their decision. Once the invention is out in the wild there is no objective way to determine whether it "would have been" independently invented or productized by someone else, so you can't even get feedback (15+ years delayed) on how accurate your speculation has been.

For sure the actual award would be a matter of judgement, as would any challenge in court. But right now patents are judged on whether they were “obvious to the person skilled in the art, at the time the application was filed”, which is both no less a matter of judgement, and a bar much easier and fuzzier to cross.

This is what fundamentally results in patent races, where any ideas that can claim to take any nominal amount of intellectual effort to think of must be rushed to the patent office before the competitor discovers it themselves. Because it doesn't matter for elligibility if three people all thought on a problem simultaneously and all stumbled on the same answer, as long as that answer took at least a little time to discover in each case.

> But right now patents are judged on whether they were “obvious to the person skilled in the art, at the time the application was filed”, which is both no less a matter of judgement, and a bar much easier and fuzzier to cross.

Yes, and right now about the only way to fail the "novel and non-obvious" requirements at the time of application is to try to patent something which already exists. (And even those cases sometimes slip through the system.) A patent might be invalidated as obvious later if someone else comes up with the same thing and can demonstrate that they had absolutely no knowledge of the patent or anything covered by it, but that's both unlikely (for products generally available to, and discussed by, the public) and an extremely hard thing to prove even if true.

In the absence of any possibility of hard evidence to the contrary, the default assumption is going to be that the invention would not be independently invented or productized by someone else. This is basically how things stand now with the assumption that the invention is non-obvious. Getting this overturned would be at least as hard as getting a previously issued and productized patent invalidated for reasons of obviousness following "independent" reinvention.

> In the absence of any possibility of hard evidence to the contrary, the default assumption is going to be that the invention would not be independently invented or productized by someone else.

Obviously if the patent office ignores the change, the change wouldn't help. That's a fairly meaningless analysis though. I am not saying the patent office should deny specifically only patents they can definitively prove will be invented in the next 15 years. I am saying they should use that as the line by which they estimate novelty. Patent applications are not like criminal cases, there need be no beyond reasonable doubt.

I'm going to go out on a limb and say that the actual purpose originally intended for patents was to get people to describe and record innovative discoveries, and eventually add them to the public domain for others to build off of.

I think transferability, renewability, and gaming of the definition of "novel" and "non-obvious", is actually what kills innovation.

The transferability is even more of a sticking point in my mind, because of it's endemic exploitation in the "automatic assignment of patent to employer" sense for disenfranchising creators of unrelated or only tangentially related work off the company clock. Then again, I can't offer a more elegant system to replace it, so I kind of have to take a Chesterton's fence-pill on that one. And besides which, having worked with engineering management, it does kind of degenerate down to "inform legal, and have them draft a waiver for an exec to sign stating the company waives interest", which is a functional, but less than ideal compromise.

Having spent a significant amount of time developing and patenting new technology — I think patents are both necessary, BUT need an overhaul.

Particularly, I think design patents are likely positive, but I think utility patents need a different mechanism. For instance, I think the patent should work for X% of revenue of a given item is provided to the patent holder, provided they don’t reach an agreement. This would ensure anyone can use patents and the revenue would be defined. I also think a patent should be limited to 3 years without it being put to use. “Use” should have a lower bar, but effectively made or reduced revenue in some manner.

In reality, you can build what you and ignore patents. Patents don’t matter until you are in business and make money. At which point, if someone uses you; you’re already successful and they’ll want a slice of that success. This is the advice I’ve always been given (from investors and attorneys, off record).

In the software domain, I feel like patents have mostly become a corporate pre-emptive defense mechanism, i.e. "don't sue us for your patents we're infringing on, and in return we won't sue you for our patents you're infringing on".

It can be even more defensive than that. "We're patenting the things we're doing, so that nobody else can patent those things and then sue us for them."

That's, um, not "promoting the progress of science and useful arts". It's at least not trying to exploit others, but it's sheer waste, having to protect yourself from being exploited.

The main problem with patents isn't their use as "first strike" attacks but rather their use for "defensive attacks". A big corporation can amass a huge collection of patents covering random trivial capabilities while a startup can't. This also lets the big corporation copy the innovations of a startup with impunity, because the startup doesn't have the resources for a drawn out legal battle, and they would fear the counter suits based on that patent portfolio. Someone else in this thread mentioned Amazon patenting a search field auto complete - that's the kind of ubiquitous functionality that could be weaponized in such a scenario.

I wonder if the USPO is asking about whether the idea of patents kills innovation, or the implementation does (e.g. how easily patents are granted, which things are eligible for patents (I know many people who think that patents are a good idea in general but generics and software shouldn't be patentable)).

Also, if patents were removed entirely, wouldn't that just benefit the largest companies? That is, whenever someone has a new idea, their idea would just be taken and most efficiently built by the largest companies, because those would be the ones with the most engineering, manufacturing, and marketing resources. What's to prevent that from happening?

I wonder if we removed the ability for companies to patent, only individuals (or groups of individuals) would help. Limit the patent protection for lifetime of said individuals. I feel like that would remove the patent troll economy.

First, people that patent something should have to be actively working on the patent; not just sitting on it like a patent troll. Patent trolls stifle innovation.

Second, it should be easy for an inventor or developer to go create something new without having to look through a bunch of patents and see if someone already thought of something related to his idea. The more roadblocks in the way of development, the less development occurs.

The people: yes

Corporations with billion-dollar gold-plated megaphones: NO, THE PATENT SYSTEM PROTECTS INNOVATION, NOTHING TO SEE HERE, MATTER OF FACT WHY DON'T YOU MAKE IT STRONGER SO WE CAN BE MORE INNOVATIVE

Patents do kill innovation. Why do you think we've seen no new aircraft carrier manufacturers come out in years? Because Boeing owns all the patents. Why do we see such fast innovation in so many competitors in any new space like search engines, then it seems like one company solidifies the whole market? That company has patents that prevent other people from reasonably competing.

The entire patent system needs to be simply scrapped.

Ideas are worthless, execution is what matters.

Patents should only be enforceable for:

- companies that have a history of utilizing/operating with the patents subject (e.g. no patent troll legal firms would be able to meet this requirements)

- only be enforceable against competitor companies that are larger than a certain size, e.g. more than 10 million ARR, so smaller companies can still be formed and get off the ground

- be different based off of the domain - e.g. it should be very hard to get a patent for software or related items

We need to do what we did with pharmaceuticals - provide a period where the patent has to be licensed so that the owner can be remunerated for their research. Should that be 7 years, 10 years, 15 years or more - I don't know and that's a good discussion to have. Meanwhile I think we can all agree that patents should have an expiration. That seems to strike the balance of incentivizing innovation while protecting research investments.

Wondering if it be viable to train a classifier on a set of patents that are known to be vague, abused, and pernicious and see what else it digs up.

This works for malware, and simple things like being owned by known patent trolls, who have selected them for their crappiness are easy indicators of what a bad one looks like, a history of litigation, clusters of ones with similarities, etc.

Surely this is somebody's ML driven patent troll startup right now?

Once phonorecords of a musical work have been publicly distributed in the United States with the copyright owner’s consent, anyone else may, under certain circumstances and subject to limited conditions, obtain a “compulsory license” to make and distribute phonorecords of the work without express permission from the copyright owner.

U.S. Copyright Office, Circular 73 - Compulsory License for Making and Distributing Phonorecords

---

Maybe something like this, but for patents?

Puget Systems popularized the concept of the submerged water computer/aquarium computer.

They released a video showing a proof of concept, but it had an obvious design flaw.

Shortly after they released that video, a patent firm patented the same thing, but with the design flaw fixed.

Some time later puget sound announced version 1 of their fish tank computer, obvious design flaw fixed, only to get forced into discontinuing it because of said patent.

So i'd say yes.

For AI/ML related patents, I’m not sure innovation is stifled insofar as people mostly don’t pay attention to them.

Most of the techniques that _every_ company is using are patented, but no one seems to notice.

Admittedly I’m sure there are cases where it did stifle innovation in this space, but I think on average perhaps not.

It’s interesting to see the comments on eink patents — I was never aware of how much it stifled innovation in that space!

> It’s interesting to see the comments on eink patents — I was never aware of how much it stifled innovation in that space!

I'm really curious which comment that people are referring to because I happen to work in the display industry. I'd never heard anything like that amongst my peers or during conferences. The main place I've seen this refrain is on HN. Please have a look at my comment history as I keep asking about which patent people are talking about. I'm still trying to figure out if the issue is real.

I have to believe that there's been research on this in the economics literature...right?? Like, maybe research on "patent trolls" specifically hasn't been studied (or perhaps it has), but there has to be research on the economic impacts of patents, and the dynamics both within and between countries. I'd be astonished if economists haven't looked at this.

I can't think of a single patent in software that actually had a positive impact on the industry. Copyright provides enough protection for any actual investment in software.

Other industries where R&D cost is high and the end result can be very compact (a chemical formula, a new material, etc.) can benefit from patents, but that R&D cost needs to be verifiable and substantial.

I see that companies get lower taxes revenue linked to patents. Isn’t they then incentivized to fight for the patents and “protect” every part of the company for tax reasons? So the real fight is against tax attorneys, because if the patent is not valid the governmental tax attorney can argue that you should pay more tax?

So it’s basically noting to do with protection against competition

They should start by making patent trolls illegal. Also shorten the patent expiry to 5 years instead of 20. If the patent holder is actively using it, it should be enough time for them to recoup their R&D costs and make profit.

Of course that's never gonna happen because too many people stand to lose - too many vested interests, bought out politicians, armies of lawyers, etc.

I'm assuming in this 20 -> 5 transition that would be 5 years after the patent is granted rather than the current 20 years after filing? Part of the issue is that a patent can take 5+ years form filing to grant, and frequently take 3+ years.

My patent attorney used to tell me: "Patents are like nuclear weapons. I want to be on the side that has the most."

Related: There's also a Copyright Modernization Committee at the Library of Congress, and Brewster Kahle is joining it! They also have virtual public meetings, including one on July 22.

https://news.ycombinator.com/item?id=27732887

Make a patent and copyright expire after 3 to 5 years. This gives the holder enough time to go to market if they choose.

Patents also preserve knowledge in publicly accessible and well documented form. Without patents the only option to recoup investment in R&D would be to make innovations a trade secrets. This will lead to frequent reinvention of the wheel and lost knowledge.

The system is broken, but abolishing it is even worse

Patents have a good intent, the real danger is non practicing entities (patent trolls).

I think that you shouldn't be allowed to acquire a patent (via acquiring the patent owner) unless you are in the business of the patent, or you go to market with an application of that patent within a limited amount of time

There should be some kind of "patent reset". Let's say next year, all patents granted until now are null and void, and new more stringent criteria are applied to new ones. I would be real curious to see the results of an experiment like this - which to be clear, will never happen.

Patents in theory or patents in practice?

The patent system has been gamified, and is now optimized for the corporatocracy.

Patents mean that you basically need venture money to being anything to market. If you're a whiz-kid programmer who makes a new OS by the time you're 16, you're just going to get buried by patent lawsuits unless you can get YCombinator money to fight it.

Yeah, the current patent-system is completely ridiculous.

To pick out one especially bad problem: it's too inconsistent, relying on legal-battles to resolve inconsistencies. And then people are just afraid to do stuff, because they don't want to have to fight those battles.

Bragging rights plays a bigger role in patent filings than most people recognize. Yes there's some business value to Microsoft/Google/etc. having patents, but if you look at why a given team patents something it usually comes down to bragging rights.

HN loves to focus on only one of the normative goals of patents - to incentivize innovation. But there is another almost equally important goal of the patent system - to incentivize disclosure. While I agree that the use of vague language in patent publications perverts the public disclosure goal, other forms of public disclosure may arise from it. For example, companies can openly talk about inventions, whether it be with the public, potential business partners, or suppliers, without fear out misappropriation.

I am not a fan of a bright line rule that prohibits software patents. Software is so explicitly tied with most inventions these days that it would be hard to tease out what qualifies for patent eligibility. Is a novel design of a robot arm that contains some software elements patentable? Instead of a bright line rule, I’d prefer to see a strengthening of 112 - written description / enablement requirements and 103 - obviousness requirements. Imagine if 112 required software patent holders to provide a hard copy of the implementation code (or at least a more detailed description of the invention). And as to 103, the policing function of obviousness needs to be ramped up. The USPTO should invest in better prior art searching techniques and hire Examiners that care about enforcement. Currently, the incentives of USPTO Examiner's misalign with the goals of the patent system, as Examiners are evaluated via a point mechanism that rewards churning out patents rather than tightly policing them.

Also, when people on HN argue about the "patent troll" problem, the real issue seems not to be with patents themselves, but with how the U.S. litigation system can be weaponized to extract fees from using patents. Thus, I think the key issue is trying to figure out low cost and effective ways to litigate patent issues. IPRs have helped, but perhaps we can develop other ways for companies to cheaply dispose of garbage patent suits that have no merit.

Mandatory licensing. Rather than grant a monopoly, a patent could give you the right to a certain percentage of all sales. That would allow competitors to serve markets that the patent holder refuses to address (ie cheap 3d printers.)

Does China or Russia sponsor any US patent troll companies? And if the answer is no, why not? It seems like a dead simple way to stifle American innovation. Use our own toxic legal system against us.

Patents encourage innovation because without them, the moment an inventor's idea gets out, it will be copied and monetized by some corporation with the resources to bring it to market, without compensating the inventor in any way.

Now idiotically evaluated and granted patents (in particular software patents, but not necessarily just software patents) stifle innovation.

Crowd-sourcing this kind of question just seems like ploy for the USPTO to make even less of an effort.

Suppose that the answer from the crowd is "yes": patents in those areas kill innovation. What can the USPTO do with the information, other than just not hand out patents in that area? Summarily tossing a patent application takes even less effort than slightly thinking about its merit.

> What can the USPTO do with the information, other than just not hand out patents in that area?

They could require the IP holder to declare the price at which they were willing to permanently release the discovery into the public domain in exchange for a one-time payment, and levy an increasing maintenance fee on a % of that price due at the end of each financial quarter.

This may eliminate unnecessary court costs and legal fees and supply chain interruptions resulting from legal disputes, while still ensuring that the original inventors are compensated and that there is an incentive for disclosure.

Suppose a defendant is notified of patent infringement and pending legal action. Instead of hiring a lawyer or halting production they could crowd-source the money to pay the inventor's fee through the patent office. The patent office then forwards the payment to the original inventor, the patent is destroyed, and there is no basis for further legal action.

Large companies may find that it is always cheaper to pay the patent office to permanently destroy each other's patents than it is to go to court, which would also level the playing field for smaller firms and startups as the discoveries would be placed in the public domain.

Why not just do that for all patents from now on? It seems like a good idea.

Why aren't patents just "use it or lose it" like trademarks? That would seem to eliminate most of the abuse in a targeted way. Unless I'm missing something...

Because sometimes you need a lot of money to setup manufacturing. The covid vaccines are easy to make in a lab at a rate of 10/week. To make them useful though you need the money to setup manufacture them at scale, and one easy way to get that is to sell a patent. Use it or lose it means anyone who might buy the patent is incentivized to wait until you lose it and then use it for free.

I think the bigger question is, will this be more than an exercise in confirmation bias, and even if it makes a valid determination, will it actually lead to policy change?

The key to the patent system are the patent lawyers, not the innovation per se. As long as the patent lawyers revolve with the USPTO, the system will not change.

Please do not be naïve.

A patent without a working prototype is meaningless and just weapons for patent trolls.

Patents last too long.

The time it takes to scale up an operation has gone way down, and the patent length should correspondingly shrink

No they don’t. But the process and cost is not working for smaller inventors.

Matt Ridley seems to think so in his book on innovation.

Isn’t that kinda the point of a patent?

Patents kill innovation, certainly, and at an ever increasing rate. The government acts as the enforcers for monopolies, as it always has been. It is based on an appeal to violence, like everything government does. It is, and has always been the principal enabler of every genocide, war, and mass abuse of human rights in history: educating people to recognize the claims of government authority as logically invalid (appeals to the populace, appeals to violence, etc. etc ) is needed.

innovators ship

and innovative entrepreneurs don't let their competitors see them as competition

Some ideas, based on some of the ideas kicked around in this thread: What about something like compulsory licensing?

Copyright has something like this[1]. One can acquire a license to distribute a copyrighted song in some circumstances if you pay the mandatory fee.

So how would this work with patents? First, let's cover the non-practicing entity, or NPE. Patent trolls are the most well known and hated NPEs but they're not only ones. You can have a small time inventor who's successfully invented a thing, designed and built a prototype, but hit a wall when it comes to the capital-intensive requirements of moving on to actual production. As much as I might want to stick it to patent trolls, I don't want to catch original inventors in the blast radius. So it can't be something that just punishes all NPEs, but just the parasitic subsets.

So Bob discovers novel and efficient way to frobnicate foobars, and gets a patent on it. But he never takes it to production. Let's say materials science isn't quite at the point necessary that the materials needed for the frobnicating foobar can be readily had cost-effectively at scale. So he sits on it.

Some years later, after some improvements in materials science, it's now feasible to build frobnicating foobars at scale. And Bazcorp, a company that's in the foobar business learns of Bob's patent, gets a compulsory license to manufacture it, and starts building and selling frobnicating foobars. Bob gets money for the frobnicating foobars now on the market, Bazcorp gets to sell frobnicating foobars, and people get to buy frobnicating foobars. Sounds win-win to me.

Now, how keep patent trolls from perverting this? Make it so the fee schedule for a compulsory license is lower if the entity holding the patent gets more than, say...10% of their annual revenue from license fees. Make it so the liscencing fee schedule goes lower the older the patent is. You could combine this with annual fees to renew the patent, escalating each renewal period that the invention in the patent is not being produced, making it more and more expensive to hold a patent on a thing and not be building the invention covered. Then add a multiplier upon the renewal schedule if the patent holder gets more than 10% of their revenue on patent licensing.

So Troll, Goblin & Associates, LLP, seeing the hot market for frobnicating foobars, does some research and discovers building them is covered by a patent they hold on the quux process. Said patent is 10 years old now, and has cost some coin to hold on to, so it will require quite a lot of money to turn a profit on it. So, TG&A goes to Bazcorp and demands one million dollars per unit. Bazcorp calls TG&A's bluff and files for the compulsory license of $0.10 per unit. Now, with the volume of frobnicating foobars that Bazcorp is moving, TG&A still gets a not-insignificant sum of money, but not enough to break even on the purchase of the patent and fees to renew it. Too bad, so sad, such is the risk of TG&A's non-creating, parasitic business model.

Another idea: What about something like jury duty as assistant patent examiners? The patent office could reach out and compel persons skilled in the relevant sciences to assist with patent review and be remunerated for said assistance. Then, you might actually get some qualified eyes upon patent applications that could spot prior art or obvious things and reject said bogus applications.

1. https://www.copyright.gov/circs/circ73.pdf

Yes.

Yes. Duh.

Yes sir. Them patents certainly do.

Not for medicine though.

Once a molecule is researched and found to work, it should be patented.

Although not if it receives public funding.

The weakening of patents has led to a rise in extreme trade secrets. Now, we have no idea how anything recent works. Nobody publishes much on self-driving. Nobody outside Tesla really knows how Tesla cars work. Progress relies on reverse engineering and teardowns.

> Progress relies on reverse engineering and teardowns.

Progress has always relied on reverse engineering and teardowns. (Or independent reinvention.) You don't learn how anything really works be reading the patents, which tend to be deliberately obfuscated if not outright misleading. Anyone working in the industry will be instructed not to read patents anyway for fear of a willful infringement claim and trebled damages. Only in very rare cases can anything of notable value be maintained as a closely-held trade secret for longer than the duration of a patent.

We definitely need patents to protect intellectual property.

There is a lot of hard work and resources needed to actualize products that can be easily copied without any effort. I have attempted a few patents in my time but it was too expensive and the process itself was daunting.

Now I use secrecy to hide critical methods and open source everything else.

Can you find examples of patents killing innovation? Sure. But the idea of a patent is to give the inventor time to bring the invention to market, and it would be impossible for any inventor to have the chance without the protections offered by patents. Show me an inventor that hasn’t obtained a patent that successfully out competed an established competitor that freely copied the inventor’s invention.

The USPTO doesn’t care about innovation, and it is not their job to care about that…so why the question and why now? Lobbying and special interests of the big companies.

If anything get rid of patent protections for big businesses, it’s not like a big tech or big pharma patent is what’s keeping the small guy from competing with them, but if the small guy can not protect their IP from being copied by big businesses then they will never be able to compete.

Patents foster innovation.

Without patents people who are technically skilled but socially inadequate would not ever try to build stuff.

Same goes for those who can promote themselves but just love the technical challenge much more and want to focus on that.

These are among the most productive demographics and without patents they'd be sitting idle because they would be too scared of working thousands of hours in vain, just setting the stage for the socially competent people to rip them off and cut them out of the enjoyment of the monetary and social rewards for the thing they have built.

At least with patents there is some sort of protection or I should say an illusion.

Say you have an idea and then one of the big guys with deep pockets and a huge marketing and PR department steals it from you or executes in a way which becomes world class, with a patent you have grounds to sue and you can also go to lawyers and convince them to work pro-bono because they'll get so much media exposure.

You can also go to a bank and ask for loans to promote your story and you can get a favorable interest rate due to the fact that you have the optionality to settle for cash.

A patent gives you options. It's also a concept which unlike so many corporate jargon terms is easily understood by the general population, and that is an advantage when you are a small startup or individual inventor going up against a behemoth...it enables you to convey social hatred against the giant company

A much more complicated instrument such as NDA/NCA gave the Winklevoss bros the ability to sue Zuck and to put themselves on the map with the movie "The Social Network", a patent as I said is much more clear and easy to understand, both for the courts and the court of public opinion

Patents also mean that the big co that copied your idea can sue you out of business for violating one of their myriad of patents, and that your patent is actually a derivative of one of theirs.

You want to drag the big company in front of the court of public opinion.

If your patent is perfectly nailed, then your version of the story will win.

In short you gotta have something to show to the journalists and lawyers who are your allies because they want to be the guys bringing down the behemoth and win the Pulitzer/lawyer of the year award.

If you don't have anything they won't write your story and they won't defend you.

Once the story is out there the entire public opinion will be on your side and you are looking at the very least at a hefty settlement which the behemoth would pay you in order to make the story go away.

A patent is a checkpoint for an individual inventor or a small startup

Reply on news.ycombinator.com