Hacker News new | past | comments | ask | show | jobs | submit login
Don't Let Congress Hand Patent Abusers Their Ultimate Wishlist (eff.org)
248 points by DiabloD3 on Sept 17, 2019 | hide | past | favorite | 51 comments



The broader issue is that the USPTO will let you patent anything. They have a long review process, but so many stupid things get approved that the agency is a sad joke.

The latest example:

https://www.eff.org/deeplinks/2019/07/someone-suing-companie...


The patent examination process is not well understood by most people.

Once a patent is received, the patent examiner carefully places the application into a room full of other patent applications.

Then kittens are released into the room with PATENT GRANTED stamps affixed to their feet.

The kittens are then returned to their holding area to await the next round of patent examination.

The patent examiners collect the applications from the floor and look to see which patent applications were granted.


The use of kittens was specifically specified by lobbyists as they knew that would make it impossible to change the procedure. Who’s going to vote against kittens? Squee!


This is easily the best HN comment of 2019. It's also 100% true and completely accurate.


Nice one. How about this Google patent for "Generating output sequences from input sequences using neural networks" http://www.freepatentsonline.com/10402719.html


This is what is protected by this patent:

"A method comprising: obtaining an input sequence having a first number of inputs arranged according to an input order; processing each input in the input sequence using an encoder recurrent neural network to generate a respective encoder hidden state for each input in the input sequence; and generating an output sequence having a second number of outputs arranged according to an output order, each output in the output sequence being selected from the inputs in the input sequence, comprising, for each position in the output order and beginning at an initial position in the output order: generating, using the encoder hidden states, an attention vector for the position in the output order; generating, using the attention vector, a softmax output for the position in the output order, wherein the softmax output scores each position in the input order; determining, using the softmax output, a pointer to a particular position in the input order; and selecting, as the output for the position in the output order, an input from the input sequence that is located at the particular position in the input order identified by the pointer."


Why does the USPTO do this? Is it a technical problem or a political one?

I remembered reading a story from the early industrial era, about an inventor who challenged another inventor's validity of patent and pressured the patent office to hold a demonstration for verification before believing the invention is true.

Software patent is probably a culprit that makes some of the previously impossible patents possible by doing it on a computer. But I don't think it can explain everything. Why has the USPTO became something beyond stupid nowadays? I don't think the patent office verifies anything these days, they only log everything and left all the actual business to courts, I'm not sure it's the best system. Is it because the model of a centralized patent office simply doesn't scale well for the tremendous amounts of patent? Or because of lobbyists? Or both (e.g. patent office doesn't scale well and the lobbyists and patent lawyers exploit it to apply stupid patents)?


It's an economic one feeding a political one.

Defining de novo property interests without physical bound inherently creates a vested rentier class interested in the persistence of those properties.

New entrants to a field generally oppose patents. Incumbants generally support them.

I've watched numerous firms (Oracle, Microsoft, Google) transition from one side of this dynamic to the other. It's a tragic inevitability, in the classical Greek sense.

The (further) tragedy of the tech world is that it thinks this is a problem only with, take your pick, software or information-technology subjects. The problems with (and debates over) patents go far further back. Arguably, patents delayed the start of the Industrial Revolution by 25 years, as England awarded extensions to James Watt for his steam engine.

Which may, of course, mean we merely postponed the inevitable global warming by a couple of decades.

The histories of light bulbs, phonograph, telephone, radio, television, and more, are other arguments against patents.

https://www.washingtonpost.com/news/wonk/wp/2012/10/10/the-c...


On the flip side, Watts developed his engine using financing from a coal industrialist, John Roebuck. Would he have secured that financing in the absence of a potential monopoly profit on his engine?

The examples you mention illustrate the importance of patent rights to securing financing that bankrolls R&D. Bell had investors who pressed him to patent the telephone. Edison’s research was bankrolled by a series of contractual arrangements with companies like Western Union—where patent rights protected technology that he was demonstrating and trying to sell to Western Union. He also sold patent rights to Jay Gould. His Menlo Park R&D lab tested 3,000 lightbulb designs before filing the first patent. The same month as that patent filing, Edison founded what would become General Electric using financing from JP Morgan and others. Although you didn’t mention flight—the famous Wright-Curtis lawsuit actually involved two patent holders. Curtis was bankrolled by, among other people, Bell, and had its own aircraft patents. See here’s for more of the history: https://poseidon01.ssrn.com/delivery.php?ID=4980020990931060...

Patents almost certainly delay the widespread adoption of a new technology. That’s the whole point. But the counter-factual scenario is not, as you suggest, what would have happened had the invention not been patented, but what would have happened if patents had not been available. Would the alternative means that exist to finance R&D (government spending, mostly) have produced the invention in the same time frame?

Also, there is the problematic fact that monopolies are, for a time anyway, good for innovation. Someone else might have invented the telephone. But Bell’s monopoly ended up inventing the transistor, C, and UNIX. Someone else might have invented the copier. But Xerox’s monopoly ended up investing the GUI, email, and networking. GE, IBM, AT&T, Microsoft, Xerox, HP, Google-these are all companies that added way more to the state of the art during their quasi monopoly periods than at inception.


Roebuck could also have had the foresight that such a machine would increase demand for coal, making him a nice profit. I don't think this is a particular compelling example.


People will make money somehow or another, but it’s not clear to me that we want to encourage these indirect monetization models. That’s basically the Android business model—don’t make money selling the OS, make money selling the data collected when people use the OS. It’s better for Watts to have had a temporary monopoly on engines than to encourage schemes where investors try to profit from a general increase in coal use.


I am afraid we get both...

Otherwise, I enjoyed your insightful reply.


Steam power, light bulbs, telephone, radio, aircraft, television, and numerous other inventions were very much in the Zeitgeist, and the fact of numerous patents being filed (or warred over) means that at best the benefits you describe are at best zero-sum: the benefits accruing to one party are those which are denied another. From an investment standpoint, this means that any such backing becomes a gamble: will an innovation pay off if a patent can be successfully filed with primacy (see Bell vs. Gray: https://www.princeton.edu/ssp/trips/data/scientificamerican0...), given primacy, and defended.

James Watt through 1800 sold a total of 500 steam engines, all low-pressure vacuum-action (the work stroke came from the condensation phase, not from the recovery steam-entry phase), with typical ratings of 20 kW, several times greater than typical watermills of the time. Most were restricted to pumping water. Watt obstructed mobile applications (ships or rail).

Expiry of his patents saw high-pressure steam engines with far higher power, mobile applications (ships, rail), tremendous increases in power (to 1,000 kW by 1850). Innovations meant both larger and smaller engines were available -- the former capable of powering large factories, locomotives, and ships, the latter for small point-of-use power sources. Much as large and small electric motors provide both raw capacity and very specific localisation.

General history of steam and specifically of Watt's obstruction of innovation are covered well in Vaclav Smil, Energy and Civilization, pp 235ff.

Gregory Clark's A Farewell to Alms notes that few of the early industrial innovators, Watt included, made much return based on either patent royalties or patent-era sales, looking at key inventors of the early Industrial Revolution: John Kay (flying shuttle), James Hargreaves (spinning jenny), Richard Arkwright (spinning frame), Samuel Crompton (spinning mule), Reverend Edmund Cartwright (power loom), Eli Whitney (cotton gin), and Richard Roberts (power loom, machine tools).

Of the list, Kay, Hargreaves, and Roberts died in poverty. Crompton and Cartwright were granted substantial payments by acts of Parliament (£5,000 and £10,000 respectively), Whitney made money through arms sales to the U.S. government, and of the lot, only Arkwright earned significant wealth, half a million pounds, after his patents stopped being honored by other manufacturers.

https://press.princeton.edu/titles/8461.html

The case against patents generally is made in "The Case Against Patents", by Michele Boldrin and David K. Levine for the Federal Reserve Bank of St. Louis:

http://research.stlouisfed.org/wp/2012/2012-035.pdf

It's truly amusing that you should mention Unix. I'll just quote from Wikipedia as the facts are very well established:

The Unix operating system was first presented formally to the outside world at the 1973 Symposium on Operating Systems Principles, where Ritchie and Thompson delivered a paper. This led to requests for the system, but under a 1956 consent decree in settlement of an antitrust case, the Bell System (the parent organization of Bell Labs) was forbidden from entering any business other than "common carrier communications services", and was required to license any patents it had upon request.[6] Unix could not, therefore, be turned into a product. Bell Labs instead shipped the system for the cost of media and shipping.[6] Ken Thompson quietly began answering requests by shipping out tapes and disks, each accompanied by – according to legend – a note signed, "Love, Ken”.[12]

https://en.wikipedia.org/wiki/History_of_Unix

That is, AT&T were specifically enjoined from making profitable use of their inconvenient operating system, which instead developed under a general sharing and collaborative model of development, some at AT&T, much at Berkeley, MIT, and other (mostly academic) sites, and through such publications as Lions' Commentary on UNIX 6th Edition, with Source Code.

Bernhard J. Stern's "Resistances to the Adoption of Technological Innovations" (1935) details numerous instances and methods of such dirty tricks, including specifically use of patents in obstruction of innovation, and is rapidly becoming among my favourite references to these:

https://archive.org/details/technologicaltre1937unitrich/pag...

Markdown: https://pastebin.com/raw/Bapu75is

And though the headline story concerns trademark and noncompetes, rather than patents, "Battles over Barbie" describes largely the same dynamics:

http://blogs.wgbh.org/innovation-hub/2019/9/6/battles-over-b...


wow i violated this patent years ago when I made a microsocial network that used an email gateway to interact with users via sms. guess I should have patented it first ¯\_(ツ)_/¯


Edison spent a large part of his career:

1. being sued for patent infringement

2. suing for patent infringement

3. being hired by companies to find workarounds for patents

It's doubtful any of that was worthwhile in the end.


I am surprised to read this: his patenting of the stock ticker and the two-way telegraph is listed on Wikipedia as the key to his fortunes. Do you mean "worthwhile" as in wrt. society or for him?


That surmises Edison had to have patented those inventions to secure his future fortunes.

I'm pretty sure the comment you're replying too was attacking the worth of patents, not the inventions.


Hollywood was also born of people trying to get away from Edison and his patent lawsuits.


The US patent system is why the US lost 10 years of aviation development and the Europeans took the lead. The US arguably did not catch up until WW2.


Why didn't the problem reappear after the war?


The government forced them to create a patent pool.


Interesting.

Where can I read more about it?



The tech companies changed the game in their favor when they successfully lobbied and changed the law to first to file vs. first to invent.


"First to invent" doesn't make much sense when the point of the patent system is to get inventors to publicize their inventions in return for a guaranteed monopoly.

"First to invent" runs counter to the patent system.


First to invent is indicative of a policy that favors individual inventors over well-funded corporations.

Little inventors might like to shop an idea first and then file a patent application if they find traction or funding. US law allows for a one year grace period after disclosure. Fail to file within one year, all US patent rights are lost. Outside of the US, any public disclosure before filing bars a patent.

Big Corp has the money to file patent applications early before disclosure.

Now with first-to-file in the US, there still is a one-year grace period in the US, but the first applicant to file is the winner not the first to invent.

Among other things, this is bad for small inventors because bad actors can file patent applications on the small inventors' inventions first. Trolls can and do scour websites to identify inventions that may be worthwhile sniping. In the old days the inventors just need to show that they invented their invention before the troll filed their patent application. Now, you have to prove that the troll literally stole the idea which the USPTO doesn't really look into.


So, my understanding of first-to-file is that any public disclosure counts as prior art for anyone else. So the only time you would have to prove the troll literally stole the idea is if your disclosure to them was non-public.

In either case, it's likely that the USPTO isn't going to notice before issuing the patent, but under first-to-file it is way easier (read: cheaper) to show that "X document was published on Y website on Z date" than it is to start doing discovery to compare private lab notebooks or whatever to decide who really invented the thing first.

This only ever affected a minor number of cases anyway, and the change mostly just made litigation costs cheaper and more predictable.

It also seems really great for open source software (where I personally work), since everyone publishes everything they do immediately anyway.


First to file, simplified some processes and bring US patent law closer to international patent law. But you are correct that it removed what amounts to an edge condition.

Public disclosure is great for making prior art to block future patents. That's why some big corps white paper just about everything whether they intend to patent it or not.


First to file has nothing to do with prior art. It’s about what happens when two companies claim to have invented the same thing.


One might easily believe that based upon the name, but a number of changes to how prior art worked in the US were required to implement it. Please consider https://www.uspto.gov/web/offices/pac/mpep/s2152.html

In particular, the paragraph starting with "Although some of the provisions of AIA 35 U.S.C. 102(a) and (b) are similar to pre-AIA 35 U.S.C. 102(a), (b), and (e), the AIA has introduced a number of important changes with respect to prior art ..."

Edit: And more importantly, Section 2152.01 right below it.


Ah, you’re right. The critical date for a patent is now the date of filing rather than the date of invention. Apologies for jumping the gun—I misread your comment as suggesting that first to file somehow changed how you calculate the date as of which a non-patent disclosure (such as an article) is prior art.


No worries. It is rather the date that you compare it to that changed (filing rather than invention). I could have expressed that more clearly.

It actually does change the dates for when patent documents are considered prior art (as described in the first paragraph I referenced), so it is easy to get the issues confused.

The effects of first to file on how prior art works are, in my opinion, way more important than the litigation edge condition it eliminated.


I don't know how it works. But does 'first to invent' effectively start the clock earlier? If so maybe the public interest is still served. You could presumably license the patent soon after its invention and hopefully it's disclosed not-too-terribly-long after that relative to the duration of the patent itself (20 years IIRC).

But if it expires sooner now, we can enjoy it royalty-free sooner.


Hadn’t most of the world moved to “first to file” already?


Copyrights and patents should be reduced to a maximum of 7 years and then opened up for anyone to use it. Society will advance far more quickly and good business owners will be able to bring competition and lower prices to everyone.


I know many artists and musicians. "Good business owners" is not that demographic, generally.

The general theme here is a short half-life. You're advocating for shorter rollover - public domain after 7 years. I believe society needs a longer rollover.

I'd advocate for the lifetime of the original author, for the reason that I believe that the author of an enduring work should be protected by copyright for the duration of their influence _not exceeding_ their lifetime. My reasoning is the reality that most creative activity happens during the earlier stages of one's life, that creative people are often guided to less lucrative short-term outcomes for the promise of greater long term outcomes. It's the case of being compensated mostly in options, great for the long term if you have long term success. But also recognizing that long term success should not amount to infinite future benefit.


I think 7 years is a little short on copyright. I think 10-15 years would make more sense.


I'm interested to hear more of your thoughts. What are the downsides of a shorter copyright (e.g. 7 years) vs a slightly longer copyright? I was trying to think of some scenarios, but I can't think of anything about that time range that might fit. For example if copyright was only 3 years, you could imagine big publishers sitting on things until they were out of copyright so that they didn't have to pay royalties. Most music, movies and TV shows are relevant for only a few years. By 7 years out, they are relegated to the rerun slot.

The only thing I can think of are things like novels. So the first Harry Potter novel was quite successful, but it took a while until a movie came out. Movie producers may be happy to wait until it came out of copyright. But that argument might be made for 15 years as well. It would be a different way of doing things, but not necessarily worse, I think. Is it reasonable that Disney can buy up the rights to all Marvel comic books for the last X years? Or would we prefer that anyone could make Marvel comic book movies after a short time -- even if that meant that the original authors aren't compensated.

Sometimes I think that even 5 years on software copyright is not a bad length. If you've been updating the software for 5 years, who wants to fork your project from 5 years ago? If you haven't, that's still pretty far out on the revenue tail even for something like enterprise software, let alone games, etc.


I might go for 30 years: roughly a generation (of humans).

If you wanted a different rule for software then you'd have to carefully define what is meant by "software" and think carefully about side-effects: would you be discouraging people from releasing source code, for example? Perhaps not worth it.


The fundamental problem which seems to be recurring is that the system is generally hostile towards consumer interests.

It's all so broken, it makes me want something dramatic to happen. Maybe the robots will take over soon and be better at governing. Or perhaps yogurt. Or maybe we can start over on Mars.


In reality what needs to happen is strong and consistent lobbying from interest groups like EFF. With a lot of calls to action and advocating. This is how consumer rights happened - with boring hard work and going above corporate fanboyism to protect yourself as a customer. Waiting for some external magic to fix things will not work unfortunately.


Yes, more lobbying will help, and that's why I donate to the EFF, but there is obviously a very persistent and difficult problem which never goes away.

I'm not really interested in "external magic". The core structure is wrong. We should create a new paradigm that works for the citizens by default.


This future is a long way off, but I expect we'll see the new paradigm you're taking about take shape as a collection of smart contracts on a cryptocurrency platform.

Then, it will have to compete with the existing ways. If it's worthy of being the new normal, it should be able to compete and win.

I think that means that economic activity on the platform with the sane IP policy would need to be more desirable for both buyers and sellers. It's obvious why sellers would want to avoid using platforms that are either so riddled with bogus IP that they can't operate (like USD), or so bereft of enforcement that they get no credit for originality (like CNY). But an incentive structure that keeps buyers from jumping ship to the latter sort... That's going to take some creativity.

My best shot at it would be this: Supposing the underlying system is a proof of stake system, perhaps your staking rewards could be reduced or increased based on the extent to which your purchasing habits transferred value to platforms that don't support the desired IP features.

So even though you may pay more for the IP compliant product up front, in the long run the system rewards you for not funding its IP-deviant competitor.

The savvy thing for a buyer to do in such a situation would be to have separate identities on each platform, so you spend your no-rules tokens on no-rules products and you spend your regulated tokens on regulated products, thereby avoiding penalties for being a traitor. Hopefully, the market cap of either economy's token would then hinge on which one functioned better independently of the other: May the economy with the best set of policies win (where winning means attracting the participants from its competition).

The trouble is, we don't have infinite resources to just spawn economic petri dishes like this. Eventually the competing economies will come into conflict over more that just which is more fun to participate in. Eventually they'll start wasting resources on fighting over access to some mineral or whatever. I'm still noodling on fixes for that, but even so, I think that such a system would be no worse than the one we have.


> The fundamental problem which seems to be recurring is that the system is generally hostile towards consumer interests.

I think you're really close, but have specified the inverse of the fundamental problem. The system is generally dominated by seller interests. There is no consumer advocacy on the level of corporate advocacy.


As it no longer promotes progress of science, I have no respect for patents anymore.


This was the podcast that took me from opposing software patents, to opposing the very concept of patents altogether: http://5by5.tv/hypercritical/67

See also: "The idea that I can be presented with a problem, set out to logically solve it with the tools at hand, and wind up with a program that could not be legally used because someone else followed the same logical steps some years ago and filed for a patent on it is horrifying."

- John Carmack


JC would just have to find a different/better way to solve the problem. Just like someone designing disc brakes or mouse traps.


I've seen lawyers look up from reading a particularly convoluted and confusing passage and say, without a trace of irony, "This part of the claims is teaching that . . ."

Teaching, yeah right.

"Teaching" is what the patents are supposed to do. In reality, there is nearly always nothing to actually learn other than how much of, say, the act of 'breathing' the claimant is trying to do a land-grab on.


I have read many American chemistry patents. Up through the 1970s or thereabouts, they often did seem to teach things in enough detail for implementation by "a person having ordinary skill in the art." Now they usually make vague, broad claims and provide fewer, less clear examples of the invention's embodiment. I don't know what drove this drift toward uselessness.

To explain this trend I might vaguely gesture toward businesses responding to stresses as the postwar golden economic era ended, except 1) old patents from before the postwar boom (e.g. 1930s) were also useful and to-the-point and 2) academic publications also grew more bullshitty in the same approximate time frame.


Things like these are truly the things that keep me up at night. The more centralised governments' and corporations' power get concentrated, the more suffering we have.


Not sure why it is hostile to consumers. Consumers seem to be doing just fine.

One can make a principled argument to abolish all patents. (If you don't believe in intellectual property rights for inventions).

However, arguments to abolish so-called software patents don't make much sense. Software is just an implementation detail of an invention. It is not the invention.




Consider applying for YC's W25 batch! Applications are open till Nov 12.

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

Search: