The intellectual property laws generally reject this "sweat of the brow" doctrine.[1] The idea is, that someone should be able to get a patent for something that results from a flash of insight rather than years of expensive R&D.
I think rejection of this doctrine is a grave mistake as applied to patents. Essentially, it fundamentally misunderstands the nature of innovation in the 21st century. A new invention does not happen because someone has a clever idea one day while mulling over a problem during lunch. That may have been true in the early 20th century and earlier, but it's not true today. Today, invention is driven by research labs staffed by expensive PhDs using even more expensive capital equipment. The purpose of patents in the 21st century is to protect and be able to transact in the results of this capital-intensive R&D.
E.g. Adonit has a new iPad stylus that's 1.9 mm versus the 6 mm typical for iPad styluses (https://adonit.net/jot/script). The capacitive touch matrix on the iPad normally cannot pick up on such a narrow point, so they use some active sensing technology that relies on measuring the electronic field emitted by the capacitive matrix and feeding that back to special software over Bluetooth. For this they have a patent. This was not a flash of insight. This was the result of measuring EM field strength with expensive equipment, buying lots of iPads to test product variability, doing lots of experiments and iterating the design. The purpose of a patent here is to keep companies from free-riding on all that R&D, not to reward anyone for a flash of insight.
By rejecting sweat of the brow doctrine in the context of patents, patent law has become divorced from its underlying economic justification: the prevention of free-riding. The magnitude of the disincentive created by free-riding depends very much on the capital investment that went into the invention.
[1] Sweat of the brow doctrine is more typically used in the context of copyright, but is applicable to a degree for patents as well. The doctrine has opposite effects in the two contexts. In copyright, it has been used to reject copyright-ability for things like phone books that might take work to compile, but aren't "original works." However in patents, it has been used to justify granting patents for "inventions" that weren't a lot of work to invent.
> This was not a flash of insight. This was the result of measuring EM field strength with expensive equipment, buying lots of iPads to test product variability, doing lots of experiments and iterating the design.
So what do you think is going to happen if someone else takes the same flash of insight ("use active sensing technology to achieve a smaller stylus"), puts in their own brow sweat to work out the details, and deploys a competing product? Adonit isn't going to sue for infringement, right? And if they do, they won't win. Right?
I'm talking about a hypothetical change to the patent system, so your question about what would happen under the current system is pointless.
As for what should happen, then under my proposal, proof of independent invention should be a defense to infringement. This falls out of the first principles (the economic concept of free-riding). The purpose of patent protection, economically, is to protect some $(big number) capital R&D investment from a competitor that trivially copies it for $(small number). If, instead of trivially copying it for $(small number), the competitor instead goes to the trouble of engaging in its own R&D for $(big number), then there is no free-riding and no reason to find infringement.
Your proposal sounds more like copyright-for-ideas than patents. In such a system, why would you need to apply for a patent? Samsung copied your phone design? Sue them for that, holistically, rather than finding some arbitrary set of "inventions" that they infringed on. It sounds like the jurors mostly looked at things this way, anyway, when deciding who's morally right. Zynga probably wouldn't like the change, though.
Useful articles do not fall under copyright, except for the separable portion that can be considered purely aesthetic.
E.g. if P&W disassembles a Rolls-Royce engine, and copies the shape of the turbine blades (which are the result of very expensive R&D), Rolls-Royce can't assert copyright because the shape of turbine blades is functional, not aesthetic.
Now you're the one who's ignored that this was hypothetical.
My point was that if we go ahead with your proposal, such that patent protection becomes about preventing copying work rather than (possibly accidentally) duplicating an idea, then what's the point of filing for patents? It could work just like copyright does now.
Rayiner: I can't respond to your next post for a while, but you didn't address the main point I was attempting to make in comparison to copyright: You don't need to file. There would be rules that you can't significantly copy someone else's hard work. If someone copies your hard work, it will be evident from the facts, just as it is evident with copyright violations, and you can sue them. No patent required.
It wouldn't work just like copyright does, even if you extended copyright to functional designs.
Consider how patents and copyrights interact with respect to software. Say you make a new audio encoder leveraging some psycho-acoustic property. Copyright protects the literal source code only. It doesn't prevent someone from doing a "clean room" reimplementation of your software, and in the process taking advantage of all of the expensive testing you did to validate your psycho-acoustic model.
Patents as they exist today might apply to any encoder leveraging that psycho-acoustic property. I.e. a competitor couldn't hold up the fact that it performed independent testing to build its own psycho-acoustic model as a defense to patent infringement.
My proposal falls in-between. You can prevent someone from reverse-engineering your program to copy the essential details, which is the fruit of your expensive R&D. However, you can't prevent someone from using the same basic idea when they go to the expense of deriving those essential details for themselves from the basic idea. What patents as they exist now protect, and what my proposal explicitly wouldn't, is the "flash of insight." The realization that multiple people may have in response to some journal paper that some newly-described psycho-acoustic phenomenon may be used to build better audio encoders.
Implicit in my characterization "copyright for ideas" is that this proposal will somehow extend copyright, as it's currently understood, to cover ideas. I understand your point that copyright as it exists today does not cover more abstract ideas.
The main difference I was attempting to emphasize between copyrights and patents is that copyrights are automatic whereas patents must be filed and granted. And my point is that your proposal, which seems quite reasonable to me, would seem to work just fine with the copyright model of automatic rights: no patent filing required. And for that matter, the distinction between elaborate designs, which cannot be copied, and "flashes of insight", which can, looks a lot to me like a principle of fair use.
I agree that patents should be unique and novel (and that many patents are issued that are not novel). I disagree on the idea that innovation can only happen through expensive R&D and by experts in labs. Such innovation is useful and should be done and protected, but seems to be more evolutionary than revolutionary. Also, will companies take the time to do the R&D if they can't patent the idea and concept first?
Its important to protect more revolutionary ideas that may be developed by someone in a garage or starting a small company. Patents are supposed to also protect small businesses and individuals with great ideas that don't have the type of budgets you are talking about.
I think in practice it's exactly the opposite. Just look at Silicon Valley. Who is doing the really revolutionary stuff? Self driving cars are Google + the DOD. Space X has burned through about $1 billion in capital, with half of that coming from NASA. Tesla is burning $50-70m per quarter on R&D.
Small companies certainly invent new things, but even then there is usually a lot of money involved, from government contracts, university spin-offs, etc. The wireless tech company where I used to work started out in a basement, and has probably spent $10m+ over the last decade developing its technology (bootstrapped with DARPA contracts).
This is not to say that small-time inventors don't have revolutionary ideas. But in practice, the stuff that happens in a garage isn't inventive but rather application of well-understood technology.
> By rejecting sweat of the brow doctrine in the context of patents, patent law has become divorced from its underlying economic justification: the prevention of free-riding. The magnitude of the disincentive created by free-riding depends very much on the capital investment that went into the invention.
Odd, I had always understood the underlying justification to be the establishment of free-riding. In return for a temporary monopoly, everyone would be able to use the same superior technique.
I agree that allowing independent invention to be a defense to patent infringement would solve almost all of the problems we see with software patents. But it seems to have some different trouble.
If you allow competitors to use the invention if they pay the cost of duplicating the research, you cap the value of the patent at the cost of doing the research rather than the value of the result in the market, because anything more than that and you cause competitors to [band together to] duplicate the research rather than paying you for a license. In general this is a good thing because monopoly rents are bad, and it allows the original party to recover their costs by licensing without allowing inefficient monopoly rents in excess of the original cost. The trouble seems to be that it doesn't provide an incentive to actually be the one who does the original research, because if the research pans out then you can only recover up to the cost of duplicating the research, but meanwhile you have to take the risk of the research not finding a solution.
The other trouble is, how do you prove or disprove independent invention after something is published?
I'm adamantly against software patents, however, when it comes to chemistry patents, I know it's not a rubber stamp process. It can take a couple years to get a patent, and lots of patents are rejected. A lot of the process involves the examiner stripping out claims that are invalid (the parties seeking the patents generally want them to be broader than they really should be).
As holder of a US patent on a purely software "invention" (aka, some game I wrote) I can categorically state that the only way it could have been granted was that it was rubber stamped because it is just a lot of talk about the algorithms I wrote, executed on a portable game system.
CEO loved the idea of getting a patent because he thought it would make us look impressive and it would defend us in advance if someone else made a similar thing and tried to attack us with their own BS patent.
Makes me sad when i consider the comparison with my grandfather's multiple rounds of back and forth with the PTO fighting to finally get a patent on a mechanical device. Comparitively yes, they rubber stamp software patents because they clearly dont understand the art. They kind of understand more traditional (aka real) inventions.
I think a bigger problem is that a well-funded entity can keep resubmitting rejected patent applications for re-examination. I think there needs to be a limit on how many "appeals" you can get.
From what I have heard, and this might be wrong since I am no expert on the matter, re-examinations creates a perverse incentive for patent-examiners to accept patents since their job-performance is rated on how many patents they process.
1) Something needs to be a true innovation to be a patent 2) The process should take a lot of man hours.
I don't feel either of those things happen due to shortage of staffing and a wild everything is possibly patentable.