The lack of an independent invention defense (or some variation of this) is pretty much 80% of what's wrong with patent law in my opinion. A sampling of the issues an independent invention addresses:
* Obviousness. Patents are required to be non-obvious to a "person holding ordinary skill in the art", but obviousness is highly subjective and difficult to assess (especially for a non-technical jury). In contrast, evaluating whether something was independently invented, while not necessarily simple, is a more objective test.
* Utility. One of the rationales behind the patent system is that we want to encourage inventors to share their ideas with the world instead of just keeping them secret for as long as possible. While the patent system requires publication of the invention, it doesn't require publication in a manner that's particularly useful. If you could only assert patents against people actually using them, then you would have a strong incentive to make them useful (e.g. source code).
* Supply and Demand. The current system grants a flat 20 years to each patented invention, regardless of how difficult or expensive it was to invent. Recognizing independent invention creates the opportunity for a more flexible system. Inventions that are hard to reproduce independently are more valuable to society and should enjoy a longer term than easier ones.
The obvious problem with all of the above of course is how feasible it is to assess whether something was invented "independently". I've addressed that to some degree in this blog post (http://www.techdirt.com/articles/20121011/14171220681/yes-in...), but I admit it's a hard problem.
I'd argue you'd get more economic benefit abolishing patents entirely (not just the software ones, I mean the ones on reproducing someone elses car or gas furnace). You pay a cost in a loss of investment valuation towards invention because it is less economically beneficial if you can't have a monopoly on your patented good, but you benefit economically by letting new inventions reach natural price equilibrium immediately, and the spread of new inventions would propagate more inventions, even without a major profit motive - you still benefit being "first out the door" on a new idea. Which we see all the time in software, and then run into this problem where first out the door has already made a bunch but is now suing me for trying to do something even remotely in the same problem domain because I see demand there, but I won't be able to meet it because I'm artificially locked out of that market.
That applies to all economic domains. Patents worked back when it was extremely hard to sell your goods to its entire market, because you were either distance limited, communication limited, or production rate limited. In the first two situations someone could just take your idea, set up half way across the country, and made all the returns.
Today, I can invent anything I want and have it on ebay overnight and ship it anywhere in the world. I can advertise on most major ad services if I want, with an audience of the entire first world. The only limiting factor for me is demand, but that just means I have to be safe when bringing inventions to market, but it is good for the economy overall because it means if I can't meet demand, someone else will pick up the slack.
It is still, like I said, advantageous to the first to market with an idea, enough so that I wouldn't predict much of a drop in novel ideas. Like I said, I would expect more if technology and designs were more open and people could cheaply iterate and extend past concepts more readily.
Being "first to market" only matters when you're building things like Facebook or Windows, where network effects are relevant. If you're making consumable or durable goods, being first to market is irrelevant. India will clone your drug within a few months of its release, but there is no way you'll make up all that capital investment by then.
Without patents, the rational thing to do is not invest any more in an invention than you can recoup in the short window before someone clones it and undercuts you on price. Moreover, it diminishes inventors and elevates manufacturers. The smart strategy in a world without patents is to build up massive overseas manufacturing capacity, because you can always quickly copy designs and then undercut the original designer with your manufacturing muscle.
'Andrewfong has a good point about independent invention being a defense. The idea of inventing something on your own only to have someone else claim they patented it is most of what bugs people about patents. In engineering circles where that is much less common, you don't have an anti-patent sentiment. When I was studying aerospace engineering in college, nobody I encountered perceived patents as anything but a good thing.
Mentioning the term "first to market" was unfortunate, but the key phrase in that sentence was "network effects", and Windows, Office, and Facebook all great examples of the advantage of network effects.
Yes, its only recently with software that patents have really come to hurt. Maybe that's because most engineering disciplines do not patent frivolously, or maybe that not many patents in these areas can generate so many billions in revenue.
(PS: Please don't use the term "India" to refer to certain Indian Biotech companies that reverse engineer patented pharmaceuticals developed in the US)
Not necessarily opposed, but there are a couple areas where this might be a net loss absent other adjustments. Pharmaceuticals would be the obvious one -- they require a large upfront research cost, copying is trivial, and the FDA approval process erodes first-to-market advantages.
Pharmaceuticals would be devastated. While there are externalities that need to be solved around our production of drugs, the patent system is a good thing overall for them.
No one spends $1 billion to make a drug if anyone can copy it simply.
Additionally there are international problems related to abolishing patents.
Ease of copying. It takes some effort to reverse engineer a car, and things like manufacturing processes are difficult to ascertain by just looking at the final product. In contrast, many pharmaceuticals are relatively simple to copy and manufacture once you know the chemical composition of a product.
If you prefer you can think of software. However software is protected by Copyright quite sufficiently so doesn't come up when talking about patents (well in this way).
But even in situations where pharmaceutical companies don't have patent protection, just providing a better product at lower prices to a larger, poorer market has proven to recoup costs just as well as charging $10,000 per dose on a patent drug.
(I saw quotes yesterday from the founders or presidents(?) of Merck and Bayer, which espoused both philosophies, and both claimed successes.)
I'm not entirely convinced that patents are necessary to ford the research and regulatory approval. Even in a market where generics are relatively trivial, there are ample opportunities for profit.
Let's stay: I am a pharmaceutical company, and you're a competing pharmaceutical company.
I invented miracle-drug and can now produce miracle-drug at 0.04$ per pill, can you can produce miracle-drug at 0.06$ per pill. You would think I should win because I can product the pill cheaper, but the full story is your very first miracle-drug pill costs you 0.06$ and my very first miracle-drug pill cost 1.5 billion dollars. I lose because I invented.
Yet you don't lose, because you can recoup that investment.
The alternative is to not make any money at all. As long as that drug produces you a net profit in the long run, which I imagine would fund additional research after recouping anyone investing in the project, then the fact someone else can also make money is irrelevant.
Think of it from the alt-companies perspective: X is sold at $0.06 a pill, but I could sell it for $0.04! There is demand, I can create supply, and everyone benefits except X.
But I don't buy that selfish interests wouldn't still invest heavily in medicine even if it is a risky investment - nobody considers the emotional fact that rich people do get sick, and are likely (in the absence of big pharma research) to directly invest to see the diseases they suffer or are concerned about cured or abated. Think of how kickstarter works - you throw money at a project that may or may not succeed, but it happens all the time because people want the outcomes regardless of the financial risk associated. Same could easily happen with pharma.
> just providing a better product at lower prices to a larger, poorer market has proven to recoup costs just as well as charging $10,000 per dose on a patent drug.
Unfortunately, it's not alway possible, there are sometimes very few people in the first place:
Cigarette companies don't get those kind of super-sized judgements against them, gun manufacturers don't. Why the hell should Google?
My favourite part is just how good the Microsoft lawyers are - "we will pay 5% of whatever Google pays" is an incredibly good deal and shafts Google quite nicely at the same time.
And anyway, isn't this really obvious - to anyone? I have scanned the patents referenced and frankly I don't see any algorithm that is not "return information relevant to the search term presented, and mark the users profile if they liked it"
No implementation details, which frankly seems to be the whole difference in software vs hardware.
"the original participating manufacturers (OPM) agreed to pay a minimum of $206 billion over the first twenty-five years of the agreement" - 4 companies, 25 years, works out at $2.1bn a year each. 10x more than the article thinks Google will pay out...
evaluating whether something was independently invented, while not necessarily simple, is a more a objective test
Only if you take a very generous interpretation of "independent". For an reinvention to be truly independent, the reinventor would not only need to not have read about the original invention, but also to have not read about anything inspired by the original invention, and not talked to anyone who read about anything inspired by the original invention, etc.
Right, independence isn't binary. There are definitely degrees of dependence. But it's more objective than "obviousness" at least. Obviousness is also non-binary, and because the jury must assess obviousness from a "person holding ordinary skill in the arts", there are additional questions as to what constitutes "ordinary skill" and what constitutes the requisite "art".
This is the core problem. The probability of two independent "inventors" (i.e., software engineers) discovering a similar solution is very high.
Before software engineering, the probability of this happening may very well have been low enough to where nobody cared.
But if you're writing code, you are relying on and creating thousands of "inventions" a day. To "claim" one of those inventions so that nobody else can use it is absurd.
The law just doesn't get that yet, which is weird, because big tech companies have been getting burned by it for decades now. I guess they just figure the benefits>costs for them, and when peons get screwed, who cares?
A much simpler solution would be adjusting the expiration of IT-patents. The patents in this case are from 1996 and expiring in 2016. If the term for IT patents was reduced to 10 years, we would have many fewer problems.
Fewer problems, but still a pain. If you independently come up with something one year in IT after the original patent holder, that's still nine years of suck.
I mostly agree. Though, you'd get some interesting reverse engineering related problems.
Ie look at the patent, come up with a problem that would make this patent obvious, lock some engineers in a sealed room with the problem, and wait for something like the patent to fall out. Was that independent invention, then?
Probably not, although you could write the rules in a way that avoids this I think -- e.g. the knowledge of the person who looks at the patent is imparted to the engineers because they're all part of the same organization.
You could probably work around this by e.g. introducing enough layers of separation between engineers and the person who read the patent/formulated the problem. It would be an interesting form of... knowledge laundering? ;).
I don't see why independent invention wouldn't lead to a legal assumption of obviousness. Independent invention seems like incredibly clear cut proof of obviousness to me, anyway.
The issue is: obvious to whom? Both Newton and Leibniz independently invented* Calculus at about the same time. That may mean Calculus was obvious to Newton and Leibniz (and it may very well have been obvious to anyone in that particular community of mathematician), but it certainly wasn't obvious to the average mathematician at the time.
* More like discovered (don't want to imply math is patentable), but the point remains the same.
If several people would claim independent invention, it is obviously a statement about obviousness. So the patent would already be invalid under the current law.
Now, since everyone involved in patents, they don't want to interpret the law that way. So there is a non-written agreement that no one will go that route. and that is idiotic.
* Obviousness. Patents are required to be non-obvious to a "person holding ordinary skill in the art", but obviousness is highly subjective and difficult to assess (especially for a non-technical jury). In contrast, evaluating whether something was independently invented, while not necessarily simple, is a more objective test.
* Utility. One of the rationales behind the patent system is that we want to encourage inventors to share their ideas with the world instead of just keeping them secret for as long as possible. While the patent system requires publication of the invention, it doesn't require publication in a manner that's particularly useful. If you could only assert patents against people actually using them, then you would have a strong incentive to make them useful (e.g. source code).
* Supply and Demand. The current system grants a flat 20 years to each patented invention, regardless of how difficult or expensive it was to invent. Recognizing independent invention creates the opportunity for a more flexible system. Inventions that are hard to reproduce independently are more valuable to society and should enjoy a longer term than easier ones.
The obvious problem with all of the above of course is how feasible it is to assess whether something was invented "independently". I've addressed that to some degree in this blog post (http://www.techdirt.com/articles/20121011/14171220681/yes-in...), but I admit it's a hard problem.
EDIT: Typos