Supreme patent silliness reported by Jamie Condliffe at Gizmodo: a (randomized) prior art generator
At Gizmodo, Jamie Condliffe reports in the post This Website Is Using Math to Create Every Possible Patent on work by Alexander Reben.
As a first point, the title is misleading. Reben's work slices and dices prior patents and patent applications (allegedly) to create prior art, not possible patents:
The system works by pulling text from the entire database of US issued and published (un-approved) patents and creating prior art from the patent language.
The comments are about as inane as the work being reported. For example:
Yeah, he’s very right. Nothing in his database will prevent a patent from being issued. It wouldn’t have before, but now that we’ve moved to first-to-file, it definitely won’t.
It might be usable to invalidate a patent if you can get the judge not to laugh you out of court when you explain you’re using random data as your evidence, but you won’t be able to afford to get to that point in the process anyway.
Of the above comment, IPBiz notes that true "prior art" can be used by an examiner to reject claims from an application prior to issue. Of course, whether or not sliced/diced fragments from earlier patents/applications, randomly combined, would be recognized as prior art can be questioned.
Reben's effort is a fool's errand.
Condliffe links to an article by Brian Barrett. One wonders if Barrett remembers why the "cold war" was called cold. After discussing actually patent litigations, there is text:
The lawsuits that Apple and HTC and Samsung and Nokia and everyone else with skin in the mobile game are throwing at each other, though? Those are dangerous. They're dangerous because Apple doesn't want cash; they've got more than the entire US government. Neither does anyone else, really. What they're all gunning for is market advantage. Which, in the most extreme case, means pulling the Galaxy Tab 10.1 off EU and Aussie shelves altogether. They're not playing Monopoly anymore. They're playing Highlander.
I spoke with Venture Capitalist and author of the influential tech blog Above the Crowd Bill Gurley, who pointed out that "The [patent] Cold War aspect basically defines certain markets as 'out of bounds.' You can't do mobile OS now. CPU is likely tough. Video CODECs are a mess." Three crucial areas, cut off from from disruptive progress because of our system's failings.
The first paragraph might seem to be about the difference between damages and injunctions. Otherwise, one might think
"market advantage" means "cash." Of the second paragraph, the presence of patent litigations (actual war) is not what the "cold war" was about (the threat of MAD foreclosed actual war).
The comments are inane.
No idea of patent maintenance fees:
something similar for tech would be a nice middleground? just make a patent last 3 yrs (or so) and renewable to a max of 12yrs. every renewal with 3yrs would cost a lot of money. This would be an incentive to innovate and invent rather than keep legacy hardware/software around because it wouldn't be easy money anymore.
Gary Kildall forgotten:
As has already been mentioned, code is covered under copyright. The issue I have with that is where do you draw the line? If I want to print "Hello, world." there are only so many ways to do that, and theoretically already covered by somebody else's copyright.
You'd have to define at what point a chunk of code becomes copyrightable (and have to follow some sort of "obviousness" test, as exists in patent law).
And a patent lawyer, apparently unfamiliar with small inventors, said:
One suggestion I think might alleviate a lot of the current problems: Implement a "use it or lose it" provision. Under the current rules, you can patent something and do nothing with it, thus locking up an invention for almost 20 years. It's this provision that provides full employment for patent trolls, and also results in these huge bidding wars for vast libraries of (largely unused) patents. If you were required to develop and market your invention within some set period of time (say, 3-5 years) or lose it, that would remove the incentive for a lot of this abuse and litigation.
The lawyer also did not understand differences between trademarks (which protect consumers as to articles IN COMMERCE)
and patents (which protect inventions hopefully going INTO commerce):
As Talia notes, there is precedent in trademark law (which, admittedly, I am not an expert on). You have to file periodic affidavits with the USPTO, showing that you have used your trademark in commerce, or your registration lapses. I don't think it would be too hard to implement a similar system for patents.
Of loser pays:
"As for litigation, Gurley recommends moving to a system—similar to what the UK and France already have in place—where the losing party pays."
Let's see; first of all, I will assume that the poster was advocating for a statutory change (just the IP system) as opposed to changing the entire legal landscape, because that would be seriously stupid. But let's examine it in the best possible light-
If we change it to "loser pays" for just IP, without reforming the laws, then-
1. We're stuck with the current system, which is horribly unfavorable to defendants.
2. We will have a system that favors rich litigants at the expense of poor litigants. Imagine if you're random content creator, and you're told that your work violates some IP by MegaCorp. If you lose, not only do you have to pay up, *but you have to pay MegaCorp's massive legal fees.* And then you're told that the law is against you. Fun!
3. Imagine you're a company, say, that sounds like Brainal Adventures. Do you bring litigation as your company? Or do you assign the rights to the litigation to some shell company with few assets? On the off chance Shell Company, LLC, loses, then Brainal Adventures owes nothing (unless there's some piercing the veil, yeah, right). If you win, then the company you sue pays Shell Company, LLC legal fees to initiate more litigation, and you have precedent on your side.
Of -- the current system, which is horribly unfavorable to defendants--, patentee-plaintiffs lose more than 50% of the time.
As a first point, the title is misleading. Reben's work slices and dices prior patents and patent applications (allegedly) to create prior art, not possible patents:
The system works by pulling text from the entire database of US issued and published (un-approved) patents and creating prior art from the patent language.
The comments are about as inane as the work being reported. For example:
Yeah, he’s very right. Nothing in his database will prevent a patent from being issued. It wouldn’t have before, but now that we’ve moved to first-to-file, it definitely won’t.
It might be usable to invalidate a patent if you can get the judge not to laugh you out of court when you explain you’re using random data as your evidence, but you won’t be able to afford to get to that point in the process anyway.
Of the above comment, IPBiz notes that true "prior art" can be used by an examiner to reject claims from an application prior to issue. Of course, whether or not sliced/diced fragments from earlier patents/applications, randomly combined, would be recognized as prior art can be questioned.
Reben's effort is a fool's errand.
Condliffe links to an article by Brian Barrett. One wonders if Barrett remembers why the "cold war" was called cold. After discussing actually patent litigations, there is text:
The lawsuits that Apple and HTC and Samsung and Nokia and everyone else with skin in the mobile game are throwing at each other, though? Those are dangerous. They're dangerous because Apple doesn't want cash; they've got more than the entire US government. Neither does anyone else, really. What they're all gunning for is market advantage. Which, in the most extreme case, means pulling the Galaxy Tab 10.1 off EU and Aussie shelves altogether. They're not playing Monopoly anymore. They're playing Highlander.
I spoke with Venture Capitalist and author of the influential tech blog Above the Crowd Bill Gurley, who pointed out that "The [patent] Cold War aspect basically defines certain markets as 'out of bounds.' You can't do mobile OS now. CPU is likely tough. Video CODECs are a mess." Three crucial areas, cut off from from disruptive progress because of our system's failings.
The first paragraph might seem to be about the difference between damages and injunctions. Otherwise, one might think
"market advantage" means "cash." Of the second paragraph, the presence of patent litigations (actual war) is not what the "cold war" was about (the threat of MAD foreclosed actual war).
The comments are inane.
No idea of patent maintenance fees:
something similar for tech would be a nice middleground? just make a patent last 3 yrs (or so) and renewable to a max of 12yrs. every renewal with 3yrs would cost a lot of money. This would be an incentive to innovate and invent rather than keep legacy hardware/software around because it wouldn't be easy money anymore.
Gary Kildall forgotten:
As has already been mentioned, code is covered under copyright. The issue I have with that is where do you draw the line? If I want to print "Hello, world." there are only so many ways to do that, and theoretically already covered by somebody else's copyright.
You'd have to define at what point a chunk of code becomes copyrightable (and have to follow some sort of "obviousness" test, as exists in patent law).
And a patent lawyer, apparently unfamiliar with small inventors, said:
One suggestion I think might alleviate a lot of the current problems: Implement a "use it or lose it" provision. Under the current rules, you can patent something and do nothing with it, thus locking up an invention for almost 20 years. It's this provision that provides full employment for patent trolls, and also results in these huge bidding wars for vast libraries of (largely unused) patents. If you were required to develop and market your invention within some set period of time (say, 3-5 years) or lose it, that would remove the incentive for a lot of this abuse and litigation.
The lawyer also did not understand differences between trademarks (which protect consumers as to articles IN COMMERCE)
and patents (which protect inventions hopefully going INTO commerce):
As Talia notes, there is precedent in trademark law (which, admittedly, I am not an expert on). You have to file periodic affidavits with the USPTO, showing that you have used your trademark in commerce, or your registration lapses. I don't think it would be too hard to implement a similar system for patents.
Of loser pays:
"As for litigation, Gurley recommends moving to a system—similar to what the UK and France already have in place—where the losing party pays."
Let's see; first of all, I will assume that the poster was advocating for a statutory change (just the IP system) as opposed to changing the entire legal landscape, because that would be seriously stupid. But let's examine it in the best possible light-
If we change it to "loser pays" for just IP, without reforming the laws, then-
1. We're stuck with the current system, which is horribly unfavorable to defendants.
2. We will have a system that favors rich litigants at the expense of poor litigants. Imagine if you're random content creator, and you're told that your work violates some IP by MegaCorp. If you lose, not only do you have to pay up, *but you have to pay MegaCorp's massive legal fees.* And then you're told that the law is against you. Fun!
3. Imagine you're a company, say, that sounds like Brainal Adventures. Do you bring litigation as your company? Or do you assign the rights to the litigation to some shell company with few assets? On the off chance Shell Company, LLC, loses, then Brainal Adventures owes nothing (unless there's some piercing the veil, yeah, right). If you win, then the company you sue pays Shell Company, LLC legal fees to initiate more litigation, and you have precedent on your side.
Of -- the current system, which is horribly unfavorable to defendants--, patentee-plaintiffs lose more than 50% of the time.
posted by Lawrence B. Ebert at 9:38 AM
0 Comments:
Post a Comment
<< Home