Patent Reviews Via Wiki

unboring writes "Fortune reports on a pilot program where the patent approval process would be opened to outsiders for review. Reviewers can vote and discuss on different proposals, through say a wiki. Given the many (recent and past) patent approval fiascos, this seems like a good idea. It'll be interesting to see how they would deal with the issues faced by Wikipedia."

Why a wiki?

Wouldn't a good old-fashioned forum serve this purpose more effectively? I mean, it's not like the reviewers are going to be editing the actual patent submission, just discussing it.

Re:Why a wiki?

Probably because a wiki is more meant to be a point of reference than forums are. There's nothing to stop them from adding a forum to the wiki, though.

Re:Why a wiki?

You and your parent poster are both right because a wiki is nothing more than a collapsed forum. When I read /. at threshold==5 it's the same thing as reading a wiki without looking at the deltas.

The bottom line is that the process will be more transparent and more open to correction than the current system.

If it would be a forum....

by page 20, they would be arguing why pink ponies are pink and if [insert favorite politician] owns one.

Re:Why a wiki?

Wikis are better at presenting a single summary of discussions. In a forum, minor mistakes don't get fixed easily (eg. if one person corrects another person, you have to read through the whole conversation to get an accurate glimpse of the conslusions drawn).

That doesn't mean wikis present only one view... a page can note that disagreement exists, and the page can present different views of a situation by different groups of people. But when there's longer amounts of discussion, having a single-page summary is far far better for newcomers, and for decision-makers to scan the arguments more quickly.

Re:Why a wiki?

Wikis are better at presenting a single summary of discussions. In a forum, minor mistakes don't get fixed easily

I took the original poster's (excellent) suggestion not to mean "literally use a forum" but rather, more generally, why not keep track of who said what? There's nothing to say you can't design a forum in which there are discussion threads and other mechanisms, such as accounts you can log into and vote. It's not rocket science to give the person a menu that says:

• Post a comment to the discussion forum.
• Register (or re-register) your present opinion about this in a multiple choice form that can be aggregated mechanically with others' votes.
• Attach a free-form summary of your personal opinion on this item in 300 words or less (with optional URL pointer to a continuation page in another venue of your choice).

I don't think everyone editing each others' text is the way to go on this since it creates an artificial sense of tension--there's no reason that my having a different view than you means we have to fight over who's view gets recorded. But allowing each person the choice of several ways to present their throughts (interactively or not, multiple choice or not, size-constrained or not) seems good because you can get summarizable info when people choose to offer it.

Also, allowing anyone to update at any time means you can keep by-day summaries of how people's opinions change and to review the history of what the discussion and opinion summaries looked like on a given day.

Link to patent review project

Curiously, neither the submission nor the CNN article gave a link to the actual project page [nyls.edu] for the Peer to Patent Project. That page has more information and a blog giving updates on progress. There's also a Community Patent Proposal Wiki [jot.com], but it seems to be down.

Interestingly, the lead sponsors for the project are HP, IBM, Microsoft, and Red Hat. Strange bedfellows, eh?

Project Red Herring

For me the whole approach looks amaterurish. It's like some persons who are clueless about patent reality thought of a new solution at the round table. It sounds nice and that is why Slashdot reported it, but in fact it is toyground action.

The patent problem has to be solved and the patent problem is NO problem of prior art, novelty assessment, "triviality" or "obviousness", patent examiner laziness or mistakes etc. However patent institutions and patent professionals like to let you enter the toyground. There you can think up solutions, but they will not solve the problems, and the institutions are safe.

Patent reform of Congress went into the same trap. They discussed the issue for the ..hmm ... third time(?) in 2006. It makes no sense to follow the red herrings. Red herrings serve the purpose that you don't get the fish.

Re:Project Red Herring

It's like some persons who are clueless about patent reality thought of a new solution at the round table.

Even worse, it's a solution to a problem which doesn't actually exist. The vast majority of people think they understand the patent process but simply don't. Publication is a vital part of the process, as is prior art assessment, demonstrating novelty and non-obviousness (to a "person skilled in the art"). You get a patent (ie. exclusive right to commercially prosper) for a set amount of time (20 years in most places) in exchange for disclosing the idea to everybody.

There is nothing to stop you filing a patent application which is an exact copy of an existing, granted patent. That patent application could go to PCT and be published at 18 months and then progress into national levels in individual territories. This could take 3-5 years and can happen without anyone in the patent office having actually examined the validity of that patent . If the patent application was published without an international search report, it might not even have been cross-referenced against the existing patent database!

Even granted patents are technically not valid until they have been challenged and upheld by a court.

If you build it, they will abuse it.

The chances of this happening are inversely proportional to the chances that it will be abused if it does happen. If a major company has a multimillion dollar product on the line, they will do ANYTHING to make sure it gets approved, even if it means sabotaging any method open to the public. It doesn't matter if it's a wiki, a forum or a voting system, they will abuse it because millions (or even billions) could be on the line.

Re:If you build it, they will abuse it.

One of the best features of a wiki is that all the edits are visible. If there's consistent corruption of particular information damaging to a claim, then that's the information to look at!

I suspect that there will be more need for accountability than there is with, say, wikipedia, but just having the facts unearthed by an army of interested persons will be valuable.

The mere fact of having some prior art or other pertinent information on the wiki won't be 'make or break', but an idea of what factors need independent verification should add enormous value to the Patent Office's research.

Re:If you build it, they will abuse it.

The chances of this happening are inversely proportional to the chances that it will be abused if it does happen. If a major company has a multimillion dollar product on the line, they will do ANYTHING to make sure it gets approved, even if it means sabotaging any method open to the public. It doesn't matter if it's a wiki, a forum or a voting system, they will abuse it because millions (or even billions) could be on the line.

Conversely, if a company's competitor has a multimillion dollar product on the line, it could be very beneficial to help dig up prior art to prevent/negate a patent and then cash in by selling (or not withdrawing) a similar product. That might sound bad, but it would allow for actual innovations to get protected (no prior art) and allow the public (as well as other companies) to pay less for derivative items because of increased competition. If the idea catches on, the grassroots/astroturf community could get used to help ensure keep sabotage successes to a minimum.

I think it's a great idea. I just wish I could see how it'd apply to biotech items. Patenting genes and chemicals found in nature still bugs me.

A wiki seems a strange choice

It's laudable that they want people to have a say in whether patents should be approved, but why a wiki instead of a normal forum? The only reason I can think of is that in a forum people usually post multiple similar comments because they don't read what has already been written prior to posting (yes, I'm talking to you, Slashdot!*), and reading all those comments would turn into another burden for patent examiners. However, even with a wiki you have to read the history of the page, or you might just get the last person's opinion. So, again, why a wiki?

BTW, it will be mightily interesting to see if this has none, some, or a major impact on what patents are actually approved. I can imagine geeks with too much time on their hands will find some obscure prior art to almost everything that companies think of patenting, thus leaving only the things that really should be patentable to make it through the process. In short, how it should be. But it all remains to be seen.

* And yes, I'm aware that somebody has probably posted a comment similar to this one already, making it oh so ironic. Yawn.

This will produce some *very* unhappy campers.

Specifically, those who would ordinarily slip through the cracks because someone doesn't catch prior art.

A significant population with an [almost] unmeasurable body of knowledge and information would do a pretty thorough job of flagging someone which the patent examiners working under extremely high pressure to push things down the assembly line. This would make the examiner's job one of validation of claims posted via wiki.

One question remains: What's going to happen if we see a couple of companies who shall remain nameless and are granted patents by filling out a pre-approved form are faced with prior art (or silly art) claims and the company receives the approval anyway? That might prove there's some monkey business is afoot. (Donating a Playboy Bunny to their favorite charity? (Charity begins at home)

Obvious problem

The content of a patent application isn't protected until the patent is approved. Submitting your patent application to a public site lets all your competition know details of what you're doing with absolutely no guarantee that you'll get patent protection of your idea(s).

I must be missing something, because this seems so obvious and insurmountable.

Re:Obvious problem

The content of a patent application isn't protected until the patent is approved. Submitting your patent application to a public site lets all your competition know details of what you're doing with absolutely no guarantee that you'll get patent protection of your idea(s).

I must be missing something, because this seems so obvious and insurmountable.

Yes, you are. First, your first statement is only partially true. An invention only has guaranteed protection once a patent has issued. However, once the application is submitted, the contents are retroactively protected if the application succeeds.

Also, any inventor who wants to seek protection outside the US has to make a public application anyways. The US is basically the only country left with a closed review process. You can opt for a closed review, but then your patent is only enforceable inside the USA (and possibly a couple other places - but not Europe, and some other large markets).

Second, the fact that the site is "public" is ot the relevant fact for patent validity. What matters is if the inventor makes one of a few categories of "disclosure". That includes most avenues of publishing, sales, etc. But if the USPTO does something that resembles disclosure (like posting it in a review wiki), they are still free to give it whatever legal definition they deem appropriate. And there's no way that they would define any part of their review process as a disclosure.

Re:Obvious problem

Everything's obvious in retrospect. People seeing something for the first time claiming "I could have thought of this" have no credibility in my book, whether or not they cry out loud.

Re:Obvious problem

Bullshit.

Already today the USPTO publishes pending patent applications, usually years before they are either approved or denied.

You can see for yourself here: http://appft1.uspto.gov/netahtml/PTO/search-bool.h tml [uspto.gov]

Publication is a very fundamental part of the patenting process, if you do not want your competition to know about your invention then you cannot patent it either. The idea is that if your invention really is novel then you will get protection sooner or later and can go after anyone thats infringing on your patent.

If your application does not get approved, then you are correct that you will have lost any edge on your competition by way of secrecy. But then again, if your application is rejected then its most likely not very valuable anyway.

BTW: This is the reason a lot of millitary research or technology deemed important for national security is never patented. Patent applications implies publishing.

Uhmmm ... not a new idea

I can't find where I posted this, but I suggested this months ago. Its a great idea, use and open source methodology to bring the USPTO up to date with things like, oh, say... prior art, and maybe perhaps just what exactly is really a bad idea? If there were respected members from relevant industries, it would be even more appropriate, and work more effectively. Mind you, Mr Balmer should be permanently banned (sorry, couldn't help myself) The point is that the USPTO cannot continue to be experts on everything, they need help, and this is a very workable way to help them.

nuff said

yeah

I would love this, but you would need experts, men who know about engineering, software and such, people you allow to vote to do this through a portal like a wiki. That way they don't even need to be in the same room. Hell you could have one man create a list of people who would get one time logins to be able to go on an edit the wiki or something like that. I dont necessarily the public should be allowed but hell make it open, forty percent of americans go out to decide the next president, what are the chances more than those who are interested will go especially if you make some sort of sign up. That probably will discourage all the but the hardcore troublemakers.

Bad idea

Wikis are great at things which involve facts where no one gains anything from lies (See Guildwiki for example).

Wikis are bad when millions of dollars are involved and anyone can edit them. It only takes some "unknown person" who "doesn't have anything to do with the company" to goto the wiki and repeatedly edit it so it seems the patent is invalid or worthless and it may just seem that way (yes I know theres checks). Look at viral marketing and Sony's "lets graffiti shit to look cool" idea, notice how companies don't care how much money they spend just so they can look cool? Well what if they could pay some kid a couple of bucks a week to edit a wiki which could influence major things down the line... Yep you guessed it, they'll eat it up.

People need to register "Use the right tool for the right job" rather than "Wow it's open source! I bet I can use this system to fix everything in the world! Cancer/World hunger/Wars I've got your number bitch!"

Some real flaws

I hate patent abuse as much as the next guy, but this seems like it's just begging for abuse.

How's this for a (new) patent abuse business model:
- Watch patent review wiki for interesting stuff
- Steal good ideas that other people have
- Instantly work on creating pre-dated "prior art" on websites, blog postings, etc
- Use shill accounts to point out the "prior art"
- Make some good cash off of other peoples' R&D

Or how's this for a "fuck with a much-hated company" mob mentality:
- Watch for patent applications from the hated company
- Instanlty work on creating pre-dated "prior art" on websits, blog postings, etc
- Post to slashdot, digg, etc, linking to the manufactured "prior art"
- Watch while thousands of tech zealots slam the patent on the wiki, citing your dishonest "prior art"

There are plenty more ways to abuse this, of course, those two just came to mind quickly and are kind of amusing.

The patent system is broken, that's for sure. But this isn't the answer. Or at least, if this is the answer, we might as well do away with patents altogether, since they will be very, very easy to dishonestly undermine. I know I'll get jumped on here, so let me ask that if you favor simply removing patents (or software patents, or whatever) from the law, please just say that and don't defend this terrible idea because it gets the outcome that you want while still pretending to support the idea of patents.

-b

Patent pending

That will work... until someone patents the Wiki.

How many free eyes are they expecting?

THe nice thing about the wiki (probably, in their eyes) is that they're probably expecting it to grow exponentially like the original wiki did.

In this case, however, they seem to be expecting tghousands of 'volunteers' to go hunting through an exponentially expanding list of stupid applications and doing the reviewer's jobs for free. I expect that this is really only going to happen to the most agregious of the 'bad' patents.

Other than that, you might be able to hope that some of the big companies will start assigning people to look at these things on an ongoing basis in the hops of slamming just about anything that moves before it get legs. Of course, if they start making agreements as to what they'll 'miss', then we'll have the worst of both worlds -- with the big companies setting up truces against each others' "volunteer" examiners, while the little guys get lambasted.

Yep. Lots of room for abuse.

Participants get sued - for knowing patent use

An important fact about patents is that the damages for knowingly using a patent are much larger than those for doing it un-knowingly. It's going to be very interesting when people who review the patents in the wiki get sued and have difficulty denying their knowledge. Or, more interestingly, when the patent trolls start demanding IP addresses behind user names to help in their lawsuit campaigns.

I'm really not sure it will be wise to contribute to this if you have anything to do with software production or ever plan to have anything to do with software production.

How will..

How will astroturfing be prevented? Or minimized?

Why not just approve all patents?

Seriously, that's pretty much what's happening anyway. The only way to find out if a patent is valid is through a court-case. I never understood why the patent office works the way it does. Why not just have people register whatever they want. If they find an infringer, they are going to have to prove that they really are infringing, and the infringer is going to have to prove that they are not in front of a judge and jury. Why try to (very ineffectively) do some of this in advance?

You could even add fines for entities registering patents which have unmentioned prior art (they obviously didn't do their research).

It would perhaps also keep investors from only investing in companies that _appear_ to have some interesting patents, but no real technology/knowledge/expertise to back it up. It would be nice to see more investment into companies that actually know how to make something and actually advance technology.

Hmm.

I'm not a patent guy (I'm not nearly technically minded enough), but I asked this when this was first floated: Why exactly should we be doing their job for them? I know, I know, the patent system is broken. But this is the best they can come up with?

Simon Phipps (Sun):"the social contract is broken"

In a recent post on his personal blog [webmink.net], he had this to say:

For it to be worthwhile for society to grant these "temporary monopolies", both sides of the deal have to be maintained. I filed a few patents back when I was at IBM, and none of them seems to me to convey the know-how for a skilled programmer to be able to use the idea readily. They are patterns designed to help a patent attorney identify infringement. While the progress of the software industry's know-how was only being advanced by corporations, there was a (barely plausible) rationale for software patents; their lawyers could decode the patents. But now the individual developer acting in community is an equal source of progress, it is clear to me that the social contract is broken.

In this setting, it seems unlikely that asking the public, and software developers in particular, to verify (or do) the patent examiners' work can be of any use. It's a remedy that does not quite fit the problem, which is granting patents on matters that once were excluded for good reason.

Moreover, the articles also linked there [webmink.net] indicate that in the field of software at least, quite probably there never even really was such a time when the "industry's know-how was only being advanced by corporations" (rather than e.g. academics and individual inventors).

This already exists for over 1 year already

Take a look at my site:
http://prevalent.de/index.pl?site=1&subsite=3&lang =en [prevalent.de]

Ok, it's for german and european patents only, but it should fit to give you an idea on how I imagined this. On the left side there is always a patent application, and on the right side you can vote the patent, submit prior art, use a forum to discuss or enter a wiki side for each patent.

The difference is, that I don't review patents before they are published. (That's not possible of course). But there are new patents that are granted, but are still within the german or european objection time.

Until today I filed four objections at the german patent office against granted patents. The cases are still running and not yet decided.

Take a look at it: http://prevalent.de/index.pl?site=1&subsite=3&lang =en [prevalent.de]

cu,

Jan

Maybe good, maybe not so good, but...

...while some good points about how this scheme can be sabotaged by entities for or against a given application there is one aspect that is critical as to how successful this concept will be, namely, how the PTO will implement this. The key aspect that signals this is hinted at in TFA: examiners, on average, have about 20 hours to spend on an application so that they will be rated "fully successful"; to meet this critical requirment in their Performance Appraisal plan they must average out, over any one year period, at least 95% of their assigned "expectancy" Now the "hot button" technologies have been rated at the highest level, so the software, business methods, biotech areas get roughly twice the time (i. e. about 40 hours).

These time figures have been in place since about the latter 1960s (although the "hot buttons" areas are more recent).
Congress has generally critcized the PTO for the excessive application pendency, so a PTO management culture developed dedicated to controlling pendency, and the productivity requirement is, along with some time limits to respond to activity once the first action has been issued, the primary driver used to effect this control.

Over the years, because additional requirements have been added to the examination process, along with the courts increasing the showings needed to legally establish non-obviousness, it has been harder to examine an application. Because the time allotted is constant the quality of issued patents has declined (yes in infringement cases the Courts apply the same, increased burden, but then the parties can devote hundreds of hours, not, at best, 40 hours or so as in the PTO during examination).

As to this current wiki proposal: if PTO management implements this as it has other examination process "enhancements" in the past, it will be "bolted onto" the existing process; just one more "thing" that has to be "checked off" in the prosecution of the application with no more time give to accomodate it. The result will be, if the examiner knows that it will be "wikied" is that less time will be spent during the pre-wiki initial exaamination, betting that the post wiki part of examination will likely develop the best prior art; after all why waste a lot of time with the prior art that will be developed pre wiki when, later, new, better prior art is developed that will just stretch out the prosecution, leading to a lower production figure (remember, it's a constant).

It's possible that PTO management will actually develop a reasonable program that is actually effective, but given their past, deeply ingrained "production line" mentality, I rather doubt it.

Improve patents

Some time ago i posted the following on a wiki
http://c2.com/cgi/wiki?HowToImprovePatents [c2.com]

# The evaluation of patents should be based on objective methods, not subjective rules.

# A business secret rule. To get a patent on an idea it most be a business secret. an idea is a business secret if it can be used in a business without giving it away. An contra example is amazons one click patent which can't be used in business without reveling it. This rule insure that the community get something in exchange for the patent monopoly.

# A bounty system. The patent systems should be a public bounty system. Every patent application should have a bounty on it. The initial bounty is payed by the patent seeker but it should be possible for others to add to a bounty.

# Internet system: The patents examination should be made public on the Internet where it should be accessible to all without cost.

# Non-obvious tests: First the technical problem should be posted and the should be a period for the public to post solutions to the problem, if the solution is posted. Then the poster get the bounty and the patent is rejected as obvious. But the posted solution remains available to the public.

# Novelty test: Then the patent solutions is posted and the should be a period to find prior art. If prior art is found the patent fails the novelty test and the finder get the bounty.

Summary dismissal of perpetual motion machines?

If so then I'm against it (until I get my patent submitted, that is, after that I'm all for it.)

Ultimately...

I believe that this system would/will suffer similar problems to the current process: It will be dominated by players with deep pockets. If you can just pay a sufficient number of people to log on and work for your patent, odds are you will win it with a forum/wiki/discussion group/whatever. Just as you can now persuade the USPTO by throwing money into lobbying them.

I dunno...

I dunno...

"Patentality" doesn't have the same ring as "Wikiality".

- RG>

Could help a lot of projects

Maybe the underpinning problems with Wikipedia stem from the fact that it is not valued enough to provide the necessary incentives to improve it.

If we put the entire patent process into a wiki (hell, even official approval) then there would be a very strong incentive to counter the flaws of the wiki.

Just talking out my ass. No real proof this work. Like everything else, it would probably just cause more spam and Nigeria scams.

Wisdom of the masses

Yes, please, wisdom of the masses is exactly what we need when patenting is in question.

On the internet, with its anonymity, the so called "wisdom of the experts" is nothing but a meaning. This meaning could belong to Generic Ted, or Bill Gates. Certainly Generic Ted won't unleash his meanings to every patent, but to a subjective share. Bill only registered because he wanted to say that Microsoft should be allowed to patent the
right-mouse-click.

If the answer to that problem is moderating, then why bother ask for the masses wisdom in the first place?

You can't let anyone, but a democratic selected group of people (preferebly wizards) have the right to change the outcome of these important issues.