Patent reform: Apply for a patent, examine some patents

Among many patent reform proposals it is common to have a desire for better examination, and more detection of prior art and obviousness. But the patent office only has so much money for so many examiners.

So here's a simple solution. If you want to apply for a patent, you must put in some time, as an expert in your field, examining other patent applications, searching for prior art and giving opinions on the obviousness. Alternately, this duty could be given only to those who actually are granted patents, to make more sure they are "skilled in the art" of their fields.

Of course, such crowdsourced examiners would have biases. They would be expected to make a sworn statement about their biases. Making a false statement could have implications on their own patents as well as the usual penalties.

Those biased against the patent would mostly hunt for prior art -- in fact they would make the best hunters. Those unbiased could make better opinions of obviousness.

Like regular patent fees, this could be biased for small inventors. (Small inventors pay lower patent fees and get some better treatments.) Large companies might have to volunteer more time from their staff, or small inventors might get reductions in patent fees in exchange for good work. Peers would examine the work of other peers to keep them honest and to rate the quality of it. And of course, unbiased patent examiners and appeal boards would still have the final, objective say.

Other volunteers could also participate in prior art searches. But with the system described above, there should be no shortage of labour. And as the number of patents goes up, the system naturally increases the labour available to do the legwork.


I don't think every inventor makes a good researcher, so having those check for prior art might not be effective. However, patent submitters (or even the general public) could be used to check the state of the art: i.e. when a patent is submitted, people in the field should be given the decription of what the device does, and if they come up with the same solution as the patent submitter, it's a "state of the art" submission that is not patentable.

Especially in the field of software, where coming up with an algorithm typically doesn't require much of an investment, a simple mailing list that patent descriptions are posted to (the description would contain what the patent is supposed to accomplish, not how it does that, e.g. "an algorithm to sort a set of numbers") would suffice: grant those patents that noone replies to, reject the patents where someone else posts the same way to solve the same problem.


I like this idea. Do you have any thoughts about how to enforce/encourage honesty? I can imagine that the "busy" inventors would simply say "yep, seems good to me" since they are too busy to carefully examine it, and they figure that one of the other examiners will spend the time to figure out if it is valid or not. Thus, this would lower the "cost" of applying for patents for the "dishonest" individuals.

The way to enforce honesty, and declaration of bias, is peer review. Part of your duty would be to evaluate the work of your other peers, and there would be reputations attached to people based on such review. You might also be called upon to review the work of other groups (in a more peripheral way) that you're not involved in, but know something about the field for.

We could even start with an anonymous review process, where you don't get to know the names of the other peers. (Later, you must learn their names to judge their bias declarations.) And where attempting to reach them out of band is punished.

"Punishment" in most cases, for both breaking the rules, and work judged insufficiently diligent by peers or PTO workers, would simply involve not getting full work credit for this particular patent. And thus you must keep going until you earn this work credit. You might have a certain time by which you must earn enough work credit, or if this is done to people who apply for, rather than get patents, you can simply make it that you don't get your patent until you have earned enough work credit.

Yes, all the peers looking at a patent might decide to slack off and declare all the other slacker's work as diligent. But there will always be one straight player in the group, and if there is, none of you will get your work credit. If it's judged deliberate, you might even get negative work credit.

For example, lets say one peer finds 20 pieces of prior art, and others find nothing, and some of the prior art is an obvious google search. Looks fishy.

However, it may simply be the PTO's job to judge this. You're working for them, and like any employer, they know how to judge the quality of a consultant they hire. You could elect to pass on any patent you are assigned that does not fit your skills (to a limit) as there will always be plenty to earn your work credit in the several years a typical patent takes.

The idea of recruiting applicants to process applications is an interesting and powerful one. Beyond patents, competitive licensing in general looks like a good domain for deploying this model.

It doesn't work everywhere, but has already proven itself in knowledge (wikipedia), and looks likely work in politics and law too.

The best people to search for prior art are your competitors, so they should be notified at the submission stage, with an open interface for adding prior art links.

Applicants could still do all sorts of work: Add prior art on other applications, process elements of other applications.
How to enforce honesty: Well, there will still be patent staff, and now there will be reputation, a very simple complaints mechanism, and reliability through agreement. And of course there is still legal penalty for making false statements etc.

If people have their work contested and judged in error, then they can have their own applications slowed down, be banned from future contributions etc. Make the balance of rewards so that the best thing to do is put up as much good evidence for their case as they can, not make mistakes. That is basically how peer review works in academia, and in jurisprudence. And, I guess too, in things like planning applications - where local owners are asked to comment on development applications. Making this into a wiki process might be an effective... although fraught, way to arrive at an agreed statement. We already iterate over contracts (agreements) why not iterate over disagreements? That is effectively what diplomacy is designed to facilitate, and what civilisation rewards.

As another idea: what about using it as open-source design optimisation. If you have an idea, push it up, like a patent, but with an Open license. Then people can edit the 'patent' design, optimising it.

Take cookery: I put up a recipe, then people comment on amounts, temperatures, times, substitute ingredients. That is effectively how Michelin * restaurants work - optimising.

Blog comments would be better, to, if they were iterated over :-)

A wiki does seem an interesting approach here. My main concern is you need to treat the biased commenters (ie. competitors, and advocates/sock-puppets) different from the unbiased, as the stakes are higher than a typical wiki article. I think you want more a comment system where contributors can add comments, and then amend their comments, but not remove other people's material. And have a duty to declare their biases.

The goal of the citizen examiner system is to get skilled but otherwise unbiased people on the case.

You do need some authentication, otherwise wiki or comments would get spam. Both from plain old spammers, and also as a denial of service attack on the patents in question. If the examiners have to weed through a lot of bogus comments it slows down the patent.

Add new comment