The Washington Post is reporting today that the U.S. Patent and Trademark Office (USPTO) will soon begin experimenting with a wikinomics approach to reviewing patent applications. Anyone who’s been keeping tabs on the intellectual property system knows that this idea is long overdue.
The number of patent applications has tripled since the 1980s, while the number of patents granted has doubled. Some might argue that the rate of invention has dramatically increased, but the more likely answer is that the quality of patents has sharply decreased as underpaid and overwhelmed patent examiners struggle with a backlog of over 600,000 applications and growing.
Most of the experts I’ve spoken to agree that the decreasing quality of patents has become a serious political and economic problem. Patents are often granted for “inventions†that have long been in practical use. In other cases, firms file patents for inventions that they never intend to commercialize. Instead they use the patent system to generate revenue by ‘holding-up’ other companies that actually make and sell products. Robert Barr, chief patent counsel at Cisco until July 2005, put it best in his deposition to the Federal Trade Commission:
“Companies try to patent things that other people or companies will unintentionally infringe and then they wait for those companies to successfully bring products to the marketplace. They place mines in the minefield. The people and companies who file these patents and extract license fees from successful businesses play the patent system like a lottery.â€
The real danger is that these questionable patents will end up driving up the costs of innovation by generating an increasing number of lawsuits, or threatened lawsuits, that genuinely innovative companies cannot avoid. The problem is so endemic in some fields (especially software and electronics) that low-quality patents have become a serious drag on the technological and scientific progress that the patent system was designed to promote.
The experimental new system is called “The Community Patent Review project” and its supported by companies such as Microsoft, Intel, Hewlett-Packard, Oracle, and IBM. Here’s how they describe their solution:
The Community Patent Review project aims to design and pilot an online system for peer review of patents. It will support a network of experts to advise the Patent Office on prior art as well as to assist with patentability determinations. By using social software, such as social reputation, collaborative filtering and information visualization tools, we can apply the “wisdom of the crowd†– or, more accurately the wisdom of the experts – to complex social and scientific problems. This could make it easier to protect the inventor’s investment while safeguarding the marketplace of ideas.
I interviewed Beth Noveck, the originator of the Community Patent Review project, when we were researching Wikinomics in the fall of 2005. We couldn’t include a write-up in the book, but I am happy to see that the project is succeeding.
Noveck said she came up with the process because she spends a lot of time thinking about how we can redefine democracy and enable more open and participatory forms of government.
“The idea of having a single person with fewer than 18 hours, and no advanced training, ruling on what could amount to a twenty million dollar monopoly (or more) is asinine” she said. She’s right, of course. Decisions on which patents to grant or reject literally determine the future of economic competitiveness. The fact that the average patent receives less than 18 hours of review is irresponsible, to put it lightly.
Noveck argues the problem would be tractable if we were to widen the group of experts who feed input into the process — which is exactly what the Community Patent Review project aims to do. Some critics of the proposal worry that the system could be gamed. But Noveck points out that current patent review system is more susceptible to abuse than one where a community of peers can review and rate each others comments on a given patent application in an open and transparent forum.
Unfortunately, most bureaucratic systems are not designed to enable mass collaboration. As Noveck put it:
“Our legal and political conception of bureaucracy and administration assumes that centralized experts have greater wisdom than anyone else. We believe that they are in the best position to make dispassionate decisions in the public interest and that they have the access to the best information.”
Noveck goes on to say that:
“While that may have been true one hundred years ago given the conditions of the communications and information infrastructure, it is no longer true today. We now have ubiquitous information networks and the ability to connect people in a way that enables us to tap into collective intelligence.”
It’s nice to see the Patent Office giving this idea a try. Let’s hope that other governmental bodies will follow suit.
