Like a lot of ideas and concepts springing from government and universities, letting colleges and universities profit from patents made by their students and faculty looked good on paper but has generally failed in the real world. Sure, there are a few home runs, like the process for inserting foreign DNA into cells from Columbia University and the development of Remicade, a drug to treat autoimmune diseases from New York University. The first has brought in more than $790 million and the second more than $1 billion to the schools where they were developed.
There’s also a study that says universities make more than 4,000 patent licensing agreements annually and collect some $2 billion per year in licensing revenues. But that study comes from the Association of University Technology Managers, the people who make a living do this. They certainly have good reason to slant the findings. They don’t want to lose their jobs.
According to another study from an outsider at the Brookings Institute, the one third of licensing fees that go to a university’s general fund (the other two-thirds go to the researchers and their academic department) barely cover the cost of operating a tech-transfer office at 13% of research schools, and the great majority fail to turn patents into an income stream.
Many that do manage to sell a patent or two end up selling them to trolls like Intellectual Vultures, a firm referred to on the web as an “infamous patent troll.” (Over 60 universities have sold “intellectual” property to that group.) Selling to trolls stifles innovation and raises the cost of business for everyone, except trolls, who even like invalid trolls. (On the good-news front, trolls lose 75% of the time when taken to court. On the bad-news side, the winning company must still pay legal fees. Time for “loser pays” in the U.S. courts.)
According to Stupid Patent of the Month, a sad but hilarious look at today’s patent situation on the Electronic Frontier Foundation’s website , Penn State jumped aboard the patent gravy train just two years ago. It tried to auction off a dozen of its patents. Only one earned even a single bid, so the college likely lost money on the auction. (Good job, PSU Technology Managers.)
The patent that was purchased is titled “Agent-based collaborative recognition-primed decision-making,” a title that makes me think it took more time to write the patent application than come up with and refine the idea. The application includes such insightful and informative gems as:
“Story building also involves information gathering, but it is more than cue-driven information investigation, because the agents are still unclear about what cues to investigate. Therefore, the key is to identify a collection of cues which the team needs to pay attention to. Our model adopts a combination of two mechanisms: hypothesis exploration and experience synthesization.”
The patent examiner originally rejected the patent. So Penn State went back to the drawing board and added the phrase, “include a team-oriented computer architecture that transforms subject matter.” In English, that translates to: “and use a computer to do this stuff.” Bingo: The patent was now valid.
The Patent Office certainly needs revamping, much like our legal system. Perhaps having patent agents require that inventors have a working model of their invention would slow or stop the flood of stupid patents. Or maybe require that patents be written in English at an eighth-grade level. Not unintelligible lawyerspeak. And colleges should try to improve their educational capabilities and stop grubbing for dollars.