Leveraging IT and Automation
From February 14 to 16, the TV show Jeopardy! featured a man-machine showdown that piqued the public fancy. An IBM computer named Watson beat out a couple of human champions at that show’s sophisticated quiz game. On February 15, using the battle as a news peg, Martin Ford published an article in The Atlantic suggesting that soon computers will be able to do the job of a lot of lawyers. Last weekend the author also spoke about it on National Public Radio.
Ford’s article reminds us that over the past several years computers have been developed that beat chess champions Kasparov and Kramnik and then speculates that: “If you agree that the game of chess requires creativity within a set of defined rules, then could not something similar be said about the field of law?
“Currently there are jobs in the United States for many thousands of lawyers…in the areas of legal research and contracts. They…spend much of their time in the library or accessing legal databases through their computers.…
“What about automation? Can a computer do the lawyer’s job? One of the primary research areas in artificial intelligence has focused on creating ‘smart’ algorithms that can quickly search, evaluate, and summarize information. We see the fruition of this body of research every time we use Google or any other advanced Internet search engine.
“We can expect that such smart algorithms will increasingly be used in the field of legal research. The software may start out as a productivity tool to make the lawyer’s job easier, and then eventually evolve into a full automation solution.” The New York Times published a similar observation on March 4, as part of its excellent “Smarter Than You Think” series.
Coincidentally, on March 8, Alexander Poltorak, CEO of General Patent Corporation and founder of American Innovators for Patent Reform, was granted U.S. Patent 7,904,453 for Apparatus and Method for Analyzing Patent Claim Validity.
Does Poltorak’s patent disclosure lay the foundation for a computer that would eliminate the jobs of patent examiners? Well, maybe it won’t go all the way, but it should point us toward more efficient use of their expertise, so as to reduce the backlog and help to realize the goal of better patent quality.
Giving Examiners the Proper Tools
As I said, critics assert that all too often the PTO issues “junk patents.” Everyone agrees that those patents that claim subject matter already in the prior art have been “anticipated” under section 102 of the U.S. patent law and should not have been issued.
When Jon Dudas was director of the PTO from 2004 to 2009, he touted a declining rate of allowance of patent applications as evidence that patent quality was increasing. However, I believe that, by and large, the patent examination practices of that era merely created a bottleneck, which led to longer pendency but not much of an improvement in the ratio of valid patents to invalid ones.
To increase patent quality, I believe that the process of examining patent claims needs to be re-engineered. A proposed patent claim has to get past several legal hurdles before it is deemed legally allowable.
Section 103 of the U.S. patent law requires that the subject matter of the claim not have been “obvious” to a person having ordinary skill in the art. Until this fuzzy test is made crisper, either by statute or court interpretation, I think the judgment call of “obvious or nonobvious” will still need to be made by a person, not a machine.
Yet, isn’t it time that the society with the most powerful information technology in human history provide to those persons, i.e., examiners at the PTO, a tool that presents an organized report of the prior art that’s most pertinent to each element of each claim under examination? In effect, it would be an automated version of the Examination Support Document (ESD) that in 2007, the Dudas-led PTO had announced it would require many patent applicants to generate.
In the face of widespread criticism, the PTO thankfully withdrew that proposal. Criticism focused on the extreme burden on patent applicants being required to manually prepare such analyses as well as the legal risk of creating an Achilles heel by making an inadvertent error or omission. After all, to err is human.
Nobody said, however, that if a patent examiner were to be presented with such an analysis, it wouldn’t help him do his job. These days, a remnant of the proposal remains in the option to seek accelerated examination by submitting a similar report.
Back in 2007, I was impressed by a publicity piece issued by the intrepid folks operating the PatentCafe.com website. They announced a commercial service to automate the creation of an ESD.
At that time, I published an editorial in Biotechnology Law Report urging that the PTO provide such technology as an in-house tool for use by patent examiners. Doing so, rather than requiring an ESD from outside, would relieve the applicants and their attorneys of the burden and risk and yet lead to patents of improved quality.
I wrote: “This would be analogous to the deployment of the automatic telephone exchange after anticipated growth of the switchboard phone system couldn’t be supported without recruiting every woman in America to work as a telephone operator.”
Sadly, I am not aware that the PatentCafe ESD product as originally announced ever reached fruition. The impetus for it dried up when the PTO dropped the proposal to require certain patent applications to include an ESD. However, the project by PatentCafe seems to have morphed into what is now Pantros IP. In addition, the patent-searching tools offered by Lexis-Nexis now include an option to employ semantic analysis. So the requisite technology seems to be evolving as needed.
Now that IBM has demonstrated the ability of their Watson system to draw sophisticated conclusions from a vast repository of text, I submit that the time has come to enlist such technology in the service of the PTO toward more speedily generating a higher quality patent examination. With one stone, this should help kill the two birds of “junk patents” and the backlog of unexamined patents described above.
What Legislation Should Cover
With patent reform high on the Congressional agenda, it is imperative that the legislature support implementation of computer-assisted patent examination.
There seems to be hopeful news from the House of Representatives. Rep. Bob Goodlatte (R-Va) is slated to spearhead the House version of the Patent Reform bill. He said in a CSPAN interview that the most likely focus of the legislation would be on “increasing the speed and quality of the patent process,” commenting further that the PTO “doesn’t use enough technology to ensure products are unique and worthy of a new patent, making the process slow and cumbersome.”
It seems to me that Rep. Goodlatte is spot-on. It is imperative that Congress encourage PTO director Kappos to invest in 21st Century systems to transform the patent-examination process. Having come to government service from a career at IBM, he is uniquely qualified to lead us toward this beneficial transformation.
Until the PTO provides a Watson-like analytical infrastructure for patent examiners, the whole patenting process will remain in jeopardy. In the legendary words of Alexander Graham Bell: “Watson, come here. I want you.”