The Supreme Court heard a case on patents this week. At issue is how do you determine if an invention is obvious and therefore not deserving of a patent. The problem is, the test is subjective. To help clarify the word obvious, the test is if "teaching, suggestion or motivation would have led a person of ordinary skill in the art" to the invention. Scalia called that "gobbledygook". Roberts asked :Who do you get to be an expert to tell you something’s not obvious? I mean, the least insightful person you can find?"
It's precisely these kinds of cases where I think strict constructionists lose. Yes laws should be precise and easy to understand but this isn't always possible. For example, the famous pornography definition. Having too many patents for obvious things really does harm the software industry and I can imagine it hurts other industries similarly. The Constitution protects us from unreasonable searches, who decides what's reasonable?
Unlike most of Congress, I read the Patriot Act. To be honest, I thought most of it was ok. Sure there are things in there that subtly change various laws and I defer to the ACLU or others to figure out what rights I'm giving up, but a lot of it was clarifying laws that were too specific. E.g., your right to privacy on the internet was different if you used Dial Up, DSL or Cable to get on the net. The law said the phone company couldn't allow tapping without a court order, but said nothing about cable companies, so you were more secure using DSL vs a cable modem. How dumb is that? As I understand it copyright law is different for vinyl records vs CDs (or at least it was). If you want laws that apply to new technologies, and don't go on for thousand and thousands of repetitive pages, you can't require every little thing to be itemized in the law. And summaries will at some point need clarifications by judges. I'll be really curious to see if they can come up with a clearer test for obviousness.
5 comments:
Was this post necessary, it seemed kind of obvious?
The root problem isn't the definition of "obvious", but that the resources (staff and dollars) of the US Patent Office has not kept up with the number of applications. As a result, overworked and relatively underpaid examiners issued a significant number of patents that should never have been seen the light of day.
Hence more challenges are coming to invalidate patents that likely would never have been issued 20 years ago.
There are lots of problems, not sure anyone of them is the root one. certainly not enough staff to keep up with demand is one. Should business processes or genes or other things be patentable at all is another.
This is JUnit test for obviousness that you seek:
a) Laws applying to general technologies
assertEquals( TechnologyAbstractClass.law1().isObvious, true )
b) Laws applying to existing specific technologies
assertEquals( TechnologyConcreteClass1.law1().isObvious() , true );
c) Laws written for 1 technology being applied to another technology
assertEquals( (TechnologyConcreteClass1) TechnologyConcreteClass2.law1().isObvious(), true);
Maybe lawmakers need to study OO design. This is clearly untestable.
New technologies ...almost by definition... return false if you call their isObvious() method....even if such a method exists. Then there is the casting problem in example c. I have no idea what to expect.
With respect to what should and should not be patentable, broad changes - and I agree there should be some broad changes - should be changed thru legislation, not the courts. Once enacted a new law could be part of the USPTO, which is an executive agency, part of the Dep't of Commerce.)
The "obviousness" doctrine actually works if the patent process is working. The patent applicant must convince a trained patent examiner proficient in the art that the claimed invention is indeed patentable. It is when the system breaks down that poor patents get issued.
The EU patent office does it well, so it can be done...
Post a Comment