Software Obviousness: Judges just don't understand computers

One thing that is often overlooked in a discussion of patent obviousness is the fact that there is a great deal of expertise that can be required to determine whether an invention was non-obvious or unique. When it comes to the software industry, these inventions are often in the form of lines of code, and these products are often less tangible inventions than those in the natural sciences.

The implication of this is that judges don't have a comprehensive understanding of software and the methodology involved in developing software and as such, they apply the same rules to computer software as they would to the natural sciences in determining obviousness. The result: way too many software patents are being granted for inventions that most people would deem pretty "obvious to try", the standard that was set up in KSR v Teleflex.

The problem with software inventions is that it is too easy for a patent attorney to claim that the notion of an invention being "obvious to try" may just be conflated with hindsight bias, since the field of computer science is transforming so rapidly. "Engineers overwhelmingly feel that the 'inventions' in most software patents are no more than 'the predictable use of prior art elements,' reflecting 'ordinary skill and common sense'" (Purvy). In other words, those who are familiar with the field of software engineering understand that a lot of the inventions are not actually original but rather just predictable use of prior art, meaning that they would be "obvious to try."

Thus, I (and the author I cite, Robert Purvy) propose that the "obvious to try" standard set up in KSR should be applied to software patents as well in order to solve this problem.

Source: http://www.law360.com/articles/525451/a-solution-to-the-software-obviousness-problem