This morning, Richard Posner, the esteemed jurist and prolific author of legal works, published a piece in the Atlantic about the sorry state of American Technical patents. The piece is not intended as an indictment of the patent system in total, but of the specific quagmire that is Software Patents. For the uninitiated, some of the largest costs of doing business today are associated with Legal Warfare. Seemingly every day a new patent troll springs up to assault innovation (trolls being patent holding companies whose sole reason for existence is the extraction of wealth from companies via the Legal system. Large organizations will transfer the control of their patents to these smaller entities so they become essentially proxy cannons in the fight for technical domination). What can be done about this?
This is a complex issue with an interesting origin. As Richard points out in his piece, there are indeed a number of industries which need Patent protection to exist (his example is Pharmaceuticals, which by all accounts deserve protection due to the costs associated with development). The crux of the issue is that the cost of development for other markets has disappeared seemingly overnight. Nowhere is this more true than in the development of software.
Ten years ago, if one wanted to build a startup, the initial funding involved $5million to EMC for data storage and $5million to Oracle for a Database. These were the costs of innovation, but with the advent of Amazon’s Web Services infrastructure and the prolific development of open-source databases these costs are gone. Today, creating a big web application starts with a someone logging onto a website, buying a server and starting coding. The initial infrastructure costs have been essentially eradicated. This is not the case in Pharma, where the costs of testing and researching have increased significantly over time, as well as the average time to market from the time of discovery.
Second, the process of granting patents involves careful consideration from learned professionals. The requirements for obtaining a patent are that the invention is “novel, useful and non-obvious”. The problem, which is immediately obvious ironically enough, is that what is “novel, useful and non-obvious” to you, might not be to me. For example, imagine reading a patent for database sharding, without an education in database technology. One might think this were an amazing innovation and that it was non-obvious, when in reality, to those in the know, it’s quite obvious. This is precisely the problem; whereas previously the patent office could determine what is and what isn’t obvious, what’s happening now is that the lines between innovation and repetition are blurred. What’s worse is that the patent process actually make this worse: here’s how.
When a patent is submitted and approved, no one is really checking for prior art. A patent is granted, and then it becomes the responsibility of individuals who disagree with the patent to challenge the owners in court. If you make a rice cooker and patent it, and I make a rice cooker and patent it 3 years later, the only thing that invalidates my patent is for you to sue me and win. This, as you can imagine, becomes a HUGE waste of time, especially when you realize there are literally millions of patents covering just telecom. So what can be done? Simple.
Crowd source patent research.
There are many experts in the world that would volunteer their time to dissect patents and determine prior art. The main issue is that when patents are granted, it is too painful to check for a previous patent, so no one does. But leveraging the power of the internet would allow helpful users to point patent authorities in the right direction. Allowing comments and input from the public on Patent approval will create a vibrant discussion of innovation and will reward new technology while relegating repetition.
Simply put, the Patent system is broken, primarily because of lax standards associated with Patent awards. The solution is a stronger system, powered on the strength of distributed American ingenuity. Intellectual property can be fixed, but it will require a lot of effort. Let’s spread the work using technology to alleviate this burden from our innovators. Let’s get back to work building tomorrow instead of defending the past.