In between watching the word finance system heave to and fro like a rubber dingy in an artic storm, I’ve been trying to get some work done. Part of this has been looking into issues surrounding software patents, both in the US and internationally. This is an area I’ve got some background in, having filed several over the years as part of my previous work at Microsoft. All is not well in the world of software patents.
There are important structural problems that waste huge amounts of time & money, do not reward inventors and entrepreneurs, and slow down innovation. This is all counter to the purpose of patents.The concept of a patent is to grant an inventor a temporary monopoly so that he or she can reap benefits from their invention, and thus be encouraged to invent and bring new and useful products to market. Society grants the temporary monopoly to the individual because it is good for society to do so - there is no "natural" right to a patent, it's a societal construct. Nothing wrong with that - let's make sure it works correctly! A secondary benefit of patents is to establish a depository of published know-how; so that once the patents expire anyone can make use of the ideas. Again, all goodness here.So where is it breaking down today?
1. Patents cost too much to file. It can easily cost upwards of $40,000 to successfully file a good patent and sheperd it through to acceptance. Qualified lawyers who know how to file and defend patent applications are necessary. This prices many small innovators out of the “protected by patents” market. – So a vital class of innovators is not served by the patent system.
2. Paradoxically, despite the cost, too many patents are filed. Big corporations are extremely fearful of being sued over patents. There is good reason for this fear; technical issues are complex to explain to laymen so big companies with lots of cash are at great risk for being sued by “patent trolls” (see next point). To defend themselves big companies have an incentive to patent even small and common innovations on the grounds that they may get sued if the don't - they "bulk up" on patents. No innovative purpose is being served here.
3. Patent Trolls. Simply put, a patent troll is an individual or organization that does not produce useful products, but rather simply files or acquires patents with the intent to sue the useful parts of society that do produce products (ask RIMM how happy they are for being sued for $600+ million for a spurious mobile e-mail patent). It’s an ugly, parasitic and lucrative endeavor.
4. Length of patents. A hundred years ago when markets were local and small, and designing and producing new products en-mass took years a longer patent term made sense. Inventors and entrepreneurs needed more time to recoup their investments. Today, the patent term of 20 years is absurdly long. Consider this historical fact… Windows 2.11 shipped in 1989…remember that? (I barely do) Flash forward twenty years to 2009, everything has changed - literally everything. The way software is created and distributed is vastly different. Computers themselves are vastly different. The uses of software are vastly different. The raw amount of software is vastly different. And between then and now, software innovation has sped up dramatically (a little thing called the Internet helped). Software lifecycles move faster now, markets are much broader, clearly an innovator does not require 20 years “lock up” on an idea encourage them to invent something new. Instead, this 20 year lockup is very good for patent trolls, and somewhat good for large corporations. For society, it is anti-innovative.
So what should be done? I propose two things:
1. The length of a software patent should be 5 years. A five year head start on capitalizing on a good idea is more than enough time to generate a healthy reward for an inventor who intends to bring their idea to market. A 5 year limit would also serve two other important goals: First, it would vastly reduce the incentive for patent trolls to file spurious patents on products they do not intend to produce – since they will only have 5 years to find and harass their victims. Second, a 5 year monopoly on an idea would force big companies to really consider what the value of filing a patent is and reduce the overall number of patents flooding the system.
2. The software code required to implement the invention should be published along with the patent. As above, there are two good reasons for this. Firstly, it ensures that the inventor actually took the time to invent something rather than simply patenting a general idea (often it is very hard to tell the difference). Secondly, upon the expiration of the patent the know-how associated with the patent is readily available for interested parties to learn from. This is pro-innovation and anti Troll. It should serve to reduce the number of frivolous patents, increase the quality of the patents, and encourage real long term innovation.
I believe both of these suggestions are important, but the first one is vastly more important. Patents are a good idea. Software patents are a good idea. But it’s time to bring the patent system up to date with the way innovation and industry works today. Software is such a vital part of our society today, and such a large part of the economy, that sensible and pro-innovation rules must be shaped to encourage innovation and reward invention and entrepreneurship.
Subscribe