Tuesday, June 12, 2007

.sft: A proposal for software patent reform

As follows:

  1. Software companies that wish to protect their intellectual property register with a new ICANN gTLD, .sft.
  2. A .sft receives "IP points" every time it produces a "significant" software innovation. For example, every time a .sft publishes a peer-reviewed paper in a major computer science conference, that .sft gets 100 IP points.
  3. Any .sft may "sue" another .sft at any time, for any reason, for any quantity of money.
  4. Lawsuits are settled by best-of-7 tournaments of StarCraft. A .sft's designated StarCraft player ("IP lawyer") starts each match with a bonus quantity of minerals, Vespene gas, and peons determined by a time-weighted function of the .sft's IP points. The victor wins a fraction of their client's requested damages determined by the ratio of their buildings razed, units constructed, etc. vs. their opponents'.
  5. IP lawyers may play Protoss, Terran, Zerg, or random race, at their discretion.

The merits of this reform are obvious. Much like patent law, StarCraft is governed by a system of arcane rules that are mostly irrelevant to the actual process of writing innovative software. Much like patent law, StarCraft's rules can only be mastered by a caste of professionals whose expertise is honed over years of practice. Unlike the legal system, however, StarCraft is swift, decisive, objective, and exquisitely balanced for fairness. Any minor loss in the quality of judgment on the margin would be overwhelmed by the reduced transaction costs of the system as a whole.

I can already hear the objections of the closed-minded, and I will respond to them in turn.

  • Q: You're introducing perverse incentives. Companies with better StarCraft players will beat companies who write better software!
    A: Paying for quality StarCraft players would simply be a cost of doing business, no different from hiring secretaries or accountants or paying your electric bill. Even though the market value of Starcraft skills would temporarily skyrocket, StarCraft does not have a government-sanctioned professional guild that artificially limits the labor supply, so in the long run the cost would still be lower than for ordinary lawyers. Anyway, if some company hires bad StarCraft players, the management's incompetent and the company deserves to lose in the marketplace, regardless of how innovative their software developers are.
  • Q: What's to stop the emergence of "IP trolls" who never innovate, but hire a lot of expert StarCraft players to sue everyone?
    A: First, only companies with substantial independent revenue streams would be able to compensate StarCraft players well enough to attract really good ones. Second, beyond a certain point, the bonus resources granted by accumulated IP points would produce an overwhelming advantage for companies that produce innovation.
  • Q: Wouldn't this give an unstoppable advantage to software firms in South Korea, which has an impressive national lead in competitive StarCraft talent?
    A: So what? The international distribution of firms holding technology-related patents in the past two decades has not been even remotely equal. Established United States firms like IBM and Microsoft hold vast patent portfolios, dwarfing those of firms in developing nations, and WIPO has been aggressively working to bring developing nations under First World IP regimes. Yet the software industry remains steadfastly innovative around the globe. Over time, other nations would earn IP points and develop local StarCraft talent, evening the score. In any case, freezing the rules over an unequal or even larcenous initial distribution is the very essence of property: if a Native American robber breaks into your home and takes your television, will you refrain from pressing charges when he's arrested?
  • Q: Your proposal permits any .sft to sue any other .sft, without any basis for infringement. That's absurd!
    A: Under the present system, lawsuits claiming intellectual property infringement can drag on for years and cost millions of dollars even if the plaintiff never specifies exactly what was infringed upon. By permitting any .sft to challenge any other .sft, we remove the fig leaf of "cause of action", which (let's face it) is mostly just embarrassing. Also, as noted above, .sft lawsuits would be swift and decisive, so the transaction costs would be low even for frivolous lawsuits. That said, I am not completely opposed to a ladder system, to save highly skilled lawyers from being constantly challenged by lowbies.
  • Q: IP points would give an unstoppable advantage to entrenched encumbents with deep IP portfolios. What's to stop them from suing everyone?
    A: Under the present patent system, IBM, Microsoft, or any number of other entrenched players with deep patent portfolios could hypothetically destroy the entire software industry in a convulsive paroxysm of lawsuits. Yet they choose not to, because suing everyone in the industry would result in "mutually assured destruction": everyone would countersue everyone else and everyone would lose. I see no reason that .sft lawsuits would be any different.
  • Q: What's to stop incumbents from suing little startups that have lower-tier StarCraft players and few IP points?
    A: Startups today usually have small to nonexistent patent portfolios, but big companies don't find it worth their time to sue them because they also have very few assets worth confiscating. As a startup grows more successful and its pockets get deeper, it should be able to afford to hire StarCraft talent and accumulate IP points, just as companies today accumulate patents and other IP as they grow bigger.
  • Q: The software industry's incredibly innovative to date, and your reform would bring it under a radically new legal regime. How can you be sure it won't stifle innovation?
    A: How can you be sure it will stifle innovation? There's ZERO evidence that it would. Imagine if we'd listened to such naysayers back when software patents were proposed. Don't you think the burden of proof is on the opponents of regulation, rather than on the proponents?
  • Q: StarCraft 2 is coming out soon. How does that affect your proposal?
    A: StarCraft 2 is currently an unknown quantity, whereas Starcraft is a classic that has withstood the test of time. That said, we should keep an open mind about the gameplay innovations, and if they prove to be successful I am not opposed to adopting the sequel someday. For one thing, the improved graphical sophistication would make lawsuits more entertaining for bystanders to watch.
  • Q: Halo would be more exciting than Starcraft.
    A: Fuck off, frat boy.