I've also been following some of the discussion on Twitter, and the overrall theme for Nilay seems to be to ask for a citation for any claim he doesn't agree with, e.g. the fairly uncontroversial claim that any non-trivial piece of software will infringe many patents.<p>What I don't ever see are his citations justifying the underlying claims of patent theory laid out here: <a href="http://en.wikipedia.org/wiki/Patent#Rationale" rel="nofollow">http://en.wikipedia.org/wiki/Patent#Rationale</a><p>1. Patents encourage R&D. The relevant criticism here is to ask: does this incentive outweight the costs of the patent system as a whole? (These costs are many. They include: any friction on innovation generally; the costs of running the patent office; patent application; patent litigation; defense against patent litigation; patent portfolio acquisition; the opportunity cost of all these people spending all this money, time, energy, and brain power on patents.)<p>2. Disclosure. Relevant criticism: is it in fact that likely that interested parties wouldn't be able to independently develop a given innovation without the patent system? It seems fairly uncontroversial to say that, as it is, a company will only apply for a patent if they think the patent protection will actually last longer than any protection they would get from trade secrecy. I.e. if a company really thought trade secrecy would keep an innovation out of the hands of their competitors for longer than a patent would, they just wouldn't patent it and keep it as a trade secret instead. So almost by definition this is pretty clearly false.<p>3. High fixed cost, low marginal cost innovations don't have enough incentive. This seems like the strongest argument for patent <i>on some things</i>, but I have yet to see any strong citations to research that isn't done by or paid for by pharmaceutical companies. It's an empirical claim that needs to be tested, and it seems to me that the burden of proof is on the proponents of patent law, not opponents to disprove it.<p>4. Designing around patents leads to more innovation. This seems like a fairly obvious case of introducing artificial inefficiency. Sure, you might get some unrelated and unexpected benefits from all of this incidental research that you're forcing, but it seems pretty obvious that this is going to be a net friction on overall innovation.