"The USPTO kicked the question to the appeals court because it wants the power to pick and choose what is patentable subject matter—rather than merely performing its well-established mandate to examine patent applications. At best, this is plain lazy. At worst, this push is economically stupefying. How can the USPTO recognize the next Google—particularly if it won't even examine the invention? Admittedly, it's always been difficult to distinguish "abstract" ideas, which are not entitled to patent protection, from "applied" ideas, which are. In short, E=mc2 is not patentable, but applying the formula within a machine (as software on a computer) to run a nuclear reactor more efficiently and safely is patentable."