5 ms·
That sounds like it should not be pantentable. At the core of the argument, software engineering is applied math. Math is not patentable. Software patents are
by wiseleo 7y ago
That sounds like it should not be pantentable.
At the core of the argument, software engineering is applied math. Math is not patentable. Software patents are workarounds around the fact that math is not patentable. Yes, the work is valuable, but it should not be patentable.
Here are USPTO's guidelines https://www.uspto.gov/web/offices/pac/mpep/s2106.html https://www.uspto.gov/web/offices/pac/mpep/s2106.html
- rayiner 7y agoAll engineering is just “applied math.” It’s the “applied part” that makes it patentable. (You cannot patent an algorithm. You can patent a method of getting getting radios to aggregate into networks. Or a method of wayfinding for a self driving car. Or a method of indexing and ranking web pages.)
- wiseleo 7y agoWhy? The purpose of a patent is to grant a monopoly for an invention that would not be obvious to someone skilled in the art. These business methods should not enjoy patent protection that prevents someone from independently discovering and implementing same solution. There is a finite number of ways to efficiently implement wayfinding for a self-driving car.
- bakuninsbart 7y agoFunny you mention wayfinding algorithms: There had been decades of progress in that regard without a single patent, only the last decade (since big business got in the game), the rules have suddenly changed.
- rayiner 7y agoI’m pretty sure the CMU folks have been parenting their stuff The whole time: https://www.cs.cmu.edu/~tjochem/nhaa/ralph.html https://www.cs.cmu.edu/~tjochem/nhaa/ralph.html
- jacques_chester 7y agoMany universities obtain patents on research outcomes as a matter of policy.
- GoblinSlayer 7y agoHardware engineering works with matter which is not subjective and so it can't be handled by pure imagination as opposed to math and software.
- jacques_chester 7y agoYou can turn any software into an equivalent piece of hardware, which meets the non-subjective matter test you're proposing. Similarly, I can convert any hardware into a software model of it, which makes it into pure imagination.
- GoblinSlayer 7y agoYou can't convert hardware into software model, the software model is necessarily an imperfect idealization, created deductively even, which thus doesn't necessarily work IRL, because fundamentally it's imagination.