4 ms·
> I get how we all intuitively want the PTO to not issue "stupid patents" but is there an objective, mechanical way to determine what is stupid? There's a very
by JoshTriplett 11y ago
> I get how we all intuitively want the PTO to not issue "stupid patents" but is there an objective, mechanical way to determine what is stupid?
There's a very simple, constant-time, objective, mechanical algorithm to determine whether to grant a patent: "return false". That would work better than the current algorithm, which closely approximates "return true".
At the very least, it should take no less effort or inventiveness to get a patent granted than to get an academic paper published in a reputable journal or conference. But it seems highly unlikely that respected academics will jump at the chance to work in patent examiner's offices the way they do to program committees.
> Do we care more about false positives (as in this case) or false negatives (a hypothetical alternative universe where too few, rather than too many, patents are awarded).
False positives. A denied patent is highly unlikely to suppress a good idea; good ideas (and a fair many bad ones with good execution) will succeed whether granted an artificial monopoly or not. But a granted patent can litigate good ideas to death.
> how are individual supposed to profit from their inventiveness when a larger entity could trivially copy the idea
Many wildly successful companies are built on technologies that could easily be copied, and yet aren't. The sole legitimate measure of the patent system is "how many useful inventions do we have that wouldn't have been invented without patents?". How many ideas (outside of the drug industry, which could potentially use some other solution) wouldn't have been created if not for patents?
Apart from that, why do you believe an inventor should be able to invent once and profit for 20 years off the work of others, without continuing to invent and promote further good ideas? Keep inventing new ideas faster than they can be copied, so that the copy is always last year's technology. Make your customers so happy that they'd never want to consider a competitor. Do any number of things that add value to the world rather than subtracting it.
> Are software inventors just supposed to give away their inventions as open source
Preferably, but copyright still exists for those who want to release proprietary software. Copyright stops others from directly copying code. Patents stop others from creating new software, even if not copied from the patent holder. The former seems far more defensible than the latter.
- throwawaykf05 11y ago>There's a very simple, constant-time, objective, mechanical algorithm to determine whether to grant a patent: "return false". There are several empirical studies that indicate this is wrong. Google and ye shall find. > That would work better than the current algorithm, which closely approximates "return true" This is objectively false if you look at publicly available data on rejection rates and understand how the process works. > good ideas (and a fair many bad ones with good execution) will succeed whether granted an artificial monopoly or not. Oh yes, they will succeed, but from past experience, not always will they reward the original inventors. > But a granted patent can litigate good ideas to death. How often has that happened? Note that if you want to mention examples like Watt or the Wright Brothers, those apocryphal accounts have been debunked. > Many wildly successful companies are built on technologies that could easily be copied, and yet aren't. And many dead companies were build on technologies that were too easily copied by bigger competitors. > Apart from that, why do you believe an inventor should be able to invent once and profit for 20 years off the work of others... Because a good idea is something that has value forever? Should an inventor not reap some rewards off that? >... without continuing to invent and promote further good ideas? Non-sequitur. What makes you think an inventor will not continue to invent? > Keep inventing new ideas faster than they can be copied, so that the copy is always last year's technology. This severely underestimates the speed at which technology can be reverse-engineered and duplicated. > Copyright stops others from directly copying code. Often the value is in the idea or the solution embodied in the code, not the code itself. Copyright does nothing to prevent that, even though it could be the idea which is hardest to create. > Patents stop others from creating new software, even if not copied from the patent holder. In practice, they don't. New software is being created at an unprecedented rate, even as software patents have been in effect for the past two decades. I'd agree we need a much higher bar for patents, but suggesting we do away with patents entirely is being dismissive of the unpleasant realities of the world.
- JoshTriplett 11y ago> There are several empirical studies that indicate this is wrong. Google and ye shall find. You're going to have to be more specific about what studies you have in mind. Considering the lack of a control group without major confounding factors, it'd be quite difficult to have a study on the efficacy of a patent system at increasing the number of inventions made available. > Because a good idea is something that has value forever? Should an inventor not reap some rewards off that? Remember that the purpose of the patent system (at least in US law) is to encourage more ideas; providing a temporary monopoly to reward patent holders is a means, not an end. At the time the patent system in the US was created, hundreds of years ago, it was believed to be a net win for the general public; at that time, it might even have been possible to read and learn from a significant fraction of the patents granted. With the current culture and wealth of inventiveness, of myriad people happy to release ideas for all to use, and the lack of any value gained by reading the firehose of patent publications, it's quite reasonable to ask if we still benefit from the tradeoff of the patent system, or whether we'd get more benefit by abolishing it. In evaluating that, not a single moment's thought should be given to rewarding patent holders as an end unto itself, only as a means to encourage more invention. > How often has that happened? Too often. I've seen far too many FOSS projects suppressed, features left out, or avenues unexplored, precisely because an area is a patent minefield. Cryptography, audio codecs, video codecs, compression technologies, encoding technologies, and numerous other useful code blocked because someone else staked a claim first. What value is gained when someone pops up and sues some browser or device vendor over some ludicrous patent (in East Texas, naturally), and wins millions? Does anyone seriously believe that those browser or device vendors actually read the patent and said "that sounds like a good idea, let's use it and rip this lone genius off!"? Even if that random troll actually did happen to be the first person to submit an application to the patent office, they've provided absolutely nothing of value. You can write code entirely from scratch, create something useful, publish it, be sued by someone you've never even heard of, and lose, all because of an artificial monopoly over ideas. > suggesting we do away with patents entirely is being dismissive of the unpleasant realities of the world. Patents are one of the unpleasant realities of the world, and patents are one of the main things making it necessary to acquire more patents, as a defense mechanism.