4 ms·
The approval of a patent registration request should work like this: The issuer provides a public description of the problem the patent solves and a secret des
by Gehinnn 5y ago
The approval of a patent registration request should work like this:
The issuer provides a public description of the problem the patent solves and a secret description of the solution.
They pay $10k in advance to request a patent registration.
The patent institution then publishes the public problem description.
External challengers can submit additional solutions for $100.
After some time (e.g. a month) submission closes and independent judges need to identify the solution proposed by the patent issuer out of all submitted solutions.
If they succeed, the patent will be registered. Otherwise, the submitter of the identified solution gets a split of the patent registration fee (e.g. $2k) and registration is denied.
The institution manages hundreds of possible judges, only a few are selected randomly by a trusted computer. If a selected judge correctly identifies the solution of the patent, they will get a bonus (more than challengers would get, e.g. $3k). Non-selected judges can submit challenging solutions.
This should be robust against bribery and prevent low effort bullshit patents!?
If a company wants to bribe judges, they need to identify the selected judges, otherwise they help not-selected judges to submit challenging solutions. This can be made very hard.
Also, the solution of the patent should solve a hard problem (i.e. a problem that is hard to find solutions for). Otherwise, challengers can easily come up with solutions of equivalent quality.
If the problem is "A mechanism to visually indicate progress", it should not be too hard to come up with the progress bar.
- staticassertion 5y ago> They pay like $10k to test their patent. Cool, so patents are (even more so) for rich people exclusively. Also, what you're describing soudns a lot like the existing system to me, but with more upfront money.
- Gehinnn 5y agoHow can ordinary people challenge patent registrations in a fair way? It is easy to claim that a solution is trivial, but it is almost impossible to prove that if you know the solution already. Feel free to adapt the sums I listed, that would not change the procedure. I do think that there should be a decent financial burden to protect an idea though. If your idea is good, you can still borrow money (or split patent revenue with investors).
- staticassertion 5y agoYes, I suppose my judgment was too harsh, sorry - I don't have a better solution, I just think it's an extremely difficult problem, perhaps not one that's solvable in a financial system like what exists today.