The examiners regularly approve absolute bullshit patents in my field that either obviously have prior work, and shouldn't be patentable anyway such as game mechanics. They clearly don't understand the work they're meant to be doing. Either patent law is horribly designed and needs to be razed to the ground, or it's being horribly applied.
With that being said, the majority of the time, the examiner made the right decision. You should check whether a patent was actually granted, for instance. Often when people are complaining about a dumb patent they're actually complaining about a dumb patent application that the USPTO rightly rejected. You should be complaining about the people writing such applications, not the USPTO.
Further, the USPTO is funded purely by fees, not taxes. Applicants want patents. That creates a perverse incentive to reduce patent quality to make it easier to get patents.
I've elaborated on these issues at length on HN before: https://news.ycombinator.com/item?id=36563611
https://en.wikipedia.org/wiki/Asymmetric_numeral_systems#Pat...
https://www.theregister.com/2021/03/13/microsoft_ans_patent/
https://patentcenter.uspto.gov/applications/16456602/ifw/doc...
Look for "Notice of Allowance and Fees Due (PTOL-85)" and click on "PDF" on the right. Scroll to page 10 and look for the "Reasons for Allowance" section where the examiner describes in detail why it differs from the prior art.
While there are minor technical differences in exactly how rANS has been encoded/decoded before, and how Microsoft does it, the fact that Microsoft was granted this means they now have a weapon with which they can cause fear, uncertainty and doubt around ANS, much to the chagrin of the ANS's actual inventor, Jarek Duda, who wanted it to be public domain and implementable by anyone.
I'm not an expert but Duda and fellow compression experts looked at the claims themselves: https://encode.su/threads/2648-Published-rANS-patent-by-Stor...
It seems to me like Microsoft got a patent on "doing ANS a little bit different" - they didn't have to, they could just do it the normal way, but this little bit of difference lets them secure a patent, and now they can pursue anyone who implements ANS to intimidate them with "how sure are you don't do ANS like we do? Let's get our multi-billion legal team, and your legal team, and find out. You have a legal team, don't you?"
In particular, this patent already had a final rejection in 2020. But Microsoft then took advantage of the "After Final Consideration Pilot" program, which sounds more like the USPTO trying to drum up trade, to get it re-re-re-examined.
This is due more to people not understanding what the patent covers. The right response in my view is to educate people. Just because someone has a patent on a particular variation of X, doesn't mean that working on X is risky or what not. Just don't infringe their variation. When I was at the USPTO, I examined a lot of little variations of common things in my area (water heaters and car air vents, mostly) and I never worried that it would stop innovation as usually the point of novelty was not particularly groundbreaking, or even necessarily of interest to anyone aside from the applicant.
Do you feel qualified to discuss patents related to software? If so, why?
I do remember your comments from past threads too. It really interesting to hear the perspective from the patent office's side, but the idea that the patent office had some secret and specialized method of evaluating novelty is ridiculous. Any expert can read a sample of granted patents and tell you that. I'd estimate maybe 5% of patents in my field have any novelty, and that's being generous.
I'm sure this has more to do with incentives and the overall system, and that individual patent examiners would prefer to do a good job. But you have to admit that the results are atrocious.
Just because they said it was granted, doesn't mean that it was. A lot of people here don't seem able to distinguish between a granted patent and a rejected patent application. Here are two examples that I bothered to reply to in the past:
https://news.ycombinator.com/item?id=38766101
https://news.ycombinator.com/item?id=36563425
> the idea that the patent office had some secret and specialized method of evaluating novelty is ridiculous
I don't think they do and I never said they do. The USPTO follows some legal standard that I personally don't agree with. I agree with you that too few granted patents have genuine novelty.
> But you have to admit that the results are atrocious.
No, I don't. You've seen a small selection of what the USPTO outputs. Only the bad cases appear in the news. In contrast, I've seen a far larger and unbiased selection and know that the majority is fine. Most applications are rejected. I probably rejected over 75% myself.
Is there a way to sample 5 random ML patents? I'd be surprised if half were any good.
I think the quality of examination and search is excellent given how little time examiners have. But mistakes still happen too frequently, and the mistakes can be highly costly. Better to stop problems upstream in my opinion by giving examiners more time.
Patent quality is related but different. I agree that patent quality is awful, but there's only so much an examiner can do to influence that. Attorneys have basically gamed the system to write vague legalese that's patentable but basically useless. And to paraphase a supervisor I knew at the USPTO, "Just because it's stupid doesn't mean that it's not patentable". I can't reject them if it meets the legal standards but is stupid.
Anyhow, I think there might be a random sort feature that can do what you want in the USPTO's public search (no time to check, though): https://www.uspto.gov/patents/search/patent-public-search
I don't have the energy to diffuse false analogies. We're not talking about a hospital.
I don't think this is a good strategy. These folks tend to have a poor understanding of patent law in my experience, and you need to understand the basics to do this right. (You fortunately don't need to know too much law to handle 90% of cases.) And these folks probably aren't very effective at patent searching even if they understand the legal parts. I think most people overestimate their own search abilities. I certainly did. Examining patents didn't level me up as much as humble me in that regard.
> More deeply, the very idea of a "patent examiner" has never made sense to me. It require being expert in all things, which is impossible. [...]
As saratogacx pointed out, at the USPTO, the vast majority of examiners have a specific technology they are assigned to. While the matching of examiners to their "art unit" is often pretty bad (I could go on a rant...), the situation is not as bad as you described. There are some generalist examiners, but as I understand it, they are in (basically) QA roles and don't need to know the technologies as much. Unfortunately, USPTO upper management seems to want to make examiners into generalists, which I doubt will work out as they want.
I agree that periodic rotations of industry folks into patent examiner positions is a great idea. It would help the patent system and give the industry folks some appreciation for what examiners do.
If you're worried about lack of expertise, you should be more worried about the courts. Judges and juries almost never have a background in the technology of the case they are working on.