[HN Gopher] IBM Patented Euler's 200 Year Old Math Technique for...
       ___________________________________________________________________
        
       IBM Patented Euler's 200 Year Old Math Technique for 'AI
       Interpretability'
        
       Author : busymom0
       Score  : 106 points
       Date   : 2025-11-13 18:41 UTC (4 hours ago)
        
 (HTM) web link (leetarxiv.substack.com)
 (TXT) w3m dump (leetarxiv.substack.com)
        
       | mv4 wrote:
       | I remember reading that IBM holds a staggering number of patents
       | - over 150,000 in the US alone.
        
       | CountHackulus wrote:
       | When I worked there many decades ago, I had pressure and saw
       | people get rewards for patenting everything you could. Seems like
       | things haven't changed.
        
       | stuffn wrote:
       | This is the primary reason IP patents should be thrown out.
       | They're far too easy to get, exist primarily for the purpose of
       | rent seeking, and generally provide nothing useful to the public
       | knowledge.
        
       | anthk wrote:
       | Good in case of Europe has a shortage on toilet paper.
       | 
       | Also, this might be tangentially related to Fractan. Any
       | Mathematician there? if there's some relation, the patent might
       | be void/null.
       | 
       | https://en.wikipedia.org/wiki/FRACTRAN
        
         | noir_lord wrote:
         | While we might get to laugh at those silly people with their
         | software patents (and describing them as toilet paper is quite
         | funny) it still has real world effects for us.
         | 
         | For many years I had to install a linux distro and then mess
         | around with freetype because there was a valid patent on
         | subpixel hinting that was valid in the US and irrelevant in
         | Europe but because the distro's I used at the time where US
         | based I got caught in the cross fire.
         | 
         | Given the huge imbalance between US tech and Europe Tech (and
         | rest of the world really) in software they sneeze, we get flu.
         | 
         | That only stopped been a problem when the patents expired[1].
         | 
         | [1] https://freetype.org/patents.html
        
           | anthk wrote:
           | That's why you should always use either European or Canadian
           | mirrors. On FreeType, maybe Clear Type related; but I
           | remember some slight hinting methods _not_ related to
           | cleartype which looked far superior on flat LCD screens IMHO.
           | 
           | Another one was Lame and MP3. And I remember the codecs issue
           | with the Penguin Liberation Front for Mandrake, LIVNA for
           | another distro and Debian Multimedia, among others.
        
       | pkaye wrote:
       | Isn't this just a patent application? Also what are the actual
       | patent claims? I was not able to find a link to the application.
        
         | kristjansson wrote:
         | https://patents.google.com/patent/US20230401438A1
        
       | ipfreely wrote:
       | Misleading. This has not been patented. The author links to a
       | patent publication, which happens to the vast majority of patent
       | filings in the U.S. There was a rejection of this patent
       | application mailed by the USPTO in August 2025.
        
         | analog31 wrote:
         | Rejection is when the fun begins. Note that I'm not a lawyer,
         | but I do have 20+ patents, none in the software realm.
         | Rejection means you get to revise your application and re-
         | submit with a new deadline and new fees. Your chances of
         | getting it accepted on the second or third try can vary.
         | 
         | Applications are seldom withdrawn or ultimately rejected
         | altogether, but may be granted with severely narrowed claims.
         | Also, you can revise the claims without changing the text,
         | within limits, which results in granting of a patent with a
         | very broad text followed by narrow claims
        
       | rdtsc wrote:
       | The patent system for software is garbage, but it also doesn't
       | seem to jive with the title. They don't claim to invent continued
       | fractions, what they patent is applying them to some domain.
       | 
       | https://patents.justia.com/patent/20230401438#history
       | 
       | > A method, a neural network, and a computer program product are
       | provided that provide training of neural networks with continued
       | fractions architectures.
       | 
       | It's a bit like the propelling device
       | https://patents.google.com/patent/US1331952A/en -- it doesn't
       | invent or patent shoes, or invent or patent springs, but it
       | patents attaching springs to shoes.
       | 
       | It would seem using continued fractions with elliptic curves what
       | the author wants to do, wouldn't be covered.
       | 
       | However, I still think it can still be challenged if someone can
       | show that continued fractions have been used in with NN before.
       | Or even better, maybe pytorch or other open source projects can
       | explicitly reject crap that's patented. If you put your shit in a
       | junk patent, take it out of the project and enjoy yourself, don't
       | spread it around. So, if the authors of the patent are the ones
       | pushing for the inclusion, then someone should challenge that and
       | have it removed unless the patent is withdrawn.
        
         | ajb wrote:
         | The problem here is whether _just_ applying some well studied
         | technique to a new area like AI is really inventive enough to
         | be patentable. There were loads of patents which basically
         | added  "in a computer" to some existing technique. Just because
         | NN's are novel doesn't mean people should be able to get an
         | exclusionary property right over the use of well known
         | techniques in NN's because they put in half a day's work to be
         | the first to apply that technique to NN's. This isn't the
         | Guinness book of records.
        
       | IceHegel wrote:
       | Whether or not the patent was actually granted in this case, I
       | have not been able to think of a compelling reason to have
       | patents for software. In fact, I think most intellectual property
       | laws need to be seriously rethought.
       | 
       | If the objective is to maximize investment by protecting
       | successful results, I don't think our system is doing a very good
       | job.
        
       | keeda wrote:
       | Before this devolves into the morass of outraged comments that
       | usually comprises any discussion of patents, here is the golden
       | rule I coined for such threads:
       | 
       | RTFC: Read the Fucking Claims
       | 
       | The claims are the only part of the patent that really matter
       | because those are the only enforceable language. Plus, this is a
       | patent application so the claims have not even been examined yet.
       | 
       | Without commenting on the merits of this patent itself, remember
       | that new applications of existing techniques are still novel, and
       | hence patentable. In fact, if you think about it, almost all
       | inventions and innovations are just applying novel combinations
       | of well-known techniques to new use-cases. (Anything that doesn't
       | fit that definition and introduces genuinely new methods is
       | usually in the "groundbreaking" category.)
       | 
       | So I'd guess "applying old technique to new problem" is probably
       | the case here. I'm no subject matter expert, and there may be
       | prior art that invalidates this, or it may not meet the non-
       | obviousness bar... but when the "200 year old math" came about
       | neural networks were not really a thing.
        
       | linuxhansl wrote:
       | Software patents are mostly garbage and unneeded. And I say that
       | having some in my name (I actually tried to get my name off them,
       | but our lawyers said it's not possible).
       | 
       | Show me one useful software patent that (a) is not "obvious to
       | one skilled in the art", and (b) benefits society by being
       | granted a monopoly. Just one!
       | 
       | Software rarely requires expensive research that would be worth
       | protecting. Rather than enabling a fair market, this takes
       | fairness out of the market.
       | 
       | Software patents are like getting a patent on "Murder story with
       | final revelation of who did it." Maybe add one or two features,
       | like a "detective with hat", etc. In one fell swoop you would be
       | able to own most murder mysteries.
       | 
       | Software (like books, stories, art, etc) is better handled by
       | Copyright law. May the one who actually _has a better product_
       | win!
       | 
       | Sorry for the rant.
        
         | JohnFen wrote:
         | I agree with your rant. I have a similar one. I personally
         | think that software should mostly just not be patentable at
         | all, since (among others) these things are not patentable
         | according to US patent law: scientific discoveries,
         | mathematical methods, aesthetic creations, and rules or methods
         | for performing mental acts.
        
           | bandofthehawk wrote:
           | In addition, software is also copyrightable, which makes much
           | more sense than patents for protecting unauthorized use. IMO,
           | patents for software should be mostly eliminated, and even
           | copyright terms should be much shorter.
        
         | fsckboy wrote:
         | > _Show me one useful software patent that ... (b) benefits
         | society by being granted a monopoly. Just one!_
         | 
         | that's a bullshit criterion, and nothing about what you said
         | applies to software in particular.
         | 
         | it's generally agreed that monopolies are bad for society, but
         | patents only grant a temporary monopoly to reward innovation,
         | in exchange for which society gets disclosure (at the time
         | patent systems were promulgated, many ideas died with their
         | inventors as trade secrets) and encourages more R&D by creating
         | a system for payoff.
        
           | bandofthehawk wrote:
           | The only "system for payoff" I've seen with software patents
           | is patent trolls. Are there cases of software inventors being
           | rewarded for their software more fairly because they had a
           | patent?
        
             | oriolid wrote:
             | I think every every company I've worked at that had R&D had
             | some kind of reward system for patents. Yes, most of the
             | software patents were nonsense but those who have their
             | names on it still did get paid.
        
               | waterhouse wrote:
               | Guessing those rewards are in the hundreds of dollars,
               | probably a fraction of the engineer salary that went into
               | the technical work.
        
           | Dylan16807 wrote:
           | It's not bullshit. You said it yourself, the benefits are
           | supposed to be disclosure and encouraging R&D.
           | 
           | But disclosure is rarely an issue with software, and patents
           | are bad at properly disclosing software details in the first
           | place.
           | 
           | And in software there's already a huge motivation to do R&D,
           | while patents are more likely to block useful work than in
           | most fields. Even if I think of highly optimization-motivated
           | fields like video encoding, patents slow down innovation more
           | than they accelerate it.
           | 
           | So can you name some software patents where those motivating
           | factors actually worked? It's a fair question.
        
         | yyurgenson wrote:
         | RSA
        
           | Dylan16807 wrote:
           | What was the societal benefit of putting 20 years of monopoly
           | on that algorithm? I don't think potential profit was a big
           | motivator in that research work.
           | 
           | And that patent got invalidated in most of the world anyway.
        
             | Aloisius wrote:
             | The societal benefit was having RSA publicly described. MIT
             | isn't a charity.
        
               | Dylan16807 wrote:
               | You think they were just going to sit on that and do
               | nothing if they couldn't get a patent? Or that they'd
               | turn it into a product without the underlying math being
               | revealed right away? I don't believe either of those for
               | a second.
        
               | Aloisius wrote:
               | There is a rather long history of the workings of
               | cryptography products kept secret, so yes, it is entirely
               | possible that the underlying math would have been kept as
               | a trade secret.
               | 
               | It is also possible that it would have never been created
               | in the first place because resources were allocated to
               | other patentable inventions.
               | 
               | Of course, in the case of RSA, a similar algorithm _was_
               | developed separately by the British government and kept
               | secret for 24 years.
        
         | josefritzishere wrote:
         | Gold star comment.
        
         | mring33621 wrote:
         | I agree that software patents are generally garbage and hurt
         | innovation.
         | 
         | Unfortunately, there are probably many people here on HN that
         | make a living off software patents.
        
         | smj-edison wrote:
         | What would you think about still having software patents, but
         | having them expire in 2-5 years? I feel like the concept of a
         | patent is still a good one, but the time it takes to go from
         | zero to a product is drastically lower with computer
         | programming. The patent length should reflect that.
        
           | OgAstorga wrote:
           | How would you feel about patenting language? I.e. If you
           | speak with certain words or certain patterns then you have to
           | pay a royalty (only for 2 to 5 years).
        
             | smj-edison wrote:
             | First off, I think that's a false equivalency, as patenting
             | is about ideas (in a platonic sense), not about instances
             | of ideas (which is what copyrighting is).
             | 
             | Secondly, we already have that in limited forms with
             | trademarks and copyrights.
             | 
             | Thirdly, I think the concept of intellectual property is
             | one of the most brilliant social innovations in the past
             | 500 years, as it aligns incentives to innovate (why would I
             | innovate if someone will just steal my work?).
        
         | greg_w wrote:
         | One cannot (in the US) get a patent for software itself. This
         | was settled a while ago. There needs to be more in the claims.
         | In fact, the patent discussed here does not claim continued
         | fractions and nobody would be in danger using them even if the
         | patent issued as is (which is not certain, because the patent
         | claims rather trivial modification of a classic neural network
         | architecture, which should be brought up by the examiner as
         | obvious).
         | 
         | Patents are propelling the society when they work as intended.
         | They made XIX century and at least good chunk of XX century.
         | Without patents, people fall back to copying each other,
         | because it is much easier to copy than to innovate.
        
       | zkmon wrote:
       | I don't think they are patenting any classical math as the blog
       | claims. It's an architecture using CF that is patented, I
       | believe.
       | 
       | Ofcourse, patents are trash.
        
       | shmerl wrote:
       | Reminds this: https://theonion.com/microsoft-patents-ones-
       | zeroes-181956466...
        
       ___________________________________________________________________
       (page generated 2025-11-13 23:00 UTC)