Patently-O Bits and Bytes by Juvan Bonni

Recent Headlines in the IP World:

Commentary and Journal Articles:

New Job Postings on Patently-O:

55 thoughts on “Patently-O Bits and Bytes by Juvan Bonni

  1. 7

    OT – but rather surprised (not) that our regular “Hyatt watchers” did not mention this case:

    link to (precedential)

    USPTO cannot overcome the American Rule and charge ALL expenses to Hyatt.

    Before MOORE, Chief Judge, REYNA and HUGHES, Circuit

    HUGHES, Circuit Judge

      1. 6.1.1

        So utility patent grants of domestic origin have not even doubled over the 20 years of Jason’s graph.


          In fact about 2/3 of the growth was from foreign origin patent applications as I said.


            And take that even with a grain of salt, as I have often chimed in that Big Corp (who are really multi-nationals, and have NO allegiance to any one Sovereign) should also be segregated from that “US origin” number.

            Heaven knows that several of my big clients are “US in name,” but that the innovation is NOT US based.

            (it is most likely even worse than you have been stating)


            There are roughly 7.7 billion people in the world, but only 0.333 billion in the U.S.A. The total world economy is worth ~$80.9 trillion (nominal USD), but the U.S. economy accounts for only ~$21.4 trillion. There are, in other words, a lot more non-U.S. people and a lot more non-U.S. economic activity in this world than there are U.S. people and U.S. economic activity. Moreover, the U.S. share of total population and global economic activity declines each year.

            It is only to be expected, then, that the U.S.-origin share of total U.S. patents should decline each year. That is not an indication that anything is wrong. That is simply the inevitable result of once-smaller economies catching up to our own.

            If—for some reason—you think it vitally important that more inventions should originate in the U.S., then the solution is to amend the immigration laws to make it more attractive for folks to move to the U.S. from wherever they presently work (see, e.g., Matt Yglesias’ new book “One Billion Americans”). On the other hand, foreign made inventions are beneficial to Americans, just as are domestic inventions, so maybe we just should not get too hot and bothered by these numbers.


              Greg, the issue you raise is not one I raised.

              The issue is that since the percentage went up so much that it should have been factored out of the analysis and that since foreign utility patent application now out number US origin application that this should be discussed as part of the data.

              It is probably good that more foreign origin utility applications are being filed in the US. But it is not indicative of the health of the US innovation or the US patent system.

              And as I said the comments with the graph were wrong in that it appears that US origin utility patents grew at a faster rate from 1980-2000 than they did from 2000-2020.

              So—-my points are about how misleading it is to include the foreign origin utility applications and Jason had the whole table there and did this intentionally.


                So this post says that on the one hand, one interlocuter is addressing arguments not made. On the other hand, it insists wrongdoing on the part of the original author for… an argument not made.

                I genuinely don’t understand how a functional adult could produce this post.

                1. insists wrongdoing on the part of the original author for… an argument not made.

                  That is NOT an accurate portrayal of the concerns being addressed.

                2. “I genuinely don’t understand how a functional adult could produce this post.”

                  Back at you Ben.

                  Not going to waste my time explaining myself again. This is your disruptive role where we go to zero knowledge and you pretend not to understand. My point is clear and was proven by Greg and others misunderstanding the chart.

                3. You’ve never once explained how the professor was asserting the health of the patent system with his graph.

                  It’s not in the original post and it’s not in his subsequent explanation.

                  It’s only in your head.

                  You spent this thread decrying others for not addressing the merits, but you won’t address the issue of the evidence for your assertion. You should be ashamed of yourself.


              Greg’s comment only shows that Greg can report numbers but lacks the ability to cogitate as to the importance of those numbers.

              Further, he is simply rather dismissive of any negative trend towards US innovation (as is his want that exhibits the Big Pharma One World Order lack of allegiance to the US Sovereign).

      2. 6.1.2

        It looks like if you consider U.S. grants that the grants for U.S. origin utility patents grew at a faster pace from 1980 to 2000 than they did from 2000 to 2020.

        I think there is no question but that given that foreign origin applications rose in percentage from 45 to 55% and that most of the looks at that graph are from foreign origin utility application that we can conclude that there are serious academic issues being raised by Jason’s post.

        Certainly, Lemley putting up this same graph and claiming it indicates that U.S. inventions and the patent system are healthy is a very clear academic ethics violation. Note that Jason doesn’t even care about this.

        Really outrageous that the academic community has turned into a marketing arm of the SV companies to weaken the patent system.

  2. 5

    Meh, on topic, I wrote a legal paper on this years and years ago. Decades. Machine learning, etc. etc. the patent and copyright value of the ‘weights’ – if you know AI – it’s all about the learned solution – i.e., the neural network ‘weights’ and to be sure the architecture and the training paradigm. Long story short? Copyrights and patents – works of authorship – and inventors respectivly – are human beings – not AI. Furthermore, if AI, were somewhere far far in the future to be deemed to have such rights – the 13th Amendment would probably come into play, see the star trek episode on Data. For now, it’s all just BS and I note – that the only ‘headline’ I got out the published, IMHO passible legal analysis, was something like “someday AI – i.e., machine learning, might be consider a work of authorship and have property rights.” Never mind, at that point, our SCOTUS would probably rule against Data in a Dred Scott like decision, and enslave the AI – the property rights belong to the owners of the AI. <— That's quasi tongue in cheek for you haters. We're a long long way from the 'singularity'.

    1. 5.1

      The way I see it, ownership is not the issue that is teasing everybody. The issue is who (or what) to name as “inventor”.

      The UK patent statute defines “inventor” as “actual deviser” of the claimed subject matter. Normally, the owner of that proprty is one legal person, the deviser another. How to address that normallity when the deviser is going to be an AI. Who shall the owner name to the Patent Office as “inventor”?

      1. 5.1.1

        True point: inventor/devisor and owner ARE separate entities.

        As noted, the US Sovereign, with our foundation in Lockeian inchoate rights, does NOT provide for a path of juristic person as owner to step into that critical originating point of being the inventor. See Stanford v. Roche. link to

        This is a critical point to understand.

      2. 5.1.2

        Reply caught in the George Carlin filter….

        (hint: devisor/inventor and owner ARE separate entities, US Sovereign fundamental basis tied to Lockean nature, and other items that I have already provided to you in conversations past)

    2. 5.2

      iwasthere: I studied A.I. at one of the top universities as a Ph.D. student back in the 1980’s.

      Basically you have it right. The number of people that have said we are just a few years from the “singularity” over the last 40 years is enormous. And Elon Musk is another physicist that thinks he understands computation. He doesn’t. The proposition that we are in a simulation is also almost certainly wrong for computational reasons that are beyond what I want to explain. (You hear ridiculous things like access to memory is constant, which it is not. The time used to access the memory is constant because the largest value is selected.)

      I probably agree with you that right now that the AI hasn’t invented anything but the inventor is still really the person that built and trained the A.I.

      I do take issue with the “singularity” term. This is another example of a charlatan taking a difficult concept that someone else thought of and figuring out a marketing label for it and then claiming he thought up the concept or idea. MacCarthy thought of the “singularity” back in the 1960’s at MIT.

      1. 5.2.1

        For all of the paper writing of decades ago, and Pile it Higher and Deeper studies of the ’80s, my experience is much more imminent, as my wheel house includes prosecution of some of TODAY’s top notch AI applications.

        but the inventor is still really the person that built and trained the A.I.

        This is objectively false.

        I do take issue with the “singularity” term. This is another example of a charlatan…

        My only quibble here is that focusing on Singularity is falling into the trap of ‘Oh Shiny.’

        It is entirely unnecessary, as the aspect of actually achieving a state of The Singularity has nothing to do with the collision — today — of a non-human inventor and patent law.


        Separate out — please — different legal issues.
        Focus on a first issue first.

        Do NOT conflate different legal issues and then turn around and try to not address issues at our door steps NOW.


            and, by the by, your reply does not address any of my counter points presented.

            Your statement of inventorship being the person who devised the AI system remains objectively false – from a purely legal definitional point.

            ‘The Singularity’ still remains a trap for the unwary, and has nothing to do with the current legal points on today’s intersection of AI and patent law.


          anon, “singularity” is, of course, related to whether the AI is the real inventor or the programmer is.

          Your arguments always take this weird structural form where they lack evidence and merely plug in words into something like a template.

          It is obvious that “singularity” is strongly related to whether the AI is the inventor. As singularity goes to the competence level of the AI and singularity assumes that once a competence level of human intelligence or greater is reached that singularity will occur.

          Whether an AI program is the inventor goes to the level of competence of the AI program. And independence from the programming and training.

          Anyway….I could flesh this out more but really this is an obvious point.


            is, of course, related to whether the AI is the real inventor or the programmer is.

            It may well be “related.”

            It certainly is NOT the driver. You are mis-applying logic in thinking that the relation requires that the level of The Singularity MUST BE achieved prior to the level of reaching the legal definition of inventor.

            You compound this by attempting to elevate humans — which clearly do not meet the legal definition — as the necessary default for the occurrence of an objective inventive event. Here, YOU provide NO evidence, and it is your own posts that reflect that which you would accuse me of: weird structural form lacking evidence with merely plug in words something like a template.

            And as to that accusation of “Your arguments always take this weird structural form where they lack evidence and merely plug in words into something like a template.

            LOL – that is a most odd reply. Maybe you should spend just a moment to THINK about the words and reply on the merits, instead of retreating to this “I need evidence” position (which is a position that you should recognize as one that Greg DeLassus pulls on you).

            Whether an AI program is the inventor goes to the level of competence of the AI program. And independence from the programming and training.

            Sure. I am not disagreeing with that – and never have. At the same time — and this is something that you STLL have not addressed — the legal meaning of inventor IS satisfied by a non-human actor, and related humans to the inventive event CANNOT satisfy the legal definition of inventor to that inventive event.

            There is NO need to ‘flesh this out more,’ other than for you to step through that very obvious logical path of the legal definitions combined with the provided facts.

            You have yet to do that.

    3. 5.3

      I think that it is a mistake to imagine that this issue will not matter until we reach the “singularity.” It is possible to imagine a machine with far less intelligence than a human, but with a more-than-human genius in one particular field (e.g., chemistry or enginee thrust). In that far-from-singularity world, inventorship of non-humans can still be a live issue.

  3. 4

    Off topic but maybe not.

    Owing to the benefits of specialization and division of labor, legislatures delegate responsibility for implementing their policy initiatives to various departments and agencies staffed by career bureaucrats, who secure their positions through civil service appointment rather than by democratic election. The early public choice literature on bureaucracy, launched by William Niskanen, assumed that these agencies would use the information and expertise they gained in administering specific legislative programs to extract the largest budget possible from relatively uninformed, inexpert legislators. Budget maximization was assumed to be the bureaucracy’s goal because more agency funding translates into broader administrative discretion, more opportunities for promotion, and greater prestige for the agency’s bureaucrats.

    As readers know, I’m a big fan of public choice theory to explain the reason why – such things like the AIA privy court was supported and advocated for by PTO. It expands and aggrandises agency power, just for example, how many new PTAB ‘judges’ were hired.

    Anyways, maybe Dennis could make a topic link, as to how public choice theory is being applied by our regulatory body – the PTO. See, link to

    1. 4.2

      On the basis of budget maximization, shouldn’t the USPTO be against IPRs? The income from patents is vastly greater than IPRs, and IPRs further undermine patents.

      1. 4.2.1

        IF there had been some mechanism to tie the hind process to the front process, perhaps.

        As is, the Office “makes money” off of both ends. By the way, if you pay attention to some of the people who speak out for small inventors, they have been quite vocal on the point that the front end and back end operating this way presents a type of shell game and false promise of what it means to obtain a granted patent.

        And they are correct. The AIA, ties in with the Oil States decision most definitely makes a mockery of what it means to have a patent grant in the first place.

        This point is also noted in the article that Night Writer has linked to several times now (Dowd/Michel).

  4. 3

    When I noticed that the DABUS application had been filed in so many jurisdictions, I started to wonder if the whole thing was merely an attempt at a test case to establish the right of an AI to be listed as an inventor. I was glad to read Susan Decker’s reporting on this point, which establishes that the answer is “yes.”

    Fair enough, of course. That is how one gets changes in the law—especially in common-law jurisdictions—by creating a situation in which courts of appeal are forced to consider the issue and decide. I agree with the consensus opinion that U.S. law does not presently permit AI to be listed as an inventor, but it is good what Prof. Abbot is trying to achieve by raising this issue. Changes to the law will be necessary, at least over the short term.

    I do wonder, however, if over the long term the advent of AI will simply make a patent system unnecessary. Right now, the logic of the patent system is that investors will not invest the necessary R&D capital unless the law gives them a way to capture the gains that flow from discoveries that result. In a world, however, in which the price of making such discoveries drops to (essentially) the price of a computer and the electricity necessary to run it, will the patent system even be necessary to incentivize such discoveries. That is to say, the cheaper it becomes to make discoveries, the less one needs a patent to incentivize it.

    On the flip side, meanwhile, if one has AI out there making discoveries and publishing them to the web, what is the scope even to get a patent. The world of prior art will soon be so crowded with (accurate and enabled) descriptions of as-yet-untested technology that it will not be possible to get a patent.

    We would need to introduce an actual-reduction-to-practice rule for enablement in order for there not to be a prior-art obstacle to every newly tested and validated invention. Is there going to be a public appetite for the enactment of such a rule in a world where (see above) new discoveries are being made and publicized even without patent incentives?

    1. 3.1

      It is a good point that perhaps AI will generate without filing patent applications endless prior art and publish it to a public server.

    2. 3.2

      “will soon be so crowded with (accurate and enabled) descriptions of as-yet-untested technology that it will not be possible to get a patent.”

      In the nearer term it’ll just make it impossible to get a patent on simple combinations of concepts. That’d be a good thing. We already shouldn’t be incentivizing the disclosure of things like [name_of_article] + [fractals].



          Ben is a paid blogger for sure anon.

          You are –frankly–not the brightest bulb anon. You just don’t see the pattern of argumentation that indicates a paid blogger. And the repetition of the same policy arguments almost verbatim from people like Ben. The classic deny, disrupt, and repeat disinformation that they are paid to follow. You know he likely has a script for most of what he posts.

          My guess is that he works for the USPTO and picks up extra money as a paid blogger, which, by the way, is cause for termination at the USPTO. You can’t work at jobs with conflicts with your work. He could be a carpenter too but he can’t push patent policy arguments. My guess is that is why Ben gets so worked up when I point this out.


            It actually amazes me anon that you can easily see how Greg’s inhouse position shapes his positions and posts. It is pretty clear that Greg is careful to represent his company in these posts.

            Ben’s arguments are so consistently the same ones that are presented by the anti-patent judicial activists with ZERO deviation that it is impossible that he isn’t a paid blogger. He has a manager that checks his work and provides him with scripts.


              Greg’s inhouse position shapes his positions and posts. It is pretty clear that Greg is careful to represent his company in these posts.

              Oh, for pity’s sake, are we going to trot out this pitiful line of projection again? I doubt that the rest of the world cares about this nonsense, but I invite those few who are interested to read the link below to see how NW embarrassed himself last time he dribbled out this drivel.

              link to


            LOL – you want to try to denigrate me just because I am not head over heals for your ‘paid-blogger’ position?

            You are DEEP into the weeds on that one, Night Writer.

            First, I never said Ben was NOT a paid blogger. I simply pointed out that Ben is most likely a patent examiner. As you yourself further illustrate, Ben MAY BE BOTH.

            I am far less concerned about Ben’s posts BECAUSE of their lack of critical thinking. They most often will implode all on their own.

            Greg “I Use My Real Name” DeLassus is a case that more rises ire.

            He IS an attorney, and he SHOULD know better, but he constantly ploys his Big Pharma propaganda and actively advocates with positions that align with anti-software based innovation. He then has the gall to hide behind ‘politeness’ (even as he insults others), and posts things like “you can’t shame the shameless” when HE is the shameless.

            He is the type that give patent attorneys a bad name.
            He is the type that fuels anti-innovation group think.

            Ben is a piker, a nobody, and is easily (if not amusingly) dispatched.

            You want to get worked up over whether Ben is a paid blogger? Go right ahead. I need not choose to join you on that point.

    3. 3.3

      Some decent points, but it is not a small quibble re:

      That is how one gets changes in the law—especially in common-law jurisdictions

      to emphasize that in the US Sovereign only one branch was authorized to write patent law.

    1. 2.3

      This would have been a worthy addition to the Bits & Bytes publication list. Thanks for sharing it, NW.

  5. 1

    I recommend the Decker article (especially for those with sensitive feelings who find my more direct style to be too much).

    1. 1.1

      I am fairly certain that no one on here would accuse you have having a direct style, so no worries.

      1. 1.1.1

        Direct – as in your face.

        (try to keep up)

        Other than that – care to engage on the substantive matters?

      2. 1.1.2

        And your silence on the substantive matters speaks more loudly than your braying with off-point attempted smarm.

Comments are closed.