The Corporation as an Inventive Artificial Intelligence

33 thoughts on “The Corporation as an Inventive Artificial Intelligence”

1. 12

   Worth noting a post by David on another thread that speaks more to the topic here:

   link to patentlyo.com
2. 11

   Might be curious as to Prof. Abbott’s view on the supposed reversal by the higher court in Australia (saw a blurb, but have not seen the actual decision).
   1. 11.1

      Having now seen and after a quick read, I must say that the AU reversal was rather unsatisfying.

      It was basically a “no, we want this result” type of reasoning.

      Thaler, Abbott, what say you?
3. 10

   Dennis, how about the UK Patents Act, 1977, Section 39, declaring who, ab initio, is the owner of an invention made by a (human) employee of the corporate applicant for patent rights? For in that situation it is the corporate employer, and not the human employee, who is declared by the statute to be the owner of the invention.

   You write:

   “…corporate ownership of patent rights are derived rather than original—they stem from a transfer of property rights from human inventors who begins the chain of title.”

   but do you say that in the case of the UK what you wrote is an accurate statement of the law?

   Is Section 39 of the UK statute something to build on? It seems to be standing up well to scrutiny in the UK courts, so far at least.
   1. 10.1

      What the UK wrote — quite obviously — would NEED be copasetic with US law – including our setting of the authority grant in the Constitution (think the nature of the patent right).

      Think Lockian.

      We’ve been over this before, MaxDrei.
      1. 10.1.1

         Errrr? Are you saying that a statutory provision along the lines of Section 39 of the UK Patent Staute is irreconcilable with the patents clause of the Constitution? Is that it?

         At 2.3 Greg makes a point about authors and copyright. If that provision of copyright law accords with the Constitution why can’t patent law adopt the same scheme of ownership?
         1. 10.1.1.1

            See my reply to Greg’s musings – you (like he) appear to miss the critical point at the start of the chain.

            This is exactly how the Lockean nature comes into the picture.
            1. 10.1.1.1.1

               That’s a bit too abstruse for me, anon.You allude, I take it, to the view of property espoused by the “father of liberalism” the 17th century English philosopher, John Locke. So, are you asserting that the UK Patents Act of 1977, in its Section 39, is hopelessly incompatible with that gentleman’s view of property? Or are you asserting that ithe UK patent statute is unacceptable because it doesn’t “deal with” the “critical point” in Locke’s philosophy of liberalism?

               I’m also baffled by your assertion that copyright is different from patent law because copyright “arises out of a human endeavor”. Are you asserting that it is impossible for a copyright work ever to arise from the endeavours of a non-human author?
               1. 10.1.1.1.1.1

                  Wow – your comment finally was released, and I have to say that I am genuinely shocked that you consider the Lockean nature of patent property rights to be “a bit too abstruse.”

                  Further, I make NO such assertion as you attempted to make with “copyright different from patent law”

                  Quite in fact – I provided (prior to Greg providing the same case) the Simian Selfie case.

                  Were you purposefully attempting to obfuscate?
   2. 10.2

      Max, the issue here is not about any difference in Ownership of the inventions of normal corporation-employed inventors. It is about the personal recognition importance to such inventors of being named as inventors on patents. As those who have spend many years needing the assistance and cooperation of employed inventors for application preparation, prosecution and litigation can testify to.
      And mere copyright Registrations are not a valid analogy. The numerous personal disputed copyright lawsuits over music authorship show the importance of that to individuals there as well.
      1. 10.2.1

         Paul, I appreciate the comment. The UK 1977 patents statute copied the German scheme of “compensation” for employee inventors. This scheme was introduced by Adolf N*z* in the 1930’s. One might argue that it played no small part in German technological resurgence in the 1960’s and that this is why the UK copied it in the 1970’s. There are libraries of case law under the German Employee-Inventor Act and in the UK at least one case has gone as far as the Supreme Court. I guess that corporations baulk at the notion of paying proportionate “compensation” to employees who make patentable inventions.
         1. 10.2.1.1

            Max, one reason U.S. companies have rarely had employed inventor “inventor compensation,” and we have no laws requiring it, is that companies that have tried it, like LMSC, have found that it creates employee disputes and even lawsuits over who are entitled to it, and it is hard to calculate when patents are cross-licensed or are only a part of a product or contract, etc. Such laws are even argued to be a disincentive to locate R&D facilities in those countries that have them.
            But you have raised with this another relevant legal obstacle to the subject proposed ignoring of actual inventors determinations in employed inventor patents.
            1. 10.2.1.1.1

               Paul, I don’t understand your second paragraph. Sorry.

               As to your first para, this strikes me as a classic “tail wagging dog” argument. To let employee inventor law determine where one sets up a corporate research lab strikes me as self-harming. As far as I know, GE chose Munich to set up its world-wide medical technology research HQ, despite Germamy having such a mature employee-inventor-compensation jurisprudence. Was GE in error there? Or was GE far-sighted?
4. 9

   Although it is an interesting idea, and although likely Congress could play a helpful role in clarifying the law in this area, it seems to me that corporate AI is the wrong way to go. As it is, the human inventors are often short-changed. They usually or at least often do not share in the profits of their invention to any significant extent, and are at least sometimes improperly not given credit for their inventions (e.g., because they left the company before the patent was filed or had a fight with their boos).

   With AI, there is essentially always a human involved somewhere. The AI generally do not program or train themselves. Also, generally, the nature of the question posed to the AI or the parameters of the invention “invented” by the AI are generally determiend by humans. Let’s at least give credit where credit is due, and figure out which of those people, should be called the inventor of the invetnion that resulted from their interactions with, or develpment of, the AI. Let’s not create another tool for further cutting the inventor out of the invention.
   1. 9.1

      David Lewis – you are incorrect on any number of points.

      While your lament of “Let’s at least give credit where credit is due” SOUNDS good – you want to give credit of INVENTION where it is NOT due.

      Thus – you miss the entire point AT point.
5. 8

   >corporations…many accompanying civil rights

   Technically, it’s the owners/shareholders who have the Constitutional rights, not the corporation. See e.g., Citizens United.

   > legal fiction of personhood status

   And thus, the modern view that corporations are a nexus of contracts, not a “person.”
6. 7

   What you are really talking about is intelligence as an emergent property, and it’s kind of the same thing as, say, an ant colony being intelligent or the universe being intelligent, by virtue of what its parts do. It’s more of a philosophical question than a scientific one.

   Taking the term “artificial intelligence” colloquially, sure, I guess a corporation is artificial and I guess it exhibits some sort of emergent intelligence as a collective unit. But “artificial intelligence” as a term of art has a more particular meaning – namely, some form of computational intelligence – that would generally exclude your examples.
7. 6

   Interesting take on it Dennis. What if an AI sets up a corporation and sets up a profitable business?

   Weird but could happen soon. I think ultimately that we need a person (human) to be accountable for anything AI does.

   The best AI movie, by the way, is Ex Machina.
   1. 6.1

      Thanks! Totally agree.
8. 5

   It would also make discovery of what uncited prior art the inventor was aware of very difficult, and attribute knowledge all prior art known to anyone in the entire corporation.

   Why is that a particular problem? Does anybody ever get in trouble for not citing some item of prior art absent an injured party proving it?

   Does it really matter what an entire corporation knows or does not know relative to some reference that arises in litigation?

   The technicalities of that form of inequitable conduct are the real legal fiction.

   PS the salience of non-human actors in patent matters can only grow each year.

   Y’all know where I stand on that….
9. 4

   As an inventor and leader in a corporation – the current lack of clarity on the interpretation of patent law in the courts leaves people like myself growing increasingly weary of the uncertainty of the value of patents. Add to that the enormous cost to enforce patent rights (at least $1M if it ends up in litigation) and the asymmetric strength large entities have over smaller, the LAST thing I want to see is another branch of uncertainty created. The focus should be on making the patent system a reliable partner to inventors and corporations versus the circus it is today by removing the growing ambiguity.
   1. 4.1

      Bingo.
   2. 4.2

      >>The focus should be on making the patent system a reliable partner to inventors and corporations versus the circus it is today by removing the growing ambiguity.

      That is actually what the CAFC is supposed to do. But the anti-patent activist judges do the opposite to weaken patents.
10. 3

    “corporations have already been granted the legal fiction of personhood status and many accompanying civil rights.”

    Piling on to a laughably bad precedent is definitely a great way to advance the “rights” of computers.

    Can we stop using the silly term “AI” and just call a computer a computer? Or is that term part of the propaganda?
    1. 3.1

       “Or is that term part of the propaganda?”

       Most definitely not — and if you don’t understand the difference then you should probably refrain from commenting until you do.
11. 2

    Corporations as named inventors is a political non-starter. It would deprive corporate employee inventors of personal patent recognition and good resumes, and make it even easier for their managers to take the credit for their inventions. It would also make discovery of what uncited prior art the inventor was aware of very difficult, and attribute knowledge all prior art known to anyone in the entire corporation. Corporations already being granted the legal fiction of personhood status for unrestricted political payments to politicians has done enough damage already.
    1. 2.1

       Great minds, Paul … just two minutes apart.
       1. 2.1.1

          Paul’s post and your post are materially different. Quite in fact, Paul’s post is closer to some of my own anti-Citizen’s United posts.

          This is a most odd version of your “one-bucket.”
    2. 2.2

       Your point regarding a potential corporate inventor’s knowledge of prior art and consequent duty of disclosure is an interesting one; an area apparently not (yet) addressed in Dennis’ draft chapter.

       If corporations were to be permitted to be named as inventors, I’m not certain that the legal standard would necessarily be one that would “attribute knowledge [of] all prior art known to anyone in the entire corporation”. To use a neural network analogy for employees in a corporation, not all knowledge of individual employees anywhere in the corporation need rise to the level of “corporate knowledge”.

       The complication would then be to create some judiciable standard of corporate knowledge of prior art that in some way rises to the level of sufficiently conscious knowledge of the “corporate inventor” as a whole. For example, the current standard, wherein only individuals associated with the filing and prosecution of the subject patent application have a duty of candor and good faith, could possibly be transferred with suitable modification to the corporate inventor situation, thereby excluding individual knowledge of those persons in the corporation who are not directly affiliated with and wholely unaware of the patenting activity of the corporation for which they work.
       1. 2.2.1

          There may well be a more immediate (and far-reaching) impact as to what constitutes a Person Having Ordinary Skill In The Art (non-real-person squared).
    3. 2.3

       Corporations as named inventors is a political non-starter. It would deprive corporate employee inventors of personal patent recognition and good resumes…

       I do not think that this is true. It is already the case in copyright that the employer is the “author” of a “work for hire.” Does this arrangement deprive corporate employees of their relevant employment credentials and recognition?

       Were Helmut Krone & Julian Koenig denied their proper recognition for Volkswagen’s celebrated “Think Small” ad campaign? Did Shepard Fairey languish in obscurity, his “HOPE” campaign unacknowledged? Is it the case that no “Morning Edition” listener has ever heard of B.J. Leiderman?

       In other words, the problem that you propose does not actually happen in the copyright circumstance? Why is this more likely to be a problem in the patent context than in the copyright context?
       1. 2.3.1

          Greg, your view falls flat — quite literally — at the start.

          Copyright still arises out of an actual HUMAN endeavor. What happens after that may well be a different story – but you have to deal with the first instance, well, first.
12. 1

    Nice — reminds me of something that I stated early on.

Comments are closed.