Supreme Court Refuses To Hear Case Over AI’s Right To A Patent; AI Inventions Remain Unpatentable
from the next-time-try-the-AI-supreme-court dept
Phew.
We’ve written a bunch about Stephen Thaler’s quixotic and dangerous quest to allow AI created works and inventions to receive copyrights and patents. It’s repeatedly failed to convince people, especially US judges, that Congress intended anyone other than human beings as the creators and inventors to receive such monopolies.
Thankfully, Thaler’s loss streak continues. Despite the surprising amicus brief from Larry Lessig, the Supreme Court has refused to take Thaler’s case and the appeals court ruling on the unpatentability of AI generated inventions remains standing.
The justices turned away Thaler’s appeal of a lower court’s ruling that patents can be issued only to human inventors and that his AI system could not be considered the legal creator of two inventions that he has said it generated.
Of course, I’m sure this won’t be the last we hear from Thaler. The UK Supreme Court just heard his similar case there, after the UK also refused to give him patents on inventions created by the AI system he created, called DABUS.
And, of course, he’s still arguing the copyright side of things as well. Given how often his attempts keep showing up, I get the feeling he’s not going to just accept defeat and move on.
Still, this is a huge win for innovation, especially as AI has become much more common and accessible to people. Let the AI invent stuff for the betterment of humankind, and not to get a monopoly to lord over us.
Filed Under: ai patents, dabus, patents, stephen thaler, supreme court
Comments on “Supreme Court Refuses To Hear Case Over AI’s Right To A Patent; AI Inventions Remain Unpatentable”
Wow, what a damper on World IP Day. My robot lawyer is totes upset.
This comment has been flagged by the community. Click here to show it.
Unconvincing arguments
I read the copyright office’s claims, and I find the following reasoning confusing:
Entering a text prompt into a neural network:
not copyrightable
Clicking on a GUI into a symbolic computer program:
copyrightable
Connecting a brain computer interface to a neural network:
??? presumably not copyrightable ???
The copyright uses some analogy to agency theory, to describe why the AI is the creator of the work, but somehow I don’t see why this is different than the two other examples.
Re: Oh, I see
Could it be…. The USPTO is using Chat GPT to reject patents?
Naaaaaa. Too easy.
Re:
Entering a text prompt into a neural network:
User enters a value they create
Clicking on a GUI into a symbolic computer program:
User triggers a specific action.
I’m guessing you don’t work with computers.
Re:
Considering what you wanted to do with copyright law, it’s very, very safe to assume you don’t know jack shit about copyright law.
Or would you like to get schooled AGAIN in a place where there are actual LAWYERS to tell you how fucking wrong you are?
(To those of you who don’t know, this is likely THE Benjamin Jay Barber, revenge pornographer and laughingstock of a certain forum.)
Re:
Your general complete lack of comprehension is a “you” problem.
Re: wrong place
The problem with discussing the copyright office’s view is that we are dealing with a question of patentability. Copyright office offers little insight into that.
(also, you may want to change your login name, it very much resembles a guy who was convicted in a notable revenge porn case)
So how much do you have to change something an AI produces to make it copyrightable or patentable?
Does going through the product of chatgpt and removing all instances of it being the product of chatgpt count? Inserting a step in an ai designed process?
Or can you just not tell anyone that “your” work was produced by an ai?
Re:
AI is not a legally recognized entity, it’s just a piece of software like any other as far as the law is concerned. So whether you use MS Word or GPT-4, you can claim the invention as your own. IANAL, YMMV, LOLWTFBBQ.
AI generated patents would either give license fees to something that does not understand money, or more likely allow people like Stephen Thaler to tax those people who actually make things.
Re:
So in other words, AI generated patents would be like steroids to Patent Trolls.
(Do I have that right?)
About the use of AI in drug research
At some point, the level of human vs. computer involvement/effort is going to have to be resolved. Given the apparent interest in the use of various forms of AI in drug discovery, there will likely be a future situation where it can be argued that the “discovery” was made by an AI – and that therefore the drug would not be patentable (per current law/doctrine). Since it’s patentability that leads to substantial profit, the pharmaceutical companies would have to either (a) abandon the use of AI technologies in drug discovery, or (b) lobby to get the law/doctrine changed. Any bets on which course of action is more likely? (I’d love to see a pharma patent challenged on the basis of AI involvement in the “invention.”)
Re:
(c) Anyone can run an AI looking for drugs, and drug companies are SOL with patents, but can make money the old-fashioned way – production.
Most data, research, and all, wasn’t generated by the drug companies anyway. They do some, but suck up tons for free (and then lock it up).
No bets on what actually happens, this planet is pretty fucked up stupid.
RIP ALL the progress...
Well there goes all future innovation, if AI’s can’t get patents then what possible motivation do they have to make anything?
Re:
Pure, unbridled spite.
Nice post by the way. I loved the article very much. It was so informative and interesting
http://brendalbranson.com/