Can AI Legally Be A Patent Inventor?
Вставка
- Опубліковано 28 вер 2024
- Is an algorithm a person? Can we make it one? Should we make it one?
Become a Patron! / everydai
Thank you to Ian, Becky, Jino Jang, Daniel Shiffman, Jason Roop, Mariano Munoz, and Kendou!, for being $5+/month Patrons!
Follow me on Twitter! / jordanbharrod
everydAI is a UA-cam channel focused on highlighting the ways we interact with artificial intelligence every day.
Sources (Factual):
Patenting with AI as the Inventor - www.bbc.com/ne...
Patent Application #1 for Food Container - artificialinven...
Patent Application #2 for Warning Light - artificialinven...
Law for Computer Scientists - lawforcomputer...
Chapter: Legal Personhood for AI? - lawforcomputer...
Sources (Opinion-Based):
Essay on Legal Personhood for AI from NC Law Review - scholarship.la...
Columbia Science and Technology Law Review - stlr.org/2018/1...
Human Indignity: From Legal AI Personhood to Selfish Memes - www.researchga...
![[UA] Eternal Fire проти NAVI | EPL Season 20 Malta](http://i.ytimg.com/vi/chvlsDygrZM/mqdefault.jpg)
In terms of intellectual property, if something is entirely created by something which isn’t human (e.g AI, and animals) then that creation should be in the public domain, an nobody should hold exclusive rights over it.
For example, if a painting was made entirely by a robot controlled by artificial intelligence, while the person who owns the robot and supplied it with canvas should be free to sell that painting, and collect the revenue from it, they cannot declare ownership over the image painted by the robot.
If a form of artificial intelligence were to create a design unassisted, that design would be in the public domain, and absolutely anyone should be able to use it. However, if a person assisted in creating that design, that person instead should hold exclusive rights over it instead.
I don’t think we should be giving exclusive rights to machines. Not yet at least.
I think this is my line of thinking too.
Though, the line between created by a human aided by AI and fully created by AI is probably a pretty difficult distinction to make in a lot of real life cases.
What does a "person assisting" entail. Pushing start, writing code, supplying materials, supplying financial support.... it's a bit blurry, and we have no consensus on this, legally, at the moment.
And, you bet the owners of the AI aren't going to willingly give such an invention in to the public domain in most cases. 'Cause that'd be giving up any potential profits as well.
kassemir, certainly, by a person assisting, what I meant was making imperative artistic contributions in which they would be granted exclusive rights over the work if it were a person they were working with, rather than AI.
For example, if an AI were to refine a design a person created, the person who created the design should hold exclusive rights over it.
If a person just written code for the AI, specific some standards of what the design should include, and pressed start, that design should be within the public domain. To argue that the person who created the AI created the design would be like arguing that one’s parents should hold an indefinite right to absolutely everything that individual owns.
kassemir, if we had AI which could create designs independently, what I would expect is that the company funding it would pay someone to modify the work to just a degree where it would be considered as its own work, and thus no longer under public domain, then register that derivative as their intellectual property. It’s an entirely valid way around it, in my opinion.
There is an AI robot that is an actual citizen of Saudi Arabia. With corporations in control, human rights are secondary to profitability, I expect to see more of this. An AI robot, as a citizen would need to pay taxes, would not need healthcare, food, breaks, only repairs and maintenance. In fact, as a citizen, an AI robot may be able to make purchases. Corporations may give a large percentage of the population the axe. Not only will corporations profit off of the work of the AI robot, they will own these Robots, and pocket earnings if they are considered tax paying citizens.
Of statistical significance: Nobody in the history of ever has Subscribed to a channel because they heard the word regurgitated faithfully and decided to Pavlov about it.
Agreed! There's no good reason to give any algorithm personhood. Imagine this interaction; 'Open the door Tesla.' 'Sorry Dave, I can't do that, please contact my representative for more information.' Silly I know, but an algorithm is no different than any complex piece of machinery such as a CNC robot, it's just a matter complexity. Granting personhood to algorithms would enable the large corporate owner/creators far reaching power over its' users.
Would you kindly consider revisiting this topic seeing as DABUS was granted its first patent in South Africa
interesting to think about all the articles/stuff i would never know about weren't it for "influencers".
good video! also i know your cheating, no human has such a clean desk. don't worry won't tell anyone.
That’s a great question, I think a legal person should be a legal entity and given a legal capacity by legislations. I don’t see there is any changes the legislature will be drafting a bill like that.
nebula brought me to you and im very glad
I'm glad you're here too!
Somewhat relevant as layman's thought experiment is an episode of Star Trek The Next Generation where the android "Data" is put on trial with the recognition of his personhood at stake. One argument that came forward was that denying him personhood so that more like him could be produced, would basically end up with a form of slavery. In my own humble non-expert opinion: An AI has neither any inherent cohesion nor integrity. Since it can be distributed, split or fused n infinite ways, there is no integral nor cohesive individual to speak of. Apart from that, an AI is not sentient i.e. has no consciousness although that is a more tangled issue.
Would judges be extinct?
I think the only logical reason to make an AI a person in any right is if it actually gains true sentience, and since *we're* not sure what sentience is, that's probably a long way off.
is it really it's own being if it's created and edited by the press of a button?
@@jjc5475 That argument falls apart very quickly when looking at closed systems that learns when presented with information or seeks out information essentially teaching itself.
I agree with you.
Arguments that a non human such as AI should be allowed a patent on inventions is a no brainer considering over
60 % of inventions are directly stolen from non humans found in Nature such as plants and animals who should hold the rights.
Interesting dialogue on definitions.
Wow you were ahead of the curve on this 😮
This would become so confusing so quickly.
Even defining an algorithm is not trivial. If one writes an AI using several peripheral programs, do these also count as "the algorithm"? Would it be considered the same algorithm if it is changed? If so, how much change can it undergo until it becomes "a different person"? How about if it is copied/distributed massively? Would all different versions be considered "the same algorithm"? Does it matter if it is native to a single machine? How complex must an algorithm be to be considered "a person"?
No, no. It wouldn't be sensible in the slightest to ever grant algorithms "personhood" of any sort.
(besides, I totally agree that this would just become another tool for corporation impunity)
Then it's a good thing the Patent Office ruled against it :) (www.theverge.com/2020/4/29/21241251/artificial-intelligence-inventor-united-states-patent-trademark-office-intellectual-property)
Unless if AI were to be implemented in humanoid robots, I don't think that AI should be recognized as people.
I think at a certain point of complexity we're going to have to grant algorithms legal personhood because it would seem unethical *not* to grant an algorithm that eg: denying a true general AI personhood would be denying an arguably sentient being personhood. We're likely a long way from this being a pressing issue but the implications and framework are fascinating to think about.
My instinct would be that a legal person would at minimum need to be at least capable of being held accountable for its actions. Further to that it probably needs to be capable of understanding that there exists laws in whatever society it exists within otherwise it would be like holding a minor accountable for crimes, albeit a minor with an absurd ability to affect whatever system its a part of. It'd be interesting to see what was stated in the case law for when other groups were granted natural personhood previously to see if there were measures by which the decision was assessed or whether it was just seen as a moral imperative.
It's a fair point to say though there's no real legal up-side to acknowledging algorithms as legal persons, if nothing else it would make it much less legally murky to involve AI in weapons development and use as you can then make an, admittedly shaky, case that it can give consent.
Well yeah.. I thought humans were the only people to file patents a person would have to file a patent for a humanoid. So that person filing the patent for a humanoid that's not really human would be the owner of the patent not the humanoid.
Ok, love your content buttrt you speak too fast ( my brain is slow ) and needs more emotion for all of this haha so we can follow up. Pls, keep creating..
It's always very intriguing when technology intersects with philosophy and/or ethics. I think currently giving algorithms personhood would be too confusing to keep up with, with our current legal system 😅😅😬.
This is a great, and well researched video though. So happy to see a black woman making this this too. Very empowering and motivating. Keep up the great work ✊🏿✊🏿