Warfare of Art & Law Podcast
Warfare of Art and Law Podcast sparks conversation about the intriguing – and sometimes infuriating – stories that arise in the worlds of art and law with artist and attorney Stephanie Drawdy.
Warfare of Art & Law Podcast
AI & IP Panel Discussion: A Global Perspective
Show notes:
3:00 David Newhoff - question of authorship
7:15 Peter Wasilko
9:00 Andres Guadamuz - blog post on AI copyright authorship
10:30 China’s focus on “intellectual achievement”
12:20 Section 9(3) of its Copyright, Designs and Patents Act 1988
13:00 Emily Gould - whether copyright is fit for purpose
13:30 UK joint evidence session on the future of AI and copyright law
17:15 Newhoff - use of an artist’s style
18:40 Wasilko - an artist’s training of a model with its own work
20:15 artist's post-stroke gen-AI recording from model training on his work
21:00 Salles Bruins' question on definition of intellect
25:40 - Ankit Sahni - China’s protection
28:30 Sahni - India’s position on creativity falls in the middle
29:00 Ankit Sahni - RAGHAV output “Suryast”
33:45 Ankit Sahni - protection of AI-assisted works by China’s courts
35:00 Wasilko - hypothetical of photographing sunsets on VR headsets
36:50 Ankit Sahni - USCO’s case by case basis
37:50 Newhoff - what is actually protectable against infringement
39:30 Sarony decision: looking at human choices used to create photos
41:00 Newhoff - ‘authorship by adoption’ is a “bridge too far”
42:15 Salles Bruins - question about training in Wasilko’s hypothetical
43:10 Wasilko - “bridge too far”-requiring license to “learn” from works
48:00 Stanford’s CodeX Group - talk on product JudgeAI
50:30 Andres - human creativity exists irrespective of copyright
52:00 Salles Bruins - copyright is a tool to enable artists to profit
53:30 Kritika Sahni - defining intellect dependent on AI context
54:50 Ankit Sahni - sui generis system of registration
58:45 Gould - applying a right like copyright to output "tough" to get right
1:02:00 Guadamuz - Ukraine’s sui generis right for AI works
1:03:45 Jason Jean - defining intellect
1:08:50 Newhoff - unconvinced that it’s a “sui generis question”
1:09:30 Wasilko - whether inputting human work makes model “assistive”
1:13:00 question of global copyright approach
1:17:15 what is the end game?
Please share your comments and/or questions at stephanie@warfareofartandlaw.com
Music by Toulme.
To hear more episodes, please visit Warfare of Art and Law podcast's website.
To leave questions or comments about this or other episodes of the podcast and/or for information about joining the 2ND Saturday discussion on art, culture and justice, please message me at stephanie@warfareofartandlaw.com.
Thanks so much for listening!
© Stephanie Drawdy [2025]
To say that I'm not feeling very sanguine about developing decent cooperation and handshakes between some of the major developers and the creative community.
Speaker 2:I'll believe it when I see it, but I'm not seeing it right now Welcome to Warfare of Art and Law, the podcast that focuses on how justice does or doesn't play out when art and law overlap. Hi everyone, it's Stephanie, and that was David Neuhof, co-founder of Rights Click and author of who Invented Oscar Wilde, the photograph at the center of modern American copyright. What follows is a recording of a recent Second Saturday online gathering that focused on AI and IP. The panel of guests that convened to discuss this topic, in addition to David, included Emily Gould, the Institute of Art and Law's Assistant Director, dr Andres Guadamuz, a reader in intellectual property law at University of Sussex and the editor-in-chief of the Journal of World Intellectual Property.
Speaker 2:Ankit Sanhi, an AI and IP enthusiast and the co-creator of Suriast. Kritika Sanhi, an IP attorney and partner at the India-based Sanhi Associates. And Stefania Salas-Bruins, an artist and attorney. As you'll hear, this panel brought an international perspective to the issues raised by AI and IP, including the question of authorship and whether tried and true principles are sufficient for generative AI or whether a sui generis right should be created. Welcome to Warfare of Art and Law and Second Saturday. Thank you so much for all being here. A special thanks to our panel guest, and we'll begin with Emily Gould, the Institute of Art and Law's Assistant Director.
Speaker 3:Hi, stephanie. Thank you so much, pleasure.
Speaker 2:Dr Andres Guadamuz, a reader in IP law at University of Sussex and the editor-in-chief of the Journal of World Intellectual Property.
Speaker 4:Thanks for having me.
Speaker 2:Ankit Sanhi, AI and IP enthusiast and the co-creator of Suriast.
Speaker 5:It's a pleasure, thank you.
Speaker 2:Kritika Sanhi, an IP attorney and partner at the India-based Sanhi Associates. Thank you for having me, stephanie. Thank you, david Neuhof. Co-founder of RightsClick and the author of who Invented Oscar Wilde, the Photograph at the Center of Modern American Copyright.
Speaker 1:Thank you, thanks very much for having me.
Speaker 2:Stefania Salas-Bruins, artist and attorney.
Speaker 7:Thanks for having me, hi, everyone.
Speaker 2:And I am Stephanie Drotty, also an artist and attorney, and we will begin with the question of authorship, and that's a question that David Neuhof had honed in on. So, david, would you like to begin?
Speaker 1:Sure. Thanks so much for letting me kick this off. So you know, ai and the question of authorship is one of the key challenges. I think that's facing a lot of people, you know, both artists and the IT community. For those who are following it, the Copyright Office very recently just about a week ago released its second report on this very question of whether or not works are copyrightable when AI is used. I can say from my perspective that I generally agree with the Office's perspective.
Speaker 1:That number one certainly under US law, human authorship is required in the expression that is produced in order for copyright to attach at all. That use of assistive tools, for example, that might change grammar or help you with color correct things like that that have been around a long time that this should not actually affect copyright protection. When you mix generative AI with human authorship and you combine those two, then we get into you know what is kind that expression created by generative AI should not be protected. But this obviously opens up questions that will get into maybe some new complicated areas in discovery processing in courts. But that's where I think we are as far as authorship, at the moment, at least under US law.
Speaker 2:Thank you, David, and when I was looking through the part two actually part one and part two of the US Copyright Office, I was pleased to see that you were referenced in many of the footnotes.
Speaker 1:Thank you. Yeah, I was pleased as well.
Speaker 2:Good to be heard. And on that, the part two, talking about extending copyright protection to generative AI works. One of the points that you made I thought was really interesting was that at a certain point, the application of copyright law itself might become irrelevant and or unconstitutional.
Speaker 1:Sure.
Speaker 1:So you know.
Speaker 1:What I was thinking about there, of course, is that that's pulled from a larger paragraph where, if we start just arbitrarily saying anything that exists as a work, of course is that that's pulled from a larger paragraph where, if we start just arbitrarily saying anything that exists as a work of expression, simply by manifesting, is protected under copyright. This opens up the possibility that, say, company A and Company B are generating works, and if everything there is protected under copyright, it raises all sorts of questions as to whether or not either one could arguably defend its own rights against the other. And that's what I was thinking about in terms of it becoming irrelevant at a certain point. Who has standing? Are all works ultimately independently created? If AI A is unaware of the work of AI B, right, it opens up kind of a can of worms in that regard, and one way to avoid that, of course, is that we decide that, no, that work that is purely generated by machines is simply not protected, that we maintain the human authorship. Doctor, that's what I was getting at in that paragraph.
Speaker 2:Yeah, and to me this relates closely to a point that Stefania raised about the question of how we define and look at intellect. So if anyone wants to make a comment about authorship or anything that David has just said, please jump in, and otherwise I might just shift the conversation a little bit to Stefania for her thoughts about it, and someone quickly I'll go to has raised their hand. Peter, I believe. Go ahead with your question or comment.
Speaker 8:Okay, I just want to comment on how difficult it is to use a generative AI system to achieve a result that I already have in my head. There are so many models, so many individual parameters. That combination has a mind-boggling number of possible permutations of settings. Before the generative AI starts spitting out images, out of which I'm going to select maybe one out of 50 or 60 that come up, I'll be doing extensive research across websites to choose which model might give me the best chance of getting the kind of image that I'm working towards. So eventually, by the time I get in, yes, the generative AI did the image.
Speaker 8:I don't have the artistic ability to express it, but I already had the kind of an image it produced in my head and went through a tremendous amount of effort and cross-com, cross comparison and experimentation in order to achieve that result. And I wonder whether an artist who documented that process might be able to come back in and use that as justification to be able to achieve a copyright on the AI at work. And has anyone looked at that problem? Has anyone tried to come and say, okay, here was my process and how I was able to get to the end result, and here's how much time and effort went into tweaking parameters of the model, and here's how many times I ran the model in order to get this final output image. And is that enough to be able to assert human authorship for copyright purposes?
Speaker 2:I completely relate on many of the points that you're grappling with the idea that sweat of the brow is not going to be a factor and yet, on a case-by-case basis, looking at all of the points that you just raised and the amount of different prompts or modifications and things like that that go into it, how we're going to go forward with a case-by-case analysis of which work is or is not copyrightable. Is there anyone else who had thoughts?
Speaker 4:I think it's quite interesting to contrast what's been happening in the United States with sort of what's been happening elsewhere. Now, funnily enough, I just published yesterday because I was preparing for this I decided, okay, I'm just going to write down a lot of my ideas and just publish the blog post and mostly what's been happening elsewhere, and it's interesting that what we are seeing to emerge is first, lots of things have been happening very quickly. It's just under three years that the generative AI sort of exploded into everyone's attention. That the generative AI sort of exploded into everyone's attention and sort of the idea that everyone had is okay, none of this is copyrightable, of course, and I think that a lot of this conversation was rightly led by the US Copyright Office. But what has been happening interestingly is, particularly, China has now had four cases One denied copyrightability and three have accepted some form of works generated by AI to copyright protection. And what has happened with them, with those cases? What is common in those cases? With those cases? What is common in those cases? So we've had one that predates sort of the degenerative AI explosion and two that have happened more recently 2023 and 2024.
Speaker 4:What has happened is that the courts have looked at what the human intervention has been. So Peter was making a good point about this. Is anyone that has worked with image generators, but also with text prompting, to get any good results? To get slop is very easy, AI slop is extremely easy, and now we're seeing this all of the time. But what has been happening is that for you to actually get anything good and I would like to think of myself as at least having achieved some merit in this is that you have to do a lot of work and you have to do experimental prompts and lots of choice of outputs. And that takes us to what is protectable in other countries, particularly what we see in Europe, in the UK, as intellectual creation that reflects the personality of the author and their free and creative choices.
Speaker 4:And in China, what has happened is that the courts have decided okay, what you have done all of these complicated prompts, but also the selection, it is what they call intellectual achievement, and so at least three decisions have now declared images that were generated with Midjourney to actually have copyright. Now, with mid-journey, to actually have copyright. Now, I've always been very keen on for us to think okay, we have to look at the basics of copyrightability. What's happening in the UK is an entirely different thing. We're probably going to lose Section 9-3, the famous or infamous section 93, depending on which side of the aisle you're sitting. So what has been happening is, I think, that with the US Copyright Office and also what everything else that has been going on, is a recognition that most of these things are tools and, just like photoshop, just like photography, just like word processing, etc. These technical tools are going to produce, sometimes, works that are protected by copyright and some things that they are not. Sorry, I went on too long that's great, andres.
Speaker 3:So I think already what we're talking about is bringing us on to the wider question of whether copyright is sort of fit for purpose in the whole AI environment and whether we can apply the long standing, tried and tested rules to the new environment, to the new environment.
Speaker 3:And I was listening just the other day to a really interesting joint evidence session of the Culture, media and Sport and Science, technology and Industry Innovation I think they're called committees who were looking at all of these questions, because at the moment most of you probably know this, we're in the midst of a government consultation in the UK all about this very issue, ai and copyright and so they're gathering evidence and they were talking to both AI companies, particularly sort of young AI companies, smaller businesses on the one side and then members of the creative industries on the other, and what came across quite strongly I felt there were mainly people in the publishing industry.
Speaker 3:They were quite confident that actually copyright law and those basic principles you know, as they have stood in the UK since sort of 1709 with its famous statute of Anne, that the first copyright statute those principles are actually doing okay and the balance that copyright is seeking to create and maintain between those rights of control of the authors and then the accessibility for everyone else.
Speaker 3:They were actually appropriate and would work in the world of AI, but they felt that at the moment there are just sort of flagrant infringements of those of those copyright rules, and that what needs to happen is sort of better enforcement and transparency, because I think there was a. What came across to me was that there was a real sense that the control you know, the control that, as copyright owners, they should have, was actually just being, you know, taken from them, because they simply don't know when their data and how their data is being used by AI developers, and that sort of lack of knowledge means that they can't really do much about it. Sort of lack of knowledge means that they can't really do much about it and they just feel, feel that, you know, it's that the control that copyright is supposed to give them, um, is is sort of being, you know, being shifted from under their feet, but they feel that, you know, if copyright were properly enforced, then actually it is still a relevant and, you know, appropriate tool for the protection of creators.
Speaker 2:Thank you, Emily and Andres, everyone else, please feel free to jump in.
Speaker 1:I would only add that, on that last point about copyright being fit for purpose, it's important because obviously we switched a little bit there from using AI as a tool to create works versus the question of infringement for mass or mass infringement for the development of AI tools. And they're two completely different questions, of course. And the first question of machine training and infringement is certainly in the States working itself. It's way through about 25 cases at this point or more, and we'll see where that goes. But you know, is copyright fit to that purpose? Absolutely. I would also say that it's still fit for the questions that we're dealing with in terms of once we start to use AI as tools, can it address this question of copyrightability using it as tools? So I don't see any real reason for a new regime per se.
Speaker 2:David, I was re-listening to our interview from, I believe, last year and we, I think at that point had touched on this issue of style as well. It's raised by the US Copyright Office recently and, I believe, their Part 1 report. Since we had touched on this before and I know you have opinions about that, would you like to raise any thoughts on that?
Speaker 1:Sure, I mean, the copyright office's first report was on on use of likeness. They were really addressing something that's technically outside copyright law, which is right of publicity questions, which is being raised in a different question, in a different light right now. You know, I don't think the view has changed all that much in the sense that that copyright historically does not protect style. However, generative AI certainly gives me a new opportunity to say, oh you know, make me work in the style of Stephanie, and certainly creates a new potential threat in the marketplace for your work. But it starts to cross over into other areas of law like fraud and again what we traditionally call right of publicity. I don't know that necessarily copyright will address that question, but there are already new proposals, new legislative proposals to address the question. We've got a ways to go to get there, I think, but I think it's outside copyright law for the most part.
Speaker 2:And, peter, did you have another thought or question?
Speaker 8:Yeah, I wonder if we could maybe try to make a trade dress kind of an argument for that. And also I wonder if it would make a difference if an artist was using a model that was trained exclusively on their own prior art so that they'd be generating it, but the model itself was trained exclusively on your paintings or your photography and then produced more in the same style again of your work, but that's because it was trained entirely on your original work product.
Speaker 1:So I'll answer. I don't want to hog the room, but I mean so. I mean, certainly, if you use your own work product to train your own AI, there's no question of having infringed anyone else's work to get there. It doesn't necessarily address the question of whether anything new that comes out is protectable, that you've engaged in any kind of authorship under the law and, as you alluded to earlier, the amount of work, that sweat of the brow that is supposed to be discounted under copyright law. And so the question and we'll see how it turns out in Allen v Perlmutter a different Allen, because that case specifically addresses the volume and complexity of prompts that he used to generate that work, which he claims to have been. You know his own mental conception, but we'll see what the court says about that particular case. He's going to be the first one answering that question.
Speaker 2:And also, on that point, the idea and this has been touched on by the Copyright Office that individuals who, for example, they name a certain musician who'd had a stroke and he could no longer perform, and he was able to recently release a song, after, I think, a decade, based off of his own prior work and creating a new song, and that to me raises such a compelling reason why you should perhaps consider certain case-by-case exceptions where you could extend copyright.
Speaker 2:So, unless there are any other points to be raised, stefania, on this topic of intellect, we talked about how China had discussed this intellectual achievement concept, the question of how we define intellect or intelligence. I just thought it was such an interesting point that you would raise, and so then I went and looked at the definition. You know the different dictionary definitions for intelligence and one that I thought does not help the case for human authorship, the power of knowing, as distinguished from the power to feel or to will and to have the capacity for rational thought, especially when highly developed, like just different definitions like that that have nothing to do with the human aspect and could totally just be put into the machine realm as these tools develop. So I'm not sure if that's what you were thinking of, but go ahead and give your perspective on this question.
Speaker 7:Yeah, thanks, I actually raised the question out of interest to hear all the experts' positions on it. It was more of a. My input is the question. I don't have a position to take yet, other than, as far as I know, these bots don't yet have what the standard definition of intellect is, even according to the programmers who make the bots, which I think is funny that we leave that out of the discussion. We talk about the person who uses the prompts, we talk about the artists who work as used by the bots, but we never talk about the coders who created the actual function that can do all this stuff that the prompter inputs. Yeah, so basically, I'd love to hear people's perspective on intellect, and that comment about the decision in China was very interesting, so maybe if we can pick up from there, that would be fun.
Speaker 2:Thank you, so anyone.
Speaker 4:Just perhaps briefly, on the question of the programmers and the developers of all these tools. Most of these have been created by employees of companies, so work for hire may or may not apply, depending on where you are sitting, but for the most part, these companies are not interested in having any ownership on the outputs that the works generate. I think that there are very interesting good reasons for that. If they claimed ownership, then I think that they would be opening themselves to more liability potentially. But yeah, so we come to an issue of this being about platform liability and intermediary liability and all of those wonderful problems for those of us that also have internet law hats. I'm a big internet regulation nerd. I can talk for hours about platform regulation intermediary liability, but yeah, that's a different question. I think that the companies are not interested at all in claiming the outputs.
Speaker 8:Yeah, perhaps more to the point, the minute a company tried to claim the outputs, absolutely no one would use their product. So the fact that from a market perspective it would obliterate the marketability of their product, no one's going to go there.
Speaker 7:Yeah, I meant more as a sort of intellectual, academic exercise of where, if we think of the true meaning of copyright and why it exists and the whole purpose of it, where would it lie from that perspective? Yeah, from a commercial standpoint, I understand, there's no, it's a moot point. But I think that the you know there's so many aspects to this problem and sometimes I think we as a whole, as globally, we lose sight of what the whole purpose of this was. And I think that going back to the basics, like the intellectual aspect of intellectual property rights, it seems so, it seems redundant to mention it, but I think we do lose sight of it and I don't think those machines are intellectual yet and I don't personally. I guess this is where my belief comes. The amount of time you spent putting prompts in, no matter if you spent 10 years consistently I don't think that would raise to the level of intellect. Personal opinion.
Speaker 5:Yeah, just to respond to what Stefania is saying. You know the original question and then also the China part. I mean with our experience of the Suryast case, of course, you know I'll try to bring in a bit of that On the Chinese cases first, I think what the quotes there are. Of course that system doesn't match any other jurisdiction or at least from what I see the kind of representation we have today amongst the panelists or the attendees, that legal system is completely different and as far as I'm concerned I'm hardly an expert in Chinese law, least of all Chinese IP law, but what I understood in my interactions with some representatives of their I don't even know how it's organized, but some sort of a trade department that had reached out to us for comment and just general off the record discussions about the Surias matter and where we are at in different jurisdictions. So one of the things that seemed to be that at least they seem to be interested in was they kind of draw a distinction between the effort that the human puts into drafting the prompts and the ability of a human to predict also connecting this with what Peter said earlier to be able to predict as to what output they'll be able to see and the lengthier or the longer or more number of lines of prompts you put in. They seem to believe that on a particular system or a particular popular generative AI program, such as with Journey, for instance, if you work with it enough you'd be able to, with several hundred lines of prompts, safely predict or presume with some amount of proximity as to what the output would look like. So they're looking at protecting two aspects the prompts as literary works and not connected with the output by itself, and then the output on a case-to-case basis, perhaps not the output itself, but if in the most recent cases I understand, what the Beijing court protected was an image which was which one part of output, but then it was used and then edited further and there was a derivative that was created out of that. So I think that, doing both the things, they want to protect the prompts as substantive prompts, not general simple prompts as substantive prompts, not general simple prompts as literary works, and then the corresponding output or output by itself, provided there is some more additions or editing that humans make to it, as something which a machine and a human have sort of, or rather a human has taken the assistance of a machine to create and thus having met the level of creativity that that legal system expects.
Speaker 5:I mean, from our experience, kritika and I, for instance, are basically based in India and the Indian law is very close to as Emily was saying earlier. We're based on common law and we're very closely based on UK law. So in a recent landmark judgment of the Indian Supreme Court, what has come out and what has been established is that India, it seems, if there were two ends of the spectrum sweat of the brow and modicum of creativity as two different ends, india kind of falls in between. So it doesn't require a higher degree of creativity, at least not as high as US, perhaps a Canada like situation, or even lower, but definitely more than sweat of the brow. So what happened in the Surya's case was this is basically how we argued it. We said look A, just to clarify, it wasn't a generative AI program, it was something that we got developed as a work for hire.
Speaker 5:And, to Stefania's comments again, we thought it was really it would be fun to name the tool after the developer, because we thought well, for academic reasons, there's going to be a lot of debate about us. Because we decided to name ourselves as the human co-author as well as the owner. There would be a lot of debate about the pictures, the images, the style images that we use, but the programmer would not feature anywhere. So the name Raghav is basically the programmer's name, and then we kind of spent a couple of hours making a neat. It's kind of stupid, but we ended up having some kind of a expanded form and then kind of reverse engineered like an abbreviation out of it. So Raghav stands for Robust, artificially Intelligent Graphics and Arts Visualizer and Raghav Gupta is the name of the person who programmed it on a work for hire contract.
Speaker 5:And one of our first arguments to the office action report received from the corporate office was they said how are you claiming ownership? Forget authorship first. They said how are you claiming ownership? Forget authorship First. Tell us, how are you the owner? So we had to establish that. Well, basically, first principles of property law and saying look, and also first principles of corporate law. Basically saying, look, this is a work for hire product and since we own the tool, we also own everything that the tool produces. Sort of like having a tree in your garden. If it's an apple tree, any and all apples that fall off, it are basically yours. So that was the argument to establish ownership.
Speaker 5:And from there coming back to the point of China and the level of creativity, so in Indian court, I mean, and before the Indian corporate office, what we tried to argue was look, you, as well as other more advanced systems that require a high amount of creativity, recognize the mere clicking of a camera button and the production of a picture as an artistic work or, in certain cases, a separate category called photographs or whatever. And so what we're doing in this process is this is not a generative AI tool, it's a style transfer tool, and what it does is it takes an input image, it takes a style image and it allows you to manipulate or decide the amount of style that gets transferred from the style image to the input image. So we said look, if the the first step, which is the input image, by itself is enough to grant me copyright which it is, because the input image was an image that we clicked off our regular phone on our house's terrace, of a sunset, literally, and that's what suryas means in hindi it means sunset. Um, what we're doing is plus, plus, is is more, is more steps in more creativity. Therefore, after that, because I'm number one making the creative decision, choosing what um sort of style image would do well with a sunset scene, and in doing so we decided to go with Van Gogh's Starry Night for more reasons than one, not just because Van Gogh is my favourite artist, but also not to let the IP office fall back on a moot objection and try and avoid the main question, the main elephant in the room, which was we want to go with a style image which was in the public domain so that they couldn't fall back on that and then refuse it.
Speaker 5:We were kind of trying to force them to address the point of how would this work.
Speaker 5:So essentially we chose Starry Night, then we chose the amount of style that got transferred from starry night to the input image and then came out the output image, which we call suryas.
Speaker 5:So what we tried to argue was this definitely goes beyond the minimum amount of creativity that a human being is required to put in to produce a work. It is original because it doesn't match any other work which is out there, nor is anyone claiming that it does, and therefore that's basically where and in what direction we went, and we believe from observation that the courts in China are trying to arrive at that kind of situation, and at a very rudimentary level. I don't know how much sense this makes, but at a very simple level, it appears that they seem to have assumed that if they were to give protection to AI-assisted works, that would encourage more investment in terms of tech companies wanting to have development in China onshore, especially with the ready supply of GPUs that they seem to have. They're a country that is rich in natural silicon and therefore they're aces in making all of those chips and everything. Same with Taiwan, but we know the political story behind that. But this is what we understand and those are my views on this aspect.
Speaker 8:Yeah, the copyrightability of a photo taken with a one-click camera by a non-expert photographer seems, to my mind, to argue in favor of allowing the generative AI tool operator to get copyright in the resulting image. Now, let's posit. Now that technology has advanced dramatically, so now, instead of it taking 20 minutes or 2 minutes or 1 minute or even 30 seconds to generate a new image, say, I put in a general prompt I want to see sunsets, and every 15 seconds or every 2 seconds, every 1 second, we have a system that's able to pop up another sunset. And I stand there with my regular camera as sunsets are flashing across my screen, and then I click the button on the sunset that I like Now, because I used a camera to take a photograph of the screen which has generated a large number of sunsets.
Speaker 8:So I waited until the sunset on the screen matched, which is analogous to my going outside and waiting for the sun to get into the perfect position, with the clouds and birds flapping by, for the picture I want to take with my regular camera. Then I take a picture with my regular camera of the screen, which again has thousands of these flashing by in just a continuous wave, maybe even multiple images at the same time. So I'm looking at a, I'm in a VR headset and I'm holding a virtual camera inside my VR headset and I have the equivalent of 16 or 17 4K panels open, all simultaneously showing different sunsets, and I turn to the one that I want and I click that button and now I've got my picture of the sunset. It was generated by the AI, but my process is so directly analogous to a regular photograph and it might even be using a regular camera, so it just happens to be a computer screen in the background that was caught with the image of my regular camera. I think we should be awarding copyright protection to that.
Speaker 1:All right, I'm going to respond.
Speaker 7:I think first of all.
Speaker 5:I think, peter, with that imagery you took me back to Star Trek days. It was almost like me sitting in the cockpit on the commander's seat and no. But I think I really appreciate your question. I think this is really valid and I will just leave it at that and look forward to hear other people's views. Except, I think there's a very important part that comes in the second report of the US Corporate Office the second report that they've published, which says we will protect books that have been created with the assistance of AI, but we'll have to see on a case-to-case basis what will get protected and what will not, depending on how much human effort has gone into creating what was created. That seemed to be their answer to what you said, but I think it's valid. In terms of Alice in Wonderland, I remember a quote it gets curiouser and curiouser from here. I think yeah.
Speaker 1:So I mean this conversation just generated multiple different questions and doctrines and including two completely separate sides of the equation. Right One is about copyrightability and, in the US, registering the copyright, which of course we have that formality where other countries do not. So that's one question. And then the other one is in the works that you're alluding to. Let's say, you know your combination of Sunset and Van Gogh, for example. What triggers in my mind is okay, now what's defensive? You go into a court of law. I infringe your work. You know you sue me for copyright infringement and now we start down the road of okay, what in your work is actually your expression, at least under US law, that is defensible there. So immediately I'm excluding anything that came from Van Gogh that's not yours and I start stripping away. And that's really one of the key questions in all of this right is what is actually defensible expression in the work? If you imagine, not just copyright attaching at the point of creation or at the moment of registration, but what could you defend in a court of law, which is obviously the whole point of having copyright protection in the first place. If you can't defend it against infringement, why bother with the protection? So you know, to Peter's comparison it's been made before about generative AI similar to photography, comparison that's been made before about generative AI similar to photography.
Speaker 1:And yes, photography was one of the first machine challenges to the concept of human authorship. Right, and you know, in the US anyway, going back to 1884, the Supreme Court very quickly and easily found look, there are the human choices that were made in the photograph, even though the defendant in that case tried to challenge the very idea that no, you used a machine, the machine made the image, the human had nothing to do with it. The court said no, you're wrong. Here are the choices, we can define them. And the plaintiff's attorney had basically provided the court with the language they needed. It wasn't a very hard case. But since then, of course, we still have the same challenge sometimes when photographs are the subject matter of a case. And still, the court will identify what is the human authorship here, what is not protected? Portraits are a classic right. We saw this with the Warhol case. There's nothing protectable about Prince's face, but there's no question that copyright attaches to Lynn Goldsmith's photograph of Prince's face.
Speaker 1:And yes, there's a certain amount of a kind of convenient fiction or metaphysics to that that we decided that humans operating cameras, even in a split second, in what might arguably be, uh, you know, almost arbitrary kind of photograph, uh, you still get copyright in that photograph. And so you say to yourself, okay, well, if I use generative ai, as peter just described, it's a different. You're crossing a new boundary and the Copyright Office actually addressed this in the report when they talked about the concept of authorship by adoption. Right and simply selecting your favorite image, that, the image that the generative AI created, it created a thousand. And you said that's the one I had in mind or that's the one I really like. That's what they call authorship by adoption.
Speaker 1:And it's a bridge too far in the sense, because it's like saying, well, I found this photograph in a shop, somewhere in an antique store, and I'm going to claim ownership of it. Or I found a piece of it's actually been cited in the copyright compendium. I found a piece of driftwood and it kind of looks sculptural. So I'm going to say it's my sculptural work. Well, of course it is. You didn't do anything to create that. So you know, I sorry if I get lost there. There were, like I said, between those two, that exchange there conjured a lot of different doctrinal questions. I'll stop there and let others take over.
Speaker 2:Thank you, David Stefania, did you want to make a point?
Speaker 7:No, I did have a question for Peter, though, in this and then a follow-up concern, depending on the answer. But in this universe in the future, where all these images are flashing by, are those images based on existing artwork that is copyright protected, or are they images, say, there's cameras everywhere just recording sunsets and flashing that towards you? Because I think, if they are based on copyrighted work, that ultimately in this universe and in this future of yours, eventually there won't be new work produced, because artists produce work and are able to sustain a living producing work because of the protections, whether registered or not, because people people value true artistic work. So ultimately, you'd run out of sunsets, right, at least painted ones. Hopefully the sun keeps coming up in the we keep orbiting, hopefully.
Speaker 8:Because when we talk about how the AI is trained, it's not like we actually have a copy of a whole bunch of photographs of sunsets. Instead, the AI is modeling different features. It might be, you know, an intensity difference in the light in one quadrant of the photo versus another quadrant. That might be one feature. And all of these features are multiplexed in a really, really high dimensional space and the purpose of the prompt is to try to map to location in the high dimensional space based upon the mapping of the terms in your prompt and correlations between them. So I'll have pictures associated with words.
Speaker 8:All of those get broken up into strings of numbers and then you look at the proximity within that space. So there isn't actually a picture representative. Instead, you have lots of features that are represented inside of the AI model. So there isn't even really a copy in a traditional copy sense existing anywhere in this. It's a matter of learning and, frankly, if we're going to do that, then why shouldn't we be able to attack the human artist who went to a whole bunch of art museums and studied in person every Van Gogh painting, and you've done in your head exactly what that large language model is doing with the photographs inside the computer system, and then what we're arguing then is that? What we really want to assert is that you're not allowed to learn from copyright works without having gotten a formal license to be allowed to learn from looking at something, and I think that's really the bridge too far with the use of intellect.
Speaker 7:has that artist looked at those works in the museum? That's the difference. It's the intellect that allowed for that processing up here.
Speaker 8:Okay, the intellect comes in in reverse engineering the correlation between language terms and where in that high-dimensional design space a given kind of an image is going to lie. So I've looked at a whole bunch of prompts, the kinds of images they go, and then inside my head I'm reverse engineering the black box of the AI to figure out where I need my words to take the systems, the tractor, to be able to generate the kind of an image that I want to come out. So I see that as the height of intellect in being able to do that process. Now I personally prefer good, old-fashioned AI with a symbolic approach where you can actually use logic and get out how the image was derived through rule application. I think that eliminates a lot of the problems that we get in this fuzzy neural computing world and I wish people would take more of a look at old-fashioned AI, certainly given today's hardware too. That was another thing. The early first generation of artificial intelligence work was starting to do some really impressive results, but they basically hit a hardware wall and there wasn't enough RAM in systems, processors weren't fast enough, so everything just basically dead-ended and we went into the AI winter and only now with newer chips and being able to leverage GPUs from the gaming world to be able to do art and interesting applications, have we cracked that? But if we took that new hardware and applied it to some of the older techniques that had been abandoned by the research community and they're still mostly abandoned, because of course now everybody wants to be doing LLM work and it's the new hammer and every single problem is a nail to be bashed with a large language model, until you get a result, even when, with a fraction of the compute resources, a traditional AI approach could derive a much more accurate answer without hallucinations. Oh, the nightmare of hallucinations. I have a friend, mark Bernstein, who is a researcher in the hypertext community, and if you ask any large language model about Mark Bernstein, his name appears enough in the literature that it will hallucinate the most insane things. I have yet to see it. Pin an actual paper that he wrote and attribute it to him. Instead, they will come up with paper after paper with plausible titles, plausible journals, plausible citations all of which are completely bogus, attributing ideas that are the antithesis of everything Mark believes to his writing, and it's just maddening. You do not have this problem in traditional, conventional symbolic AI. You do with large language models. It's a mess and it certainly makes them unreliable and not fit for a lot of purposes. But again, everybody coming out of school figures if I want to get a job, I want to get into a large language model, newfangled AI company and we're going to throw that problem.
Speaker 8:We had a talk a while back in the Codex group out of Stanford and there was a vendor coming in. They had a product called Judge AI and they were trying to use a large language model to model the behavior of jurists. And they put up a scenario in front of us and they gave us a case of potential litigation in a third world against a first world megacorp for alleged pollution, without any actual damages being established or evidence on the record of a direct connection. And they asked the judge AI what would be a just result. And then they surveyed the group Now politically on the call, 85% of the people were strongly in the liberal political camp. A small fraction were in a conservative camp. The conservative camp in that case were little Scalia's.
Speaker 8:We looked at the statute that they flashed up in the example and said well, obviously the megacorp is going to win here because there's no demonstration of damages, there's no connection between the allegations on the record before us, so they should lose. Judge AI concluded that, because of the potential ramifications and the indigenous people that are being affected by the pollution, the just result would be to shift the burden of proof and make the megacorp prove they had done everything possible to prevent potential pollution, even if an actual pollution incident hadn't been demonstrated. And I looked at this whole thing and said, yes, it was perfectly aligned with the politics of the group on the call, but it had no connection to the logic of how the statute should be applied in the circumstances, to the facts on the record. And it's doing politics, it's not doing law. Can I just?
Speaker 7:Sorry, can I just circle back to? Since you did answer my question, Peter, thank you. So I think that in that universe, the sunsets were not based on copyrighted work. They were purely coded effects that you were searching out. I think, that makes a big difference.
Speaker 8:Yeah, well, they were based probably on learning from looking at sunsets and looking at some copyrighted images of sunsets, but all of the learning just was deriving a model of what sunsets are like as an abstract sunset level, not using sunset from the training data itself. The training data was only used to derive rules about what are the qualities that people associate with the word sunset.
Speaker 7:Yeah, I think that makes a difference to me, so thank you for clarifying that.
Speaker 4:I just wanted to do a very, very quick pushback, stefania, sorry about this. I wanted to push back on the idea that people create because of copyright. People create because of copyright. I think that human creativity will be alive and kicking well well. Well, after all, systems of property disappear, going back to the Star Trek idea where all property is going to disappear. In the post-scarcity world where you can replicate everything with the press of a button, I think that we've demonstrated, if anything, we've demonstrated, as a species that gave us a couple of things, and we'll try to make art, music or literature out of it. I know I still write poems, very, very bad ones. I've never even remotely thought about publishing anything. Don't get me started on all of my failed artistic endeavors, but I think that a lot of people create and generate. Human creativities exist irrespective of copyright.
Speaker 7:Oh, I absolutely agree with you and sorry if I was too brief in what I was trying to say. Absolutely, humans will continue to create. What I was trying to say is I was linking, I was using copyright to link creating art like you do, like I do, like a lot of us do, to actually being able to survive in society and make some money. So copyright is a tool that allows artists to profit and, you know, even for tax purposes, you're okay. Being a money losing artist like you're, able to be on the red for a long time because it's something that's valued and copyright allows for that. That's what I'm saying. I don't think humans will stop doing it. I just think that artists will not be able to survive as much from a financial standpoint.
Speaker 4:I completely agree with you. Yeah, those are two different things. No one has ever accused me of being an artist, but that's a different question.
Speaker 2:And one point I was just going to raise about, uh, the ragav tool that, uh, I know ankit has said before, is that, uh, as you described it, that that tool had less capability than many cameras that are available today, I believe, and so so, just going back to the less hyped-up AI tools that are out there, I don't think Raghav falls in that category from your description, and I was also curious. Kritika, I know you were part of the Suriast project and did you have anything you wanted to raise about that or anything that's been said today?
Speaker 6:yes, yes, we agree with Stefania where she said that copyright is a human intellect. You know it's all about creativity, you know innovation and originality. You know, basically, that needs to be protected and, and in my perspective, you know the definition of intellect, it depends on the definition of, in AI context, what it will be, since the laws are still at a very nascent stage, you know they are still developing and it's very, I feel, every jurisdiction will have their own perspective, like, as in India, like in our jurisdiction, like they have accepted, they have copyrighted our work. But it only depends that whether AI generated work should enjoy, you know, same protection as human created works, same protection as human created works, depends on, you know, the definition of intellect in AI context. That will be, you know, actually developed, at whatever stage. You know, in every jurisdiction, I feel, that objective of IP law is only to protect, you know, human intellect. You know their creativity, their originality and, of course, the innovation that actually is, you know, created.
Speaker 2:Thank you, kritika and Ankit. Did you have a point?
Speaker 5:I just quickly wanted to again circle back to what Stefania was saying. I think you know this makes one, in one sense, a case for a sui generis system to come up because, at the very least, what it does guarantee is that the works that are created with the assistance of AI will be classified in the corporate offices in jurisdictions where, of course, you can apply and register. There are a few jurisdictions, but the US is one of them, canada is one of them. You can apply and register. There are a few jurisdictions, but the US is one of them, canada is one of them, india and a few others. But so far as a sui generis system is concerned, and in systems where you can apply for registration, for copyright, at least in the regulator's records, the corporate officer's records, it'll correctly show as a work that has not been created out of pure human talent but has been created with reliance, and possibly heavy reliance, on machine and in this case, ai. And that would be the key in, or sort of the starting point in, ensuring that human creativity is always rewarded, recognized and incentivized higher than an AI-assisted work. Of course, practically how it pans out we'll have to see, because much of these systems work on self-declarations, very similar to a trademark application. You claim to be the owner, you claim in certain systems to be using the trademark since xyz date and it's really on self-declaration till someone challenges it and then you have to prove it in trial.
Speaker 5:But I think the sui generis system is the starting point, because the other option I'm just asking myself to my mind the other option is to then disincentivize the adoption of technology, which would run contrary to how humanity is developed. We have not been. We've not remained at cave paintings. We've gone far ahead with special effects in movies and so on. Everything, animation and all of these things involve extensive use of technology.
Speaker 5:But that doesn't mean these things don't have economic value and they're not given protection. The right thing is. The simplest example could be in many jurisdictions, for instance even in India, photographs or other works that rely on technology, for instance sound recordings. They enjoy a much shorter duration of protection and in some instances they are also subject to other limitations of scope of protection, such as there are more exceptions that apply to them by the statute or there would be compulsory licensing sections that would apply to such works. So I think we first need to look at how an AI-assisted work or a computer-assisted work could form a separate category of right, and then, from that point, we could of course resume the debate and move forward as we were moving forward, that's really interesting.
Speaker 3:Sorry, Stefania, I was going to say it's very interesting and I was thinking, as you were talking, I was going to ask you what kinds of rights you thought might attach to that sui generis right for AI-generated work. In the context partly of the current debates in the UK at the moment about computer-generated works, and it looks like, as Andres was saying earlier, it looks like they might disappear and I think that I mean I don't think they've been sort of very well used and I think there are sort of fundamental sort of conundra, really that from um applying a right a little bit like copyright, which is based in these ideas of originality and authorship, to something created by machine. So maybe it could work. That's to be generous right, but I think, uh, I think it might be quite a quite a tough one to get right no, I, I absolutely agree.
Speaker 5:Another question that often puzzles me now, since what david said is, for instance, I related to that um. We had a nice discussion with a uk-based um body. We were closely with them, called uk music uh, and they've been, like most of us, particularly perturbed by by all of these developments and what I learned them. I'm not much of a music enthusiast, as much as I am for art, for instance, but what I learned is they've had, when you speak about AI in the music context, they said we have this magic tool called Auto-Tune which can make any non-singer bathroom singers sing, like you can make it to the Grammys. And they say look, what is it?
Speaker 5:Basically, since we had the kind of studio recording and the kind of structure that we have, the songs they're recorded in layers. Much of the music at times is digital because it's synthetic in the sense there's no actual drums or whatever. It's almost AI generated. In fact it is actual drums or whatever it's almost AI generated, in fact it is. And so is autotune as a tool by itself. And what I learned from them is and one question that came up was when autotune is applied to the voice layer of the song, because there's a bass layer, there's a tune layer, there's several layers and then there's a vocals layer. When an autotune filter is applied, it's ear powered and it basically just sort of sharpens the rough edges a bit and gets it in tune with a particular note that you're trying to sing. Or you tell the autotune that this is what it's meant to be, but the singer was off sing, or you tell the autotune that this is what it's meant to be, but the singer was off.
Speaker 5:So when you look at what the US corporate office had to say in our decision, like the Board of Review in the Surya's decision it's there on the corporate office website, or even what they say in the report is unless you can discern and separate a human effort from the machine effort, all of it falls in public domain. A human effort from the machine effort, all of it falls in public domain. So I asked myself this question how and under what circumstances can you separate what the autotune AI did to the vocals layer? What part of it was their contribution? What was the failing which? Arguably the entire vocal layer of the song will fall in the public domain. I'm not quite sure, despite all of the advent of technology and all of it shifting focus from human effort to machine effort and making things easy for people who are less talented to produce. But I would say throwing everything in the public domain is definitely not the intention of the law either, and that's one of the questions that comes to mind.
Speaker 4:Just briefly on that. First, I'll make perhaps the obvious comment that the Grammys are getting into. The Grammys is a very low bar nowadays. I'll just mention that. Interestingly, going back to Sojournerist Rights, ukraine has actually implemented a Sojournerist Rights for AI works. Nobody knows how that is working. They implemented it don't quote me on this I think January 2022, if I remember correctly. So, besides fighting a war, they managed to change their copyright law.
Speaker 4:Because of everything that is happening in Ukraine, we don't know it. We just don't know what's happening with that. But they created it's a 15-year right. It has to be. It's a registrar's right, so you have to go to the IP office. I think it's not a copyright office, it's just generally an IP office and you have to request it and it's for very limited commercial reuses. So it's mostly attribution and some commercial rights, but it is also very limited in time. The answer is we don't know how that has gone. I've heard from some colleagues last year from Ukraine that it hasn't been applied that much, but that could be just that creators are just much busier with other things over there. But yeah, it is the only country that I know that has implemented some sui generis right. Okay.
Speaker 9:And Jason, did you have your hand raised? Human intellect versus just like what the AI model like can create, and I thought it more. It more so like boiled down to like ego, like everyone has an ego, but a machine doesn't. The only thing that the machine can do is basically function off of what you like give it. So I don't really think it's creating anything. If anything, it's just like trying to depending on, like the word prompt that you give it. It'll try to make it amalgam like close to, uh, the prompt that it was given, um based on its, I guess, like repository of just like images that like it scoured from the web.
Speaker 9:It made me think back to like two instances. There was an art competition, a piece of AI generated imagery was entered in and it won because it was also trained off of another like well-known artist and the, I believe, like the prompt that was like given out was just like make something that's like close to the heart, and I thought that's kind of ridiculous because the machine doesn't have one. How could it, like it interpreted that out was just like make something that's like close to the heart, and I thought that's kind of ridiculous because the machine doesn't have one. How could it like it interpreted that? And the second instance was a well-known, like Polish fine artist.
Speaker 9:Uh, someone trained an AI model off of, like his art style, um, and also tried to impersonate him for money. And even though he tried to bring up the subject of like copyright, of just like, you can't do this, you can't just like make, you can't just train like an AI model off of my works and then like, try to like, blend it and like merge it together to make this like strange amalgam for for, for money. I think that he lost because I I didn't read much into it, because I didn't know much about copyright all the time, so, like, all the words didn't really make that much sense to me. But, um, I don't know, I just feel like, when it comes to defining, uh, human intellect, one thing that I always consider is the ego. Like, what do you want? As living beings, we're in a constant state of need. A machine doesn't really need anything.
Speaker 2:Thank you Any comments?
Speaker 8:Maybe the artist should be going after them on trademark grounds.
Speaker 1:Yeah, trademark doesn't actually protect that in general, at least under US law. It's a little trickier. I was going to go back, though on the subject because, again, it's important to always distinguish between what we're calling assistive AI versus generative right. So autotune, I would argue, certainly under US doctrine so far, would probably fall under assistive AI and wouldn't really affect the copyright ability of the final song. Where you have a lot of expression, you have selection and arrangement, you have all these, all these rationales for why that finished song has has copyright protection, and the auto tune shouldn't really affect that in a similar way, copyright protection and the auto-tune shouldn't really affect that In a similar way.
Speaker 1:I watched a demo not that long ago of a special effects guy in motion pictures. He showed how he took a scene of I don't know, it's probably 100 frames for the scene and used an AI to interpolate some of the work in between the frames for stuff you wanted to clean up. Likewise, and in fact the copyright office almost explicitly said that that kind of use in a larger work like a motion picture really should not affect the copyrightability of the motion picture. On the other hand, when you start using generative AI to create expression, that's where we get into the question. And you know, I personally don't think that it's as novel a set of questions per se from a doctrinal perspective. I mean, yes, it raises some new challenges in terms of the application of the law. But I'll give you an example.
Speaker 1:We, not that long ago, through Rights Click, helped register some photographs of the sculptural work. This woman finds old saws and I mean literal old saws, not the expression old saw and she turns them into sculptures. And anybody who knows copyright is going to tell you categorically, you know, the sculptural aspects of that are protectable, but the saw part of it that she found is categorically not protected. And so we already have a framework for talking about this. If I present you with a visual work and I disclaim it under a limit of claim doctrine and say you know, generative AI did X and I did Y and I'm claiming my copyrightability in the Y part, it's not really a novel area of law per se and we have a framework for discovery and for litigating all that. And so I'm not convinced that it's a sui generis question per se. And all of this notion of whether or not generative AIs, as machines, think and feel and have motivation to create art. I'm fairly cynical on that notion as a rule, philosophically and socially.
Speaker 8:David, do you think it would make a difference if you're using text to image versus image to image? What if I start out by sketching what I want roughly with my horribly subpar personal artistic ability and I feed that to a gen-AI system, which then turns my sketch of the room and the setting into something that looks like it was done by Van Gogh or turns it into what looks like a realistic photograph? And can I then argue that that's just assistive AI, because it started with the badly drawn hand sketch? And then is some court going to come in and assess whether I have enough artistic ability to be able to say that that's assistive versus purely generative?
Speaker 1:So I think the question is not. It's an interesting question that you're asking, but personally, you know, to me right now, assistive is fairly clearly defined in that you're, you know, you had a mental conception of what you wanted to produce and you're using certain tools to expedite longstanding. You know tools that have been used forever. Photographers have been using photoshop for how long and sitting there and changing the values and color values and whatever in their photographs, um, and if I use a tool to expedite that process because it's learning my, my style, um, fine, that doesn't really affect. But you know, I think the the real question is again going to the Alan V Perlmutter case. There, the artist used text to generate a visual and his claim is getting very close to a claim similar to finding copyright ability in photography. Right, cinematographer? Right where he's saying I I prompted it again and again and changed the complexity of my prompts and I had an image in mind that I wanted to create and I just kept telling it. The same way, a director might tell a cinematographer what to do, or even a photographer in a studio might tell everybody how to arrange everything and what they want until they get what they want.
Speaker 1:You can see the parallel right now in Now. In part, it depends on the integrity of the person making that claim that says yeah, this is what I had in mind, this is what I had to get, because, of course, the generative AI, unlike a camera, can produce something on its own without you necessarily telling it to, and that's part of the gray area. But whether you're using image to text or text to image, I don't think necessarily matters so much as you being able to make a claim that you can. You've got enough control over the tool to generate something that is copyrighted copyrightable, forgive me. Um, and right now, the the feeling, at least according to the office and according to others, is that right now, generative ai is a roll of dice. It's just you don't have that level of control. You can push it up to a point and claim that yep, that's what I had in mind, but that the machines aren't there. That may not be correct, the office may be wrong on that, but that's a technological question as much as it is a legal one, right?
Speaker 8:And that's what discovery is for. Yeah, I think, as newer models have more and more knobs to twist and all, we'll get closer to hopefully having that kind of an outcome.
Speaker 2:Unless there are any more points on this, I wonder if we would shift to the idea of a global copyright and the challenges or incentives?
Speaker 4:Oh, God, please. No. Copyright is strictly national. We can barely keep up with our own countries and, as someone that loves comparative law, you'd be putting me out of a job.
Speaker 3:I think it's. Yeah, I know, I completely agree with Andres and think it's probably pretty unachievable.
Speaker 6:Yeah, but then I think we can still consider it. You know it's basically crucial due to you know AI is technology. It's a very borderless you know it's like borderless in nature and you know it has so much of impact to. You know potential to impact. You know societies worldwide, so it cannot. You know AI cannot exist in a vacuum. I believe that maybe steps towards making you know having a global approach can be taken. Like you know, there are so many. We have like WIPO, we have WTOs, these bodies that have actually standardized the guidelines for several aspects. You know we have treaties for, like burn convention, for copyright, so why not for AI?
Speaker 3:I completely agree, krishka, that the conversations are really important and are being had, but I think that it's probably a long way off to actually get sort of global agreement on all of the principles given. But as an aspiration.
Speaker 4:Yes, yes, yeah, going back to that, I know that WIPO are starting the discussion. So they have been holding these international conversations, or mostly online, on different aspects of AI and copyright, participating in a few, and last year for the last SEC, for the last big copyright gathering at WIPO, they had a big launch. They are starting to think about this. I know that there are some efforts at WIPO to sort of a big launch. They are starting to think about this. I know that there are some efforts at WIPO to sort of start thinking about a treaty, but the way things are going, at least at the international level, it's very slow. So it took WIPO the best part of 20 years to produce a treaty on traditional knowledge and the broadcast treaty. I think even longer.
Speaker 4:It is possible, but I think it would require quite a lot of very difficult politics to try to get some harmonization. So at least I think a treaty would be a good idea. There are some concepts that need to be harmonized, or better harmonized. I think that the concept of originality I mean we have quite a few concepts represented just in this panel, so it may be possible to start looking at that. But yeah, that may be something that at the current stage of international IP politics. It will be very slow and, as we know, a few months in AI is a decade.
Speaker 3:That's exactly what I was going to say. Who knows where we'll be in 20 or 30 years' time. By the time, ttc might be in progress.
Speaker 4:We'll be all using AI agents. This conversation will be held by our respective AI agents that will give us notes at the end of the meeting.
Speaker 2:One of the other points that I think, andres, you had suggested is what is the end game? And so, as we're closing out our time, I wonder if everyone wants to give their closing thoughts and include perhaps their thoughts on what the end game for them, how they think it would be envisioned.
Speaker 4:I guess I'll get started on that. I've been really interested in what the end game is going to be. I always make jokes that we're going to have five years of strife and then stability will ensue Stability, stability Sorry, is this thing on, but anyway it is possible. I have no idea what is going to happen, but I think it's going to be a combination of potentially licensing agreements from large collective societies, collective negotiation bodies. Publishing appears to be pushing quite heavily towards that, because publishing is dying. Publishing is dying.
Speaker 4:Potentially also some form of legislation. I think exceptions and limitations are going to be part of that. Having more clarity for developers. That is going to take place of legislation. Don't ask me what that is going to look like. And now that the US government is very, potentially very heavily pro ai, we don't know what may come up with there.
Speaker 4:But um, and potentially, I think, more technical solutions make it easier for um, for creators to have some form of very clearly say, I don't want my works to be trained some form of I know opt-outs are not everyone's favorite solution but some form of technical standard or, I don't know, model poisoning doesn't appear to be working quite well, so it's just a combination of tech. This is what always happens a combination of sort of practical licensing, technical solutions sorry, technical solutions as well and legislation, and potentially the case law as well, is going to start going in one direction or the other, and potentially the case law as well is going to start going in one direction or the other, and for the moment we just can't expect the case law to be all over the place. So that's sort of where I'm. Eventually we'll have to have a resolution. We can't have this situation where everyone is suing each other. That never lasts for longer than four or five years.
Speaker 3:Perhaps I'll come in now because I don't have much more to say.
Speaker 3:I think Andres has summarised it brilliantly.
Speaker 3:I think and I hope that human creativity and the products of human authors will continue to be valued and the products of human authors will continue to be valued, and I think there will be a kind of coming together of what we might call the sort of two sides so the creative industries and the AI developers, and I think that's kind of already starting to happen through the development of licensing mechanisms and you know the sort of uh, the integration, I guess of of those two communities, um, and I think that maybe um sort of intermediary bodies that kind of facilitate, because in a sense, in a way, the kind of the cat is out of the bag and these AI developers already have lots of the, the data that creators might want, might have wanted to control, and so it's a way of, I think, facilitating a dialogue between those two sides.
Speaker 3:And I think maybe bodies that kind of somehow curate data, this sort of data breakage idea, will become profitable businesses. That will be something that we probably see more of, be something that we probably see more of um and more competition within the market, within the ai development market. We've we've seen deep seat recently, which I think is I mean, I'm no techie but it seems to me, from what I've read, to be a bit of a game changer, um.
Speaker 8:So yeah, I think kind of we'll muddle through in terms of the law and cases cropping up over the next five or so years and eventually things will kind of settle yeah, I think I could see a future where some of the AI, gen AI tool makers start partnering with artists, maybe marketing individualized personal models, where it would be a split between the AI company and the artists that they train the customized submodel on to be able to produce artwork in that artist's style.
Speaker 8:Maybe there might be a role for non-fungible tokens in here somehow. I'm not sure exactly how that would play out, but I think that the ultimate customers want to see the artist be able to make a good, healthy living for themselves. So, even though anyone can go out and get a poster of the Mona Lisa, the Mona Lisa original has tremendous value and for the same notion, even though we might be able to find Gen AI versions that look like Stefania's art, she'll still be able to find organizations that will want to patronize her as a human artist by virtue of her being a human artist, who brings that talent to the table and her own perspective behind it, as opposed to just the soulless images that might be similar in style to her work but not actually be her work. So I think artist trademarks and that level and partnerships with some of the AI companies might provide a way to massage the economics, so it becomes a win-win all around.
Speaker 2:Anyone else, david?
Speaker 1:Sure, I'll be the cynic in the room, which is not news, you know. In particular, if you ask many of us Americans about the end game right now, we're not really thinking about art. But in that context you know Andres mentioned a moment ago you know that the US has moved toward an AI-focused, technocentric message and you know as many as Rick and I know very well the large tech companies in the US use very highfalutin language like innovation, and that's a generalization that doesn't have any particular meaning. And unfortunately, generative AI for making creative work gets rolled into AI for allegedly curing cancer and solving climate change and all of it just becomes one big conversation that is not really being had, it's just kind of being promoted. And so you know, in context to you know what Emily said a moment ago about dialogue between you know, the creative community and the developers.
Speaker 1:With all due respect, creative community have been waiting for that dialogue for about 20 years, since the days of Napster and mass piracy, and that dialogue still hasn't really happened. It kind of sort of did, but not really. And certainly the kind of conversations that are coming out in Washington, as I say, they're so generalized that you know we're just gung-ho on AI as kind of a you know moon landing kind of initiative. I honestly don't know where creative and copyright discussions fall within that broader mess of incomprehensible dialogue. I would love to see copyright and artist conversations still happen in Congress that are separate from the noise that's coming out of other parts of Washington, shall we say. But we'll see. We're not there yet. Let's see what comes in the next. You know several months.
Speaker 2:Thank you, David Ankit.
Speaker 5:Yes, I think I'll just summarize using a quick example or anecdote. Basically, you know, I remember when I was born, I still had my great grandmother alive and what she used to say was you know, there's a very high-ass river in India and what used to happen in the olden days was that you would often go to the riverside. It was really beautiful it still is and you would have all of these painters lined up, and when the cameras came in, the first word that got out was hey, this will be the end of all the painters that line up alongside the river. Nobody's going to get either scenes of the river painted or themselves painted and portrayed and carried home. Nobody's going to get either scenes of the river painted or themselves painted and portrayed and carried home. But then, when the cameras came in, you realize there was always a space on your wall for the pictures and there always still remained a place for the art.
Speaker 5:And what I can say is hopefully, from this point onwards, there would be a third place for those of us.
Speaker 5:I mean, I'm personally not someone who has a taste for that, but probably the next generation, the Gen Alpha, as they say will probably have a taste for AI-assisted or AI generated art, and they will have a third place on the wall, hopefully considerably smaller than the space occupied by the paintings or the photographs, but it will simply occupy another space on the wall and each of these will be side by side. Basically, to sum up and say that human hunger, or demand for each different kind of art form, fine art form, will remain the same. If my cultural or art hunger gets satiated by buying fine art paintings, it will not get satisfied by buying photographs that get framed in the fanciest frames or, worse, ai generated art. So I think all of these will coexist, as time has witnessed and as time has seen. And, by the way, we still have as many painters on the banks of that river in India. Should you come, you will find as many painters there.
Speaker 2:Thank you, Ankit Stefania.
Speaker 7:Yeah, I guess my final words will be and it's more just hope. Painting's not going to go anywhere, painting's here to stay.
Speaker 2:Amen, I agree, kritika. Did you have any closing thoughts?
Speaker 6:Yes, I think you know a balance needs to be. You know balance needs to be, you know, created the right balance where both technology and human creativity will have to coexist. Where the technology has to be encouraged yet human creativity needs to be rewarded, balance needs to be created.
Speaker 2:Both have to coexist there will be links in the show notes to learn more. If you were intrigued by this podcast, it would be much appreciated if you could leave a rating or review and tag warfare of art and law podcast. Until next time. This is stephanie drotty bringing you warfare of art and law. Thank you so much for listening and remember injustice anywhere is a threat to justice everywhere.