Watch Video

See Event Detail

Lindsay Compton: So, everyone, welcome! My name is Lindsay Compton.

Sir Michael Allan Galvez: **Sorry. Where is that coming from? There’s another test here. Gonna close that. There we go. That was weird. 

Lindsay Compton: Good evening everyone. Welcome to the Society of Illustrators. Thank you for joining us for Generative AI and Image Creators, the Future of Artist Rights. My name is Lindsay Compton and I’m the Deputy Director and Director of Programming and Partnerships. As artificial intelligence continues to transform how we create and consume images, artists and designers are facing urgent and complex questions about all of their work, ethics, compensation, and the very nature of creativity. This town hall was organized with the hope that everyone will be feeling empowered with a clear understanding of the current states involved and practical insights to help navigate the evolving intersection of AI and the creative industries. We’re fortunate to be joined by a distinguished panel of artists, designers, technologists, and advocates whose work sets out the center of art, technology, and policy. 

So I’m going to go introduce Carl. Carl is an illustrator, creative entrepreneur, and founder of the visual storytelling studio Epic Made. Through his personal brand “Colorful Carl”, he explores fine art, AI, and watching. He currently serves as the Eastern Regional Representative for the Graphic Artists Guild. And then we have Ismael Kherroubi Garcia on the screen here. He is the founder and co-leader of the responsible artificial intelligence network also known as RAIN, and CEO of Kairoi, an AI ethics and governance consultancy. He brings a wide range of perspectives on his experience in fintech, museums, theater, and academia. Now we have Tim O’Brien. He is an award-winning illustrator and an extraordinary, with an extraordinary legacy of work including numerous Time Magazine covers and the iconic Hunger Games book series. is the second longest serving president of the Society of Illustrators from 2015 to 2022 and has been adjunct professor at Pratt Institute. Tim has a leading voice in the effort to keep generative AI out of illustration competitions and uphold standards for original artistic creation. And then we have Rebecca Blake. She is a veteran graphic designer and former advocacy liaison for the Graphic Artists Guild, where she tracked legislation and authored numerous articles on copyright and visual artists’ rights. She also served on the board of ICoD, the International Council of Design, and ASCRL, the American Society for Collective Rights  Licensing. She was recently hired by ASCRL as Senior Vice President for Interstate and Foreign Licensing and Business Development. And then we here at the helm, we have Sir Michael Allan Galvez, is the Marketing Director of Creative and Digital Media at 54 Below and BFV Management, a photographer whose work has been published widely on a variety of Smithsonian archives. He is also a founding member and colead of RAIN, and serves as Associate East Regional Representative for the Graphic Artists Guild. 

We’re looking forward to a thoughtful and nuanced discussion and we encourage questions during Q &A. And before I turn it over to Michael, I also just wanted to give Rachel Sheeran a shout out and thank you for all the thorough material for the event. So without further ado, please give a warm welcome to our guests.

Sir Michael Allan Galvez:  Hi, good evening. Thank you guys for all being here, whether you’re in person or online. I’ll be your host this evening. And thanks, Lindsay, for that wonderful brief introductions for all of us here. I just want to say it’s my absolute pleasure to be here with you all and our four panel and our three panelists. Four panelists? Yes, three in person and one online joining us. To set the stage, I’d just like us to look back at what led us to this moment in this room. In 2021, the world found out that over 5 billion images were scraped from the web and included into data sets to train several text to image, generative artificial intelligence models. This was done without consideration of copyrights nor compensation. 

This disruption caused by generative AI in the creative industry is vast and it is no exaggeration to say that generative AI poses a once in a lifetime existential issue to the status quo we currently all enjoy. Use cases of generative AI have caused a concern and have found themselves across headlines through different outlets for the last four years. Some of these include articles surrounding lawsuits like Anderson versus Stability AI at Al, Getty Images versus Stability AI, Oldsmith versus Andy Warhol Foundation. All of these are revolving around using those data sets of scraped images that was just mentioned. If you’re local to the area, you might have even seen or heard about the picket lines that were downtown at Hudson Yards in 2023, when the Writers Guild of America and the Screen Actors Guild held a joint strike out of concerns over the use of generative AI in the writers’ room and also in capturing our actors’ likenesses to use forever without what they feel is proper consent and fair compensation. By the way, if this is all foreign to you, do not worry. You can ask all of our panelists here in the Q and A later that’s coming up. So more recently, just last week, it’s been reported that the news website Business Insider laid off 21% of its workforce across all departments, a number of them also being writers, so that the company can go in their words “all in on AI”. 

On the flip side of this, US courts have sided with artists and ruled that images generated by AI cannot be copyrighted*, which sets a remarkable legal precedent for all of the challenges to come. There are also several responsible use cases for artificial intelligence in the arts. One such use case is actually in the realm of art authentication, where AI is being trained to recognize forgeries to great degrees of accuracy, bringing a powerful tool in potentially protecting artist compensation in copyright. Another great use case will be detailed by one of our panelists in just a few minutes. I believe that the blog article published by the Institute of Computing and Cyber Systems by Michigan Tech summarized this entire situation at its best, directly from their words: “If an AI can be trained to steal billions of images from the internet, an AI can be trained to recognize unlicensed/copyrighted images and choose not to include them in the dataset. However, the predatory AI utilized here is evolving faster than the guardrails – big AI corporations have financial resources that artists do not” (Townsend 2024).

Hard to sugarcoat this. This is a big mess. Further complicating this is the involvement of the government. Last year, the European Union, across the pond, adopted the Artificial Intelligence Act, a first of its kind, common regulatory framework that details the risks of AI and aims to provide a more ethical way forward. Critics argue that the AI Act doesn’t go far enough, but I believe that it’s a significant step forward. In the United States, only a handful of state-level governments have been able to enact some form of legislation related to artificial intelligence. If we look at the federal level, the picture gets even bleaker. Just last week, the House passed its version of the Budget Reconciliation Bill, known as the One Big Beautiful Bill, which contains a provision freezing regulation of artificial intelligence in the country for the next 10 years. 

Now that’s a lot of information. And while we can’t cover everything tonight, we will try to center our discussion on a few main points. One, will the use of generative AI stay predatory to artists like the way it started, or will our shared future see a more ethical and responsible regulatory framework put into place? One that makes creativity more human again. How will social, cultural, or legal efforts play a role in protecting our work and our rights? And this is exceedingly important. What can we all do in the meantime as individuals? So that’s where we seem to be in the present moment. Wow. So Carl, how do you feel about all of this? I’m throwing it right at you. 

Carl Cleanthes: Yeah, I mean, it’s a complicated thing, like you said, you know, there’s, there’s the first and foremost thing that needs to be said, like you said earlier, is that these data sets were made very unethically, right? We have copyright law for a reason, and nobody in these data sets were given consent or compensation or credit. And that’s the mantra that I think we’ve been repeating and I think it’s really important to stand on that. 

That said, you know, I see two sides. It’s definitely hurting the creative economy. The bottom level of this, entry level artists right now are probably in the worst space they’ve been in terms of being able to get a paying gig, whether you’re a designer or an illustrator, all that sort of stuff that we thought Fiverr was a big problem originally. This is that on steroids, right? I see on a production side of things though, like on the agency level, right? So I own a creative agency, we do animation, illustration work, and I see a future where my small team can produce much larger projects based off the data set that we can build in-house that references work that we own and did, right? So in my head, there’s a space here soon where we’re going to be able to make key frames and characters concepts and, you know, develop our story and it’s going to help us do it quicker, better, faster with my same team. So I think that that’s really powerful. 

But on the other side, are these huge corporations going to now demand 10x of the staff and use less people and do big layoffs? It’s the double-edged sword, right? So where does that come to a precipice? And that’s sort of the whole conversation lately. But we’re really trying to figure out how do we protect our colleagues in the meantime and figure out how do we navigate this the best way possible. 

Sir Michael Allan Galvez: Yeah, so I love all of that, but also could you… I know personally that you have some experience with responsible use cases of this. Can you tell us the details of that? Is there a responsible use case for this in the arts?

Carl Cleanthes: Sure. I mean, you’re talking about is a passion project of mine. So, you know, having that foresight knowing AI, generative AI is here. I own a company and I have to make sure that we make enough money to pay the bills and take care of this. Everybody’s family that works with me, right? Like I take that very seriously. And there’s no world where it’s like, well, I can lay off half of them and get that. That’s just not how I’m wired. But you know, some people are, but so in that space I’m experimenting, right? Like I want to understand this technology. 

So I started doing research on what are called LoRa models. And LoRa models are low rate adaptation models that are a smaller data set that sits on top of like a stable diffusion model, which again, is not an ethical model, but this is sort of derivative of that. So again, step in the right direction, I think. But I took all of my father’s paintings. My father was a watercolor artist and did these amazing visceral, just like emotional abstract, crazy paintings. I collected 700 of his works of art from his whole lifetime. He was from here in New York as well. And I put them all into a LoRa model and created this thing that now generates artwork again from my father. And the first time I did it, I just started crying. It was beautiful. it’s like, there’s something there that’s really magical. 

But like, what is that? And that’s like a post-human creation. But like, it’s not really my father, but it felt like it. So I don’t know. I think that’s but that owning that data set and all that stuff that went into it, I own the intellectual intellectual property rights up. So that is a more ethical approach and a step in the right direction. Right? 

Sir Michael Allan Galvez: Thanks for that Carl. What about you Ismael from the beautiful land of Canada? 

Ismael Kherroubi Garcia: Hi everyone, it’s Ismael here, calling from Cape Breton, Nova Scotia, Canada, you can tell from my accent. This is the unceded territory of the Mi’kmaq people. I wanted to say above all, well, just start with a huge thank you to the Society of Illustrators for hosting the Responsible Artificial Intelligence Network for the first hybrid event outside of the UK. And a huge thanks to everyone at the Graphic Artists Guild who’s helped get this event together. I hope we do many more. Now I do want to give just a moment to the fact that I am joining online. This is quite an unfortunate turn of events. I really wanted to be there in person. Would have been my first time in the States. Would have been wonderful to see New York and meet all the panelists and all the audience there. But there’s just been too many problems coming out of airports in the US recently and really didn’t feel like it was the safest thing to do. 

Now, even though that’s an aside, it is unfortunately the framing of a lot of today’s discussion. We saw at President Trump’s inauguration in January that he had some of the world’s most powerful men in the world of AI lined up behind him. And this might be part of an argument for the challenging of the uninformed use of generative AI tools in the arts and beyond. When you use them, there are real power dynamics that are being reinforced through their use. 

But I think echoing a bit of what Carl said, I don’t want to get too carried away with you know, it’s not that we shouldn’t use these tools. My question is always going to be, how can we use these tools thoughtfully? What are the conditions that allow us to use these tools properly? And I’m sure we’ll hear more from the audience and from the panelists as to how they’re used. They can definitely replace clunky tools, help brainstorm, test ideas, overcome small budgets like, Carl just mentioned, and I’m going to propose two responses to what the conditions are that allow us to use AI in the arts. 

So firstly, artists, and I might be putting a bit too much pressure on all of you, I’m sorry, but I think you need to find the resources to help improve your literacy, your ability to critically evaluate use cases. And I will repeat this, I think throughout the call today, I really genuinely think that you making it to today’s events or watching this video online on repeat or live, you’re doing part of that. You’re trying to join the conversation. Secondly, we all have a role to shape the policies, the governance of AI models, AI use cases, whether that’s at governmental levels or organizational levels. And ultimately, underpinning everything I say today is the notion that art results from a creative process whereby we bring something into existence. And this process, to me, is just fundamentally human. So really looking forward to the conversation. Thank you. 

Sir Michael Allan Galvez: Thanks so much, Ismael. That was great. What about you, Tim?

Tim O’Brien: First, thanks for all the people that came to this society tonight. It’s great to see you in person. For everyone listening, thank you for taking the time. It could be one of the least sexy things to talk about at any given moment, but it’s very, very important. I want to say that I’m here not as a representative of Society of Illustrators, or the Illustration Conference, which I’m the Vice President of, or Pratt Institute, or the H, or any other organization. Here’s a freelance illustrator with a long career. And I’m also offering my own thoughts. This is not my own self-interest. Though I think it’s fine for everyone to speak to their own self-interest, I’m doing this mostly because I’m an educator and I care about the future of our industry and I care about people who want to be illustrators and I care about the arts and the purity of art. 

And I’ve been alarmed by the generative AI outputs, like most people have. The AI has been perpetrated by corporations, by the media, by Adobe, by every social network, and every search now has a button encouraging you to press it. And I’ve been on numerous panels discussing AI, the dangers of it, the implications of it, and I have some… pretty hardened positions that haven’t changed from the moment I first saw it. I don’t think it’s possible to engage with generative AI in any way as one believes that it was built immorally and illegally. I don’t think one needs to learn how to use it to be against it. I don’t think one needs to learn chat GPT or Midjourney to think that the output of these texts to image word extruders are in fact not generating meaningful additions to humankind. The argument that generative AI is here, the toothpaste out of the tube, the horses out of the barn, is not a reason to value it or make it part of an illustrator’s workflow. 

We do have to respond to it though. And we don’t have to use it, we don’t have to celebrate it or use text image tools. It’s embedded in Photoshop now, there’s no reason you have to use it. Just be stuck for a moment. Be willing to not know something for a second. I oppose the casual acceptance of generative AI in our day-to-day work as illustrators. Like I think any seasoned illustrator or art director who posts experiments with text and image generators is a grand disappointment. We all need to call that behavior out and it’s not easy, because some of these people are our friends and also there’s sometimes our target group we want to work with. 

But there’s a polite way of doing it by saying you really, really admire that person and you think that that kind of activity is beneath them, it’s beneath their quality of their work. There should be a way that we can always have that reaction so that we can continue to make it less than cool.The one thing that has changed in the past few years, it’s not cool. It’s now cemented as less than great. When something is discovered to be made by AI, public hate, they call it out. And that’s our role, primary role as illustrators is to sort of call it out and have that opinion. 

I’ll cut it off there, I kind of came in hot. But I think this conversation today is the beginning of us having these things in public. We haven’t really talked about it as an organization here at the Society of Illustrators. So anyway, that’s my position. 

Sir Michael Allan Galvez: I appreciate that. I think it’s very important to talk about this passionately because I feel like the entire point of the talk is to make sure that everyone comes out informed in a way to make sure that everyone has a way forward. So I appreciate all the thoughts we’ve had so far from our three panelists. Lastly, Rebecca, what do you think about all of this so far? 

Rebecca Blake: So I have so many thoughts. It just to be really clear, again, I’m speaking as myself, I’m not speaking for ASCRL, I’m not speaking for the Graphic Artists Guild. Although I do have to say that my opinions are certainly formed by 30 plus years on the board of ASCRL** and over 15 years working with the International Council of Design. 

So many thoughts about this, but the huge, huge thing I think what we need to consider is, one, generative AI just fell on us like a thunder clap. This was part of a decades long reduction in artists’ rights that has existed as the tech sector has developed. We need to understand how we got here to start. We need to understand that germane to this entire discussion on generative AI, should one use generative AI, how can we use generative AI ethically and responsibly? How can we develop best professional practices of generative AI? Because unfortunately, this is not going to go away. And the artists have to be the ones to develop the professional practices and the ethical standards. Even as I agree with Tim, that you have to hold your nose because the base model was unethically sourced. 

But as we approach this, we need to understand this is part of a larger labor issue. And I am standing on my Graphic Artists Guild, labor union trade association roots when I say this: there has been a transfer of wealth across every industry that has occurred over decades. And we are part and parcel of this. Every single illustration that we make is of value, which is why people want to take it and use it and why they don’t want to pay for it because they can’t afford to pay for it. The AI’s generators say that it couldn’t possibly license all the images we need in order to develop this platform. 

But if we want to effectively advocate for ourselves as working professionals, working artists, then we have to understand the boring stuff that we have been ignoring for years, which is the copyright behind it, which is the labor issues behind it, which is the transfer of wealth, which is the technology that has stripped your ability to control your work. Case in point, Meta built their AI model on all of those images that everybody’s been happily uploading to Facebook and Instagram. And Facebook and Instagram, lo and behold, strip your metadata out. If you’re sensible, you’re putting your metadata into your illustrations, you should. It’s one way you can control it. It’s called copyright management information. There’s tutorials on it, embed your metadata. Social media platforms strip it out, ostensibly because it increases the file size. That’s bullshit. They have enabled larger and larger images to be uploaded. Do not tell me two kilobytes of metadata is going to be the deal breaker. It is stripped out so you cannot tell where your images are being used. All of those images uploaded, no metadata, difficult to source, is what Meta used to develop their AI platform. And if you sign on Facebook or Instagram, you gave Meta the right to do that. Because the terms that you use give them broad sub-licensing rights, and that includes for generative AI. 

This is what I mean about how we have to pay attention to the boring stuff. Because we have things of value. We have our ability to create and the creations we produce and that is being stolen because we have to an extent locked ourselves defenseless. So I think we need to understand that this is a moment at which we can start seizing control of that. 

There’s… I am hugely pessimistic at this point on our ability to protect ourselves with generative AI because of the sea change that has occurred in the political landscape since January. The elections have consequences. It is no coincidence that the Copyright Office put up their third report, a generative AI that said you cannot assume that using creative works or text and data mining to train your systems is a fair use. That is what the tech sector wanted. Lo and behold, the following Saturday, the Register of Copyrights is illegally fired. No coincidence there. There’s absolutely no coincidence there. 

But having said that, I gained hope from articles about how generative AI is producing garbage, because guess what? It is feeding on garbage, because guess what? The majority of people using generative AI to produce images are people who aren’t artists, who don’t have the passion, don’t have the skills, don’t have the training, didn’t put in the hard work, and are putting in little text prompts and producing candy-colored, soft-focus garbage that is now swamping in the internet. And this is what the the Generative AIV is feeding on. This is a chance for us to start leveraging the value of the work we do, the originality of the work that we do, and hopefully gain some traction. But it is going to be a very difficult row to hoe with who we have in the White House and the lack of action on the part of Congress. So that’s how I generally feel about it. Yes, pissed off, somewhat pessimistic, but I’m seeing opportunity.

Sir Michael Allan Galvez: I guess from this, you know, it feels like out of all industries, a number of artists feel like the creative industry is getting the brunt of the consequences, right? of this AI revolution. Do you feel that this is true? And if so…

Rebecca Blake: Yes.

Sir Michael Allan Galvez: I figured you would say that. That’s a follow up. And if so, why do you think that art or creativity or however you call it is…

Rebecca Blake: Because creativity is hard. 

Sir Michael Allan Galvez: Okay.

Rebecca Blake: That is why. Because to produce and by the way, this applies to musicians, it applies to writers, it applies to photographers, it applies to software engineering engineers. There’s an art to software engineering. There’s discipline and dedication to software engineering. The act of creation is difficult and it is a stony path with personal failure after personal failure after personal failure, you’re working, you’re working, working, like, “holy crap, I’ve got the nose in the wrong place” (and I’ve done that before). And then these wonderful things where it all comes together and a beautiful work comes out. And this is what we do for a living, but it is hard. And the rest of the world wants it easy because it’s a value. They want the value without exerting the effort.

Sir Michael Allan Galvez: That’s a nice little zinger there at the end. But I feel like Carl has a response.

Carl Cleanthes: I just, in that same vein, just, think, you know, the difference, you know, I saw a report that basically said, you know, the arts programmers, even legal counsel, they’re going to take big hits from generative AI, both on the language side and the visual side, right? And, but I think artists have historically been undervalued, disenfranchised a lot more than say high paid programmers and high paid attorneys. So we were already in bad spot relative to some of these other places and now we’re getting hit by this as well. So I think there’s an added insult to injury there in the first place that should be also looked at, think, or to take into consideration. 

Sir Michael Allan Galvez: Tim do you have thoughts on this?

Tim O’Brien: Yeah, I have some thoughts about this. I think that to think that we are the only sector that is being harmed is wrong. There are all kinds of entry-level jobs all over the world that people count on to survive. And those are being cut away every day. And there are sectors of illustration that are immediately affected, and those would be animation and any kind of process work like that. 

Carl Cleanthes: Concept art.

Tim O’Brien: Yeah, concept art. I do feel like the guardrails are held in publishing. So I want to say that, that I’m not completely pessimistic. I’m quite impressed with how authors demanded that AI not be involved in their projects, which pushed publishers to change their contracts to not allow AI. And this this has happened in almost every major publisher. If you use AI, you have to tell them you used it and how much you’re not allowed to generate a cover. And that is just sort of the way it’s going to be for a while. And that’s a really encouraging thing. 

Also, people take personal pride in what they do. And no one’s running around taking victory laps for pressing a button. And anytime someone sort of talks about using it, it’s always a bit of like a reluctant, like, well, I had to. It’s all filled with shame. I don’t think that our industry is particularly hit more than other people, but I do feel like it’s easy to show this. It’s hard to talk about a data processor, but you could show images that are generated. As we all remember, Lensa, in about 2022, was asking people to upload all their photos so that they could make illustration avatars for themselves. And I was seeing all the people I went to high school with uploading all their photos and making very reasonable portraits of themselves. And it made me nauseous. It made me nauseous for anyone who wanted to be an illustrator. And that alarm changed over time because it’s now it’s sort of cemented in our industry as a thing you played with. It’s a novelty. 

Sir Michael Allan Galvez: Yeah, I feel like the current trend is doing versions of Studio Ghibli-inspired artwork that’s coming out of image generators. 

Tim O’Brien: That trend is called a crime.

Sir Michael Allan Galvez: So, Ismael, I know from personal experience that you have lot of thoughts about this. What do you think? 

Ismael Kherroubi Garcia: I’ll only add one thing and agree with almost everything already said. The only thing I’ll add is it’s been surprising. I think that might make it more impactful on the design sector, on the illustration, on the arts. It’s been surprising that we’ve gone from, hey we can use large language models to underpin text to text generators, so AI chat bots, to help you write emails quickly or help write resumes and cover letters, to, oh we did the boring stuff, now we’re going to do the creative stuff too, because no one likes drawing or doing art. It’s been a really surprising shift in the rhetoric. It’s been… we were automating in things we don’t enjoy, and now we’re automating literally the only thing keeping me, you know, keeping most of us like interested in life. We’re human, we create things, we enjoy the process of creation. You might be as boring as me and just enjoy writing a report on AI, or you might be, you know, everyone in that room will be actual artists. That to me is what’s been most impactful. It’s not that it’s worse for the arts, it’s that it’s such a surprising turn and so unnecessary.

Carl Cleanthes:  But I’d look at that. I feel like what Rebecca said, that there’s a lot of garbage going into these datasets now as they grow and more garbage coming out. And I’m less doom and gloom than some people might be. Because I think at the end of the day, AI is never going to generate really innovative, amazing, nuanced things that an artist could create. Like, it’ll do something really fast. And it might be beautiful-ish noise. But it’s not going to innovate. It’s not going to create something that has deep, meaningful aspects to it like real art has, you know, unless there’s a real artist really infusing themselves into that experience. And then there’s an artist there. 

But, you know, to Tim’s point, like there should be great pride still that people take at any point, going “I don’t use any AI for this. This is all analog. You know, I do this with a watercolor and colored pencil.” Like, I think that’s going to have more and more value. As we have more more AI generated crap, the people that really hold their traditional skills, I think is going to have an amazing value moving forward. You can’t really emulate that, even with the printing technologies and stuff they have. I think people appreciate that wabi-sabi of the handmade details. I don’t know if I used that word correctly, the enunciation. 

Sir Michael Allan Galvez: Oh, no, that’s a very interesting point. I think on a societal level, is there a world or, you know, or is there one thing that we think someone, whether an organization, a government, can do that starts or catalyzes an ethical framework of AI in this society, jn the United States. Is that something that you think is doable? Is that something, how, how will that look like? What, what is that? You know, what is that first action that triggers that? Rebecca or Tim. 

Tim O’Brien: I think you, I heard you open your mouth before I did.

Rebecca Blake: Well, you know, it’s something I put a lot of thought to when I was on the board of International Council of Design, we actually developed a code of professional practices for designers and it feels like you’re putting on the set of obligations, but what we can realize is when you develop a code of professional conduct, professional practices, you actually empower yourself. You elevate yourself to you becoming the partner and the advisor of your clients because you are coming from these points of standing. 

I think that is what is needed in general in every creative discipline, but I think the generative AI really puts the crux on it. For example, Tim talked about LoRAs. These are fine tuning models where you can use as little as 10 or 20 images and you can develop a model that will then generate images in the style of a certain illustrator. So an illustrator who spent 10, 20 years developing, crafting their own original style, all of a sudden will see the Internet swamped with images that look like they could have made them. They have lost their unique value proposition, which is their vision. Because somebody developed a LoRA that they are now selling based on training on their materials. This is a really clear example where if you are a visual artist, is it ethical? To use somebody else’s work to develop a fine-tuning model? I think we all can say, no, it isn’t ethical, but we should be the ones to go ahead and say that. 

And I wanted to propose that we have this concept in the United States called fair use. You guys have probably heard of it. Fair use is the reason why a rap artist can sample Roy Orbison’s “Pretty Woman” and come out with this amazing piece of work, which is a commentary on Pretty Woman and takes it in a totally different level because they have now imbued this work with new expression, new meaning. And most importantly, they are not competing with the original work. They are not hurting the estate of Roy Orbison, which derives a value from that. This is all under this concept of fair use. Fair use is how we have beautiful collage work. Fair use is how we have beautiful montages. Fair use is how illustrators can incorporate copyrights in their work, but if they generate something that is new, new expression, new meaning, new vision, new creative decisions, then that passes over into an area where it is considered a fair use of that work. There is no right long, you know, hop back and forth and this is fair use and this isn’t fair use, it’s, it’s, it’s a grey area. Yeah. 

But the fair use concept when somebody evaluates whether this use of somebody else’s work is a fair use or not, there are four factors that are considered. And this, by the way, is the case in every single copyright infringement case where somebody says, “I got to use that because that’s fair use”. Those four factors have to be considered each time. And the first factor is (I always have to look this up because I always get it mixed up). The first factor is the purpose and character of the use. This is where this idea of transformation, expression, and meaning comes into play. Have you imbued what you have created? with such originality that it is a new work, even though you’ve used this. The second one is the nature of the copyrighted work. If you’re something that is factual, facts have less copyright, so not original, they’re facts, versus something that is entirely creative. So if you are using works that are entirely factual, then building from it, then that tends to help more with the fair use decision, does it mean you pop across into a little area where you’re in this little magical fair use land? It just means that it’ll help support you. The next one is the amount and the substantiality of what you use. How do you use the entire work? Which by the way, is a really pissy thing with the AI image generation. The only way they can work is by ingesting the entire image. How do you use the entire work? How do you use a piece of the work? And if you only use a little piece of the work, how do you use the heart of the work? If anybody ever tells you it’s okay if you just use 20% of something, it’s a very used bullshit. It’s about 20% of the heart of the work that is not a fair use. And then the last thing is the effect on the potential market. Have you hurt the original artist? 

We can take these four factors and then use them as a framework for developing ethical practices with generative AI. But in order for generative to develop professional practices that have any standing, the first question that has to be asked is, is generative AI being used as a tool or is it a crutch? Could you get where you’re going to go without it? Have you brought that much originality into it? 

And the very last thing I want to say on this, the Copyright Office has done a really, really interesting analysis of generative AI and copyrightability. And what they have said is that a work has to be generated by the human being in order to have copyright. This is built on decades of pre-existing law. Believe it or not, there was somebody who tried to get some text copyrighted by the Copyright Office that they said had been dictated to them by celestial beings. And the Copyright Office said, I’m so sorry, but gods are not human, and therefore we’re not writing any copyright stuff, I kid you not. That actually happened. People trying to have works created by elephants, trying to get a snapshot taken by a monkey all to be copyrighted and all of those copyrights were rejected. So there is a long precedent that goes behind the Copyright Office saying it has to be a work of human creation. 

Having said that, the Copyright Office did grant registration to a work entirely created on a generative AI platform. Some piece of God awful work called “A Single Piece of American Cheese”. Hideous. Why did they grant it? Because the individual who created it, who owns actually a generative data company, so they wanted to actually have a use case, the person who did it imbued so much decision making and creativity into using that platform, a long iterative series of prompts, taking the output, manipulating it on platform, taking segments of it, epositionkng things. And not only did they do that where there was decision making every step of the way they documented it with a video so they could show it to the Copyright Office. It is a hideous piece of work, seriously ugly. But it fulfills the Copyright Office’s requirement for human creativity and most importantly, intention. Not like “I put in the same string of words and look, I got a rainbow color pony out of it.” But actual intention, intention going towards a goal, and that’s the difference in it. So all of this I think is what needs to be considered as we develop professional practices and ethical practices. And the point is to protect artists and to protect the culture of what we’re creating.

Sir Michael Allan Galvez: Tim, did you have additional thoughts?

Tim O’Brien: Just one thing. Additional. That was great. I’m a realist illustrator, and I do a lot of covers where I need elements and to do that I have to do some reference searching. I have to purchase images. I’m working for a client like Time Magazine. They’ll acquire the rights to images from very great photographers that have access to people that I don’t have access to, and people are being paid along the way. The cost of the illustration to the publisher is way more than my fee. It’s my fee plus all those other rights being purchased. 

If generative AI paid for the usage of all the images it used, then it could be ethical. Then it would be competitive because it would be expensive to churn out an inaccurate image and ugh that’s not right. I’ll do it again  $1,500. $1,500. That’s how it should work. Because it doesn’t exist without the work that we’ve made. 

Any generator that generates music won’t exist without music being put in. Any image generator won’t exist without images going in. It’s all been stolen. And so the only way I see it working is if there’s compensation. If there’s no compensation, it’s not fair. Now, if anyone wants to give away all their work, artists should have the right to say, yeah, you can have all my work. And they should be able to say, “I’ve decided, I don’t want you to have it. I changed my mind. I was crazy last week.” So that should be revocable. 

I don’t, I just don’t, I’m not ridiculous, I don’t think it would ever happen. So I just don’t see it as a viable moral decision for any society. Now, this is the stuff, I’m against assault weapons. I don’t think anything I can say can make them go away. But the fact that some people use them isn’t a reason to suddenly change your opinion about them. 

We’re in this difficult moment here where some people think it’s fine and we’re going to find out how much our art director friends really love us in the next five years, because we love them, we really do. They’ve provided us careers. Me, for 35 years they’ve provided me a career. And I’ve broken my back and my heart and my knuckles for them and I continue to do it. And there’s a whole new generation of people who are willing to do the same thing. So, yeah, I just don’t think it… I think there are things that could work. I think that compensating artists would be a good starter. 

Carl Cleanthes: I’d love to comment on several of those things. On the art director side, I’ve heard several stories and talked to several art directors in my agency as I work in the world. There was a phase where they thought they could just hire prompters to do things for them. And what they found out is they might be able to get a great image very quickly, better than normal, in a faster time. But these people that are prompting that are not already super talented artists that are crazy good at Photoshop and can do all the things that they can make those images without the prompting have no acute control over the details. They can’t do the things they need to do and it just caused them more frustration and delays. And then finally just having to circle back and actually hire an artist. 

So I don’t think we’re going to lose our jobs there. I think technology is going to get better. But again, I think the training and the skill that artists have is going to make us the best at using these technologies. Because the marketplace, unfortunately, is going to demand us to interact with these things on some level. I think there’s going to be new positions where people use, if you’re an in-house artist at a company, I think you don’t have as much say if you’re a freelancer, for example, where you go, hey look, I’m not going to use these things as part of my brand, it’s not who I am, we do it this way. But the fact of the matter is, I know you don’t like the toothpaste out of the tube analogy, but I don’t know how we get back to another place. 

I think the shame methodology is great. I think, you know, coming up with a better ethical data set is something that we should do. There are organizations working on collective licensing. I don’t know if those are solutions. It’s not going to replace, you know, $100,000 a year of income for illustrators, but it’s going to create data sets where people can put their information into it, be given compensation, credit and consent and be paid as it’s licensed out from. And I think organizations will pull that off, but to the same kind of mentality that you’re speaking of, like “hey, people should have shame for using these things”, We collectively can come together and go, “hey, let’s, let’s demand as a people, an ethical data set and companies to use that” and then try to build a movement I think trying to support that. Now it might be woefully optimistic, but I mean, if we’re going to fight the fight, like let’s come up with a solution that might have a chance. I don’t know what the other solutions are other than things that are just wholly built on smaller data sets for local things that companies themselves own entirely. Yeah. 

Sir Michael Allan Galvez: Yeah, so as Ismael, I want to punt this question to you. It’s essentially the same question as everyone, but I just want to frame it in the sense that as someone who runs an AI ethics and research governance consultancy, have you either encountered you know, do you have any ideas about the specific topic of, in your experience, is there that ethical framework that is already in existence or you’ve already maybe worked with? 

Ismael Kherroubi Garcia: The answer’s usually no, either I don’t know about it or it’s not there. 

Sir Michael Allan Galvez: What would you do?

Ismael Kherroubi Garcia: Yeah, I think what’s interesting about what we’re trying to solve here is use AI better if we want to use those tools at all. And that presupposes a sort of top-down approach to policy and governance making. So I would work with an organization and the manager who happens to have a bit of funding and happens to care a bit about responsible AI will say, “okay, we want to come up with a policy” and they’ll say, “great, who else is part of the conversation?” And they’ll say, “well, no one, I’m just hiring you to come in and do it.” And I’ll say, “yeah, no, okay”, I’ll walk away. Like this is a waste of time. Like I’m happy to take the money. I need it, but you’re not gonna get any value for it. If we write a policy together, I deliver three or four presentations, write a policy, put it on their intranet and never talk about it again. 

What I think is missing and going back to the initial question, okay, is there some sort of global policy, is there a sector-wide policy we could create together? Whatever we create has to be done with the input of artists, illustrators, whatever subsection of the artist art sector we want to work with. And the process needs to be inclusive. The people in this room, in the room that I can’t see, I’m sure you’re all very lovely, I don’t know how many people there are, but you all need to have these discussions where you’re able to question what there is already. What’s a LoRA question, the impact of putting more AI generated images out there, understand what all of the copyright and fair use policy conditions that Rebecca pointed to. 

We need a deep understanding of this and generally we can call whatever this understanding is “AI literacy”. Okay, and I genuinely think that I said I’ll say this again, your coming to this event is an interest in improving your AI literacy. The thing with AI literacy, so it is on the one hand no one knows what it is so please don’t ask me any follow-up questions Michael, but secondly, it’s not got to do with AI, it’s got to do with the specific industry sector organization. That’s the conversation I would have. 

Now, the Graphic Artists Guild, you have a statement on responsible AI. Society of Illustrators, have the statement of responsible AI. The next step, if it hasn’t been done already, and I’m not assuming that it hasn’t been done, is to ensure that everyone who comes into the organization, who wants to be part of your conversations, is exposed to that policy and is explained why it matters, why it’s important, but also maybe the policy needs updating as the community grows and changes and shifts. 

So that to me is what might be missing in a question about, again, going to the client that always exists and that’s the easiest one to find: “I want to come up with a policy. It’s just me, the CTO”, and I’m like, “yeah, you’re not of any interest. You’re not going to actually succeed in responsible AI at your organization.” Whereas as the Society of Illustrators, Graphic Artists Guild, us at the Royal Society for the Encouragement of the Art Manufacturers and Commerce. We can bring people together, we can create policies that respond to their interests and defend their interests, and evolve as their interests also evolve. Long-winded, somewhat confusing response. Thanks.

Sir Michael Allan Galvez:  I think that you’ve actually touched, and this is a very good segue to the last very quick question I was going to ask the panel here before we open the floor for questions. So I think the big thing, right? The big question that I kind of said was exceedingly important at the beginning is that we talk about AI literacy, we talk about current policy initiatives, but what can artists as individuals do now, immediately, on their own, to pragmatically move forward and not fall into either a cycle of despair or a certain optimism? Could each of you give us one to three easily actionable items that anyone in this room right now or online can do within the next week or even the next month? Whoever wants to go first on that.

Tim O’Brien: Copyright your work. Pretty easy, believe it or not. It’s pretty easy to do. You get sort of used to doing it. And I made a lot of money defending those copyrights. So it’s worth the effort. You show it, copyright it. Put it online, copyright it. The founders knew it was important for creators, creative people to be able to profit from their creations. It’s part of… the beginning of this country, what they thought would work. And so we’ll see how many of those people in Washington really care about the founders now. 

Support efforts to disrupt generative AI, that means using things like Nightshade and Glaze. Fuck it up a little bit. And add metadata, as I said before. 

And I don’t like the word shame because it’s sort of an aggressive action, but I think we should speak out when you see it being casually used. You’re working, you don’t use the button on Photoshop to generate something. They’ll notice that no one’s using it if we don’t use it. When you use iStockPhoto, don’t generate a photo, just look a little longer for something that’s not perfect. The more we engage with it, the more we casually start using it, that’s the slope. And I’m worried about that. So I think being against it vocally, publicly is where we should always be. That goes for the uncomfortable thing. 

As an organization, if one part of the organization teaches people how to use it, incorporate it into their workflow, and the other part of the organization feels like you should set up guardrails, then you lose your moral authority to do that. Maintain your moral authority by being against it. And it’s not a luddite’s position, it’s a thoughtful position. Those are just some three basic things, and I’m sure we’ll hear some better things. 

Carl Cleanthes:  I have a couple of things. I think the principles of being successful in art has always been about building yourself up and building a brand and making someone enjoy the idea of working with you and getting a piece of your work. Those things are never going to go away. So, the value of someone scraping, let’s say, Tim’s amazing illustrations into a LoRA and generating a new one, that’s not going to resonate with the same value as someone who actually commissioned Tim for a piece of art. So the principles of building yourself up as an artist are not changing and are more important than ever. 

I, also being a huge fan of technology, while I super support copyright, that is the fundamental piece of this, there are some additional things that you can use, things like blockchain technology. I know NFTs themselves have a bad rap, especially in the creative community. I think they’re massively misunderstood, and there’s a lot of grifting that went on in space in the past few years. But the long story short is that you can create an immutable record for fractions of a penny of all of your official works of art on your website and moments. That’s very powerful. And that would stand up very powerfully in court. So those are things that you can use to protect yourself in this space moving forward. And I’m happy to share those thoughts with people and get into the weeds of any of that with anybody at any time. So I’m a huge advocate for that space. 

But yeah, just copyright your work, protect your work, build yourself up like you would. There’s a lot of doom and gloom out there. Just, you gotta stay positive. People do art because they love it and they’re passionate about it. And don’t let this stuff deter you from that. You’ll find a way, if you’re passionate enough and you keep trying, you keep figuring it out and you don’t quit. You can do it. So I don’t want people to not want to be artists because it seems terrible out there. Like you can figure it out and there’s a community that will help you. 

So then join organizations like the Guild, be a part of the Society of Illustrators.. We’re out there banding together with different opinions, trying to make all the things work for each other. Like we’re a community. So that’s a path forward. 

Sir Michael Allan Galvez: Ismael. Do you have anything as I position myself into the audience? 

Ismael Kherroubi Garcia: Of course I do. I always have something. First of all, you already did it. You came to this event. So congratulations. Second thing you’re going to do is share this event once it’s recorded and online. That’s always helpful. Third of all, and emphasizing the really important part of Michael’s question, how do we avoid fear and undue excitement about AI, you need to read up, you need to understand what on earth AI is, you need to find the communities. Yes, to what Carl said, find the artistic communities to improve yourself as an artist. That’s perfect. We need to keep doing that. But also find the communities having meaningful conversations about AI. So today, Responsible AI Networks, Society of Illustrators, Graphic Artists, we’re coming together, we’re making this space. 

But also there’s the AI Now Institute, there’s the Algorithmic Justice League, and there’s the Distributed AI Research Institute. Those are three organizations out of the US. Look them up. AI Now Institute’s report from yesterday was beautiful, and it was about the power dynamics underpinning AI, which are just so much more important than, “Ooh, when did they last update this tool or, can I use it on Google search now?” So yeah, those are my thoughts. 

Rebecca Blake: Yeah, echoing what other folks have said here, I just want to make one little correction. As soon as you put something in tangible form, including as a digital piece of work, it’s copyrighted to you. And fun fact, Tim, the only place in the US Constitution where the word “right” appears is in the passage on copyright. But what you do need to do is register your copyrights with the Copyright Office in order to be able to protect them. And while it is true that blockchain can certainly develop, show an unbeatable record of your ownership, you can’t bring in copyright infringement lawsuit until you can register your works. So do think about registering your works. Don’t get to do all of them. I know it’s expensive. The Graphic Artists Guild is trying to get lower cost options for you, but it is a cornerstone. 

I agree with what everybody else has said here, but I think one of the things that we really need to do as a community is we need to start paying attention to what’s going on and educating ourselves on it. If you want to understand what the copyright issues behind AI are, what the labor issues are behind AI are, go to people who are experts on copyright and labor and are sympathetic to artists’ points of view. Please don’t go to famous artists. Because famous artists are wonderful people, but they very often post incorrect things about copyright. Copyright is complicated. You need to go to the sources that understand it to really understand what’s going on. I have seen artists pick up the bandwagon and promote absolute myths about copyright because of things artists said. It’s incredibly damaging to your standing, especially as a professional community. And I’m saying this as a graphic designer. There are good sources, there are good advocates for us. 

And the next thing we need to do is we need to come together as a community. If there are going to be professional standards and ethical standards, they have to come from us. We are talking about generative AI for images, generative AI for images, not generative AI in general. The musicians have different issues than us. The text and book writers have different issues than us. There are different ways that generative works damage musicians and writers than us. So we have to develop our standards. Sure, there can be global standards for creative industries, but if we’re going to parse it down, and we want to be protected, we have to be the ones to set those standards. 

And we can’t do it unless we understand how generative AI works and unless we understand the legal underpinnings behind our rights. So that’s the really big thing you need to do. And it’s a little bit boring and really sorry, but actually it gets kind of interesting when you look at it. 

Sir Michael Allan Galvez: All right I think I’m going to to follow up on this with questions from the audience. Raise your hand if you have a question, I will come to you. Make sure you tell your name and what your question is. I think I saw yours first. 

Jed: Good evening everyone, my name is Jed. I am an illustrator. I dabble in animation, design, and occasional photography. And I want to say thank you to all of our panelists for giving us this talk. I have a question in my head that I would like to question you with, Carl. This is about a passion project where you are basically gathering your father’s work and are conjuring spirit. So it almost sounds like this seems like a very unusual form of, frankly, a kind of therapy. It almost sounds like art therapy. So all of this discussion of like, we’re losing all of us as artists, losing our value because our work is being stolen, because there’s not an automatic plagiarism machine that we can all access using a button in our apps, which we rely on for our livelihood. With that said, so you’re using that, using it as almost as a form of healing processing. Would you want to folks’ to experience it the way you experienced it? 

Carl Cleanthes: Yeah, the way I’m experiencing it is it’s been a beautiful thing. I mean, it started with a more analog collaboration, right? Like my father died and I think two weeks later, I took one of his old sketches I found and I just made a copy of it and just drew on top of it. I hadn’t done art in like seven years since art school because I built up my studio and hired a bunch of full-time artists. And next thing you know, I’m not even on the art team. So I went on this art healing journey to refine my own creativity. And yeah, it just evolved over the years. 

And then when I found out about LORA models through my deep, long conversations with Rebecca, extensively trying to learn the advocacy space in the arts. (And thank you, Rebecca you’ve been absolutely amazing to be mentored by, I should say.) But yeah, I was like, well, what if I could make new art with my dad, right? So if you checked out my YouTube channel, I’ve been making art Wednesday nights for three years, making collaborative pieces with my dad. And I started to run out of sketches to use with them and I was like maybe I could make some new ones in his style with a LORA model and then I just started experimenting. Its brought m to tears often, the healing journey is powerful. 

I wish people could experience that part of it. It’s weird because there is this part of this stuff thats like soulless and whatever, but like that didn’t feel soulless. It’s weird. I’ll get there one day to be able to talk about this.

Tim O’Brien: Can I say one thing about that? I think that’s quite moving and I, know, sharing that. That’s a personal use, you know, I have no problem with someone doing that. That’s beautiful. But I think if the surviving Beatle children in 10 years decided to start to create a music generator to make new Beatle stuff, I would be furious. It would ruin it for me and I wouldn’t want to hear it. I would hear some of it, but I would feel that way. But if they made it and they were sharing amongst themselves and “Listen, I can hear Dad and all that stuff”. That’s fine. 

But I just feel like we as artists, we bring these things personal to us. We don’t want to see a remix of an author we love. You know, so this is beyond illustration, but this is about creativity and artists and the, you know, if you, Alex, if you wanted to use your father’s art to make something and it was moving for you, would be great. If you started sharing everywhere, I would say, “Man, I love Robert’s work, but…” 

Alex Grossman: You mentioned that, When I started today, I’ve been talking to lot of artists and so forth about my dad’s work and everything. I remember speaking with [unintellibile]… 

Tim O’Brien: Robert Grossman, and this is Alex Grossman, his son.

Alex Grossman: …and he said, “You know, so you’re going to use something, you know, your father’s animations,” he said, “Please do not try to recreate or you know, I mean, do a sort of documentary right now.” People, you know, have said, “Oh, you got to have an animation, got to have new animation.” I’m like, “There’s no way I’m going to try and emulate his stuff or do anything.” And I’m also doing a book on him and I’m surrounded by sketches and all these sorts of things. And it’s interesting because I was kind of moved by your story, Carl, but, you know, I wouldn’t see myself doing that. You know, I mean, I’m trying to kind of further his legacy, but I don’t see taking any of his work and trying to, you know, fuck with it in some sort of way as anything that would be advantageous to me or to anybody out there. So I kind of agree. I mean, well, whole-hardly agree with you, Tim. 

Carl Cleanthes:  I personally just don’t feel like it diminishes the original. If anything it celebrates it and it’s just something different, right? Like I don’t think it takes away if we were gonna make a Beatles album for the… Because it’s out there doesn’t mean the other work isn’t as magical as it ever was. And maybe this feels less quality or less whatever, but like in this case, when I make these pieces, I’m actually collaborating. Like, I’ll take these pieces and do collaging and things of that nature and make a whole new piece on top of it. Like, I haven’t, like, sort of published a bunch of the ones that, like, just generate right out of it. It’s just interesting to see, so. 

Tim O’Brien: I can say that I think that’s beautiful to make something based on your, some you love’s, work. I have a student at Pratt who’s graduated this year who’s made her thesis based on her grandmother’s needle point. It’s beautiful. So I don’t have anything on top of that. We’re talking about the… We’re talking about public things. 

Sir Michael Allan Galvez:  It seems like the station has personal use for commercial use. It varies much more for personal choice. 

Rebecca Blake:I also think what’s really key here is the attribution. Carl is acknowledging that he’s using his father’s work. Tim’s student is acknowledging she’s using work. That’s huge. The thing that’s so dehumanizing with the AI in this generation is people’s work is being taken and sort of munched into the algorithm without permission, without any attribution.

Carl Cleanthes:  If somebody else did that with my art I’d be furious.

Sir Michael Allan Galvez: Just want to bring it back. We have a question. We had a couple of questions in the audience. Sir. 

Steve Brodner: My name is Steve. I’m an illustrator and I teach at a school. I’m very concerned not about people really loving AI because as has been pointed out here tonight, people are not loving the automatic use of the robots in the department. So that doesn’t touch the issue. But what I am concerned about is our kids. Our kids who are being conditioned to find everything quick and easy and everybody can do. And the reason they’re in our school is not because they have been spending their lives doing something because they can enjoy making money. 

I’m concerned that administrations will find it intimidating to hold the line on the end and say to them, you can use this in your free time or when you graduate, any way you like. But when you’re in here, you’re gonna learn skills. And Tim and I are illustrators, which means that you are combining ideas in complicated ways that have to do with experience, emotion, and intellect. A robot has never had an experience in the real world. And as of right now, I need that to, I’d like to have this discussion in the panel of pulling the line in schools and how can we frame the new technology, let’s face it, trillions of dollars that will be poured into this. It’s going to dominate them. in so many parts of the economy. It’s been an upturn, up, up. It’ll destroy everything we know of, including new forms. But in art schools, we tend, we are endangering our culture and the intellectual growth of our kids and our society by taking skills which can be easily done with falling back on the easy. So I’d like to hear what you think. 

Rebecca Blake:  I’d love to answer this. So the thing that gives me a little bit of encouragement is there are some artists who are using AI, generative AI, entirely creating works on, (well, one of these not really), creating works largely on platform. And they’re creating works that have a vision and an integrity and a narrative. Paul Laidler – Tim, I don’t know if you saw his talk at ICON. But I think in order to create works like that, I look at these works, I  actually came across a work on the internet, I go “Oh My God, I bet that was done by Paul” and it was. 

That wasn’t an easy task. It wasn’t an issue of just putting in text prompts and gosh oh golly, this  just popped out. There was, first of all, I’m sure there was an iterative process. But the conceptualizing, the intellectual rigor that went behind the creation of those images, to use generative AI to produce something that has creativity and value, it should not be a shortcut. It’ll be a shortcut for the kid who doesn’t want to learn how to do watercolors, and he wants to use generative AI to spit out something that looks like a watercolor, sure. But to actually use it to create something with integrity, you need that, you need to have that mindset, that creativity. 

Another person is Jessica Helfand, who’s been doing a series of studies of immigrants, laborers, and they’re beautiful. And what she’s done is she’s actually trained a model on, I think she collected a lot of vintage photographs. I don’t know what her generative AI process was, but then the outputs she then recreates in oil paintings, which is this incredibly beautiful series of portraits. And again, you see one at a distance, you know, that’s her, there’s integrity behind it. There’s a vision behind. So done correctly, it should not give people a shortcut. 

Having said that, my tremendous fear is – I think what you touch on – is that students are going to want to use it to not to do the hard labor in developing the skills. And that is a tremendous problem. But the other problem, the other problem I find – and I’m almost done Tim, I promise – the other thing I’m worried about is generative AI is removing the low-hanging fruit, the illustration projects that are the bread and butter of people be fresh out of college. And if that is the case, then who’s going to be able to afford an illustration background, an illustration education?

Because you’re going to be entering into a marketplace where you’re basically going to have to support yourself with some other source of funding until you get to that sweet spot where you are actually able to turn a living. And if that is the case, then does this mean that our illustration students are going to be coming from prosperous backgrounds? And if that is the case, then what does that do to the representation in our illustration student body? Because that’s the seed corn. That’s where our cultural richness comes from. What is it going to do to the ethnic diversity, the educational backgrounds, the economic diversity of our student bodies? Those are the things I do really worry about. 

Tim O’Brien: Okay, sorry. I, like you, Steve, I’ve been teaching a long time, I think 34 years for me. I think you’ve done quite a few more. So I have a good perspective on how hard my students work. Did they work harder in the past? Did they work harder now? My students work really hard now. And the thing that I found very interesting since 2020, is that more of them are wanting to go back to hand skills. They’re feeling anonymous and sort of hidden behind the Procreate and the Photoshop and the onset of easy image generation. They wanted to make it with their own hands. And much to my chagrin, because that just drives me crazy, because I have to teach them how to use things and make gradation by hand. So it’s hard work, but they’re not afraid of the hard work and the work that they’re doing is quite beautiful. So that’s the student part. I’m not worried about the students. 

I’m worried about the schools. The schools have been dangled a very shiny little object. And most of the emails and sign-ins that you see when you teach at an institution is, “we’re having a session teaching you how to incorporate AI into your workflow, teaching you work. How do we incorporate? How do we incorporate?” What’s the incentive for the school? Why do they feel that they have to do that? Well, they’re afraid of being behind the eight ball. They want to promise their parents that their students be prepared. And these are student replacement tools. These are illustration replacement tools. The  schools are trying to teach people to be illustrators.

And so I made my job, and since I stopped being president of this place, to be more involved in school. And I joined the proud senate and the… I joined the AI committee and we had to create some guardrails. And mostly what I did was look at language, how we use language. How do we talk about it? If we say, AI should be used this way, then I insert a line, if you choose to. I made sure that it is still presented as an option. And I never want any language to be put out there and says that it’s assumed that we should and it’s going to be part of your work flow if you choose to. And it has to align with the learning outcomes of a course. If you have a painting course that’s supposed to teach someone how to paint something, we don’t crack out the image generator. If you’re working in your architecture part of the school and you need to generate ideas, they might have a different lens. But anything that teaches students how to make things, it’s not really part of it.

And I don’t really know any students that are very interested in using it. And anyway, most schools have a rule that says you have to attribute. And if you have to attribute the work to AI, it’s another badge of shame. And no one wants to do that. They barely want me to help them with their work. In fact, they might draw on one of their pieces, they hate it all. They want to make their own thing. So I’m not worried about students. I’m worried about the schools and I think the schools are being a little bit dazzled right now and I think things are evening out. There’s fewer signs in the elevators. 

Steve Brodner: Just want to say that I’m in a SVA group. We had a conference in AI. It was all skewed to pro-AI. This was disturbing to me. Our provost seemed squishy that day. But meanwhile, now, because of the conversation we had, there is in the rule book in the bylaws when the kids get introduced into school, “No AI tolerated”. Period. So we had a conversation. At least if this administration stays there, we’re going to be alright. 

Carl Cleanthes: Yeah, I’d like to comment on this as well. I feel like every evolution in technology’s had this this fear, right? You know, when we went from, you know, transfer paper and grease pencil animation to digital animation and computers, there’s a big fear that we’re going to ruin it. And there’s going to be the skills are going to get lost. 

This is a new tool. There’s issues with the data set, for sure; we’ve discussed that pretty extensively. But as ethical data sets come about and they’re being worked on, they’ll be here eventually, I think it’s important that artists learn these tools and the idea that they need to be, it’s gonna become a new craft. It’s not going to replace illustration or painting, just like beautiful graphite drawings weren’t replaced by photography, like we were afraid it was gonna happen when that came through. This is going to be a new type of art. And it isn’t like, I punched in the keyboard, made a rainbow unicorn, like that, you’re not an artist all of a sudden. 

But much the  away that a collage artist makes something transformative, you can make something transformative with AI. I’ve done it personally. I have pieces of art on my website right now that I spent nine, 10 hours conceptualizing using images that I rendered. It’s not, like, it’s not like a big, like I’m not, you know, trying to license them out anywhere or anything, but I’m experimenting because I know this is the economy that I’m in, right? I still have a policy with my business where we produce, we don’t put out any AI generated pixels to any of our clients within my business. But as an individual, like I know this is the economy we’re in, this is the world that’s moving forward. 

I’m up here advocating for doing it the right way and navigating that conversation in the right way and being part of these, you know, like the Graphic Artists Guild and making sure that… you know, there’s a saying in the advocacy space that if you’re not at the table, you’re on the menu. Like we’re at the table. So yeah, I think this is a new tool. 

There will be people in the coming years that become renowned artists that do amazing work that are using generative AI But it isn’t just punching in a prompt. It’s conceptualizing an idea. It’s deep layers of collaging and manipulation. And we’ll see that. Like I’m 100% positive of that. Like, will some people hate it? Yes, they should. I would love that. I want people to be vocal about that, right? Like that’s how culture evolves, thats how technology evolves alongside of it. I love that the Society of Illustrators has a “no AI policy”. We absolutely, they need that. We should keep that here. So yeah.

Rebecca Blake: Just throwing a little dash of cold water on what you just said, Carl, which is just to point out that there actually has been a shift in the labor market. The Association of Illustrators in the UK did a report – they do it by biannually – in the 2023 report, already showed that there was a significant loss of income back then among illustrators due to generative AI. The World Economic Forum came out with their every-other-year report. And in that report, if you went back two years ago, they indicated that graphic design was showing moderate growth. Well, this year it is the 11th most, 11th at highest risk of declining, declining incomes, declining growth because of generative AI. It’s done a hairpin turn. 

So I am not saying that the future is all bleak. But I am saying we are going to have to address the labor issues and the monetizing issues and the collective licensing issues in order to protect our careers so that we can actually learn these tools and become leaders and do everything else we need to do. 

Carl Cleanthes: And the good news is we’re not alone. That’s happening across so many industries, like we talked about earlier. 

Sir Michael Allan Galvez:  Thank you all so much. So, I want to give Ismael final words before we close.

Ismael Kherroubi Garcia: Really briefly, Steve, hi, I can’t see you, that’s fine, thank you for coming. I want you to leave from this event with a bit of a less fear about what’s going on or at the very least a bit more understanding of where that fear comes from if I can. I’m going to play therapist for a second. 

The fear is from two places. Tim’s already pointed to one. Universities and schools are being bombarded by the AI companies. They’re being sold absolutely every generative AI tool. It is brutal. On this, Ben Williamson from the University of Edinburgh on LinkedIn is a great person to follow if that’s something you want to learn more about. The other thing, there have been discounts or free, yeah, some of the AI tools have been given for free to students during exam periods. That’s just a thing that’s happening now, right? So schools and students are being bombarded by this thing. And that’s the source of it. 

Now, I think to Tim and Carl, what they said about, the kids want to get their hands dirty and Carl was speaking about, we need to hone in on our AI, on our artistic skills. And yes, we might or might not use AI in the future, but we need to hone in on our artistic skills first and further. That’s what’s most important. I think the child that makes it through school just blurs and kind of does it and just goes through the whole process, comes to class and stares out the window, but then passes the exam with a, I don’t know how you grade in the US, but with a 50 % mark or whatever. That student’s always going to exist. I’ve never had any interest in that student. And I don’t know how you guys as teachers can actually manage that student. I think you’re all very impressive. 

AI is just another thing that if anything emphasizes some structural issues about education, yes, exams and some of the teaching isn’t meant for every single student. Children will learn in their own different ways. And we need to be, I don’t need to, I don’t know enough about the pedagogy to actually bring anything useful to this conversation. But all I want to say is you are doing an impressive job. And it’s good that we are aware of what AI might mean for the education process, but the education process itself is what you are going to be best at improving. 

So, yeah, like Carl saying, focus on honing in on your artistic skills. I’m going to say whatever you do in life, if you’re a teacher, if you’re an artist, whatever, focus on those skills, on the structural issues that are already in place, because there are plenty already. So let’s do some meaningful work.

Sir Michael Allan Galvez:  Thanks so much Ismael. So the reason I wanted to give them final words is because we are actually going to end this formal panel portion and mingle in the gallery here with our panelists and you can feel free to ask them any further discussion. Ismael of course will not be able to join us so he will be saved by Canada and we’ll be mingling and drinking. But thanks everyone for joining us and thanks for being with us.

Ismael Kherroubi Garcia: Thanks all. Goodbye. 

* Note: Before the advent of generative AI, courts have concluded that copyright protection exists only for works of human authorship, and that has been echoed in the Copyright Office’s guidance on registration. In its guidance on registering the copyright of works created by generative AI, the Copyright Office has stated that works generated via generative AI do not have the human authorship component required to vest the work with copyright. From their guidance: “When an AI technology determines the expressive elements of its output, the generated material is not the product of human authorship. As a result, that material is not protected by copyright…” However, the Copyright Office has also stated that if an AI-generated work contains elements of human authorship, those contributions are copyrightable. To read the Copyright Office’s guidance on registering AI generated works, please visit https://www.copyright.gov/ai/ai_policy_guidance.pdf. 

  ** Note: Rebecca Blake misspoke; she spent 30+ years on the board of the Graphic Artists Guild, not ASCRL.