Yes, there’s a problem, but It’s not just about AI.
I’ve heard passionate arguments against AI usage from fellow artists. I’ve also read in detail about the lawsuits filed by creators against companies who used their work without permission. I agree that this is wrong and that it has hurt the true validity of the tools. These concerns are real, and they’ve shaped how I approach the technology.
A central question remains: if we could make the source imagery for AI training completely copyright-free and ethical, would that actually end the argument over the use of these tools in art making? Or is the real issue an underlying belief in purity in the creative process? As a graphic designer, I know that purity in creation was disrupted long before AI ever existed.
From Generative Art to Generative AI
I was exhilarated the first time I started using AI in my art. I’ve been working as a generative artist since the late 90s, so I’ve seen a lot of shifts in how tech intersects with creativity. Back in the early 2000s, I was building generative art installations that used text, images, and sound. I was dreaming in code, digital sensors, and databases as these were the core elements to create art-making robots. I was even invited to exhibit my projects in the US and international media arts biennales in Split, Croatia and Wroclaw, Poland. But this new wave of AI tools? It felt like a leap. A serious one.
When the most recent wave of AI tools burst onto the scene in 2022, I found myself re-reading Walter Benjamin’s essay The Work of Art in the Age of Mechanical Reproduction and picking up Lev Manovich’s article Who is an Author in the Age of AI? for the first time. They both helped give me clarity on the debate developing. The first was written over a century ago, and the second in the present moment, but both challenged me to think beyond the surface of the debate.
That said, I didn’t jump in without questions. I was skeptical about where the image data was coming from. Most of the image models were trained on LAION400M, a huge dataset scraped from the internet. It was meant for research, not commercial use. As an artist, I care deeply about copyright and creative ownership. That part bothered me.
But the power of the tool was undeniable. AI helps me iterate quickly. It pushes my image-making forward and challenges me to try things I wouldn’t have done on my own. New poses, wild color combinations, unusual compositions. Sometimes, I don’t even recognize what I’m capable of until I see what AI reflects back at me.
Through my experiments and explorations, it became very true that it’s not about replacing my work at all. It’s about expanding it.
AI as a Catalyst and Sharpener for Creativity
One of the things I love most about AI is how it helps me start. Sometimes, it’s like having an oracle. I might not know exactly where a project is headed, but I can toss in a few ideas and see what comes back. That response helps me clarify what I want. Or what I don’t want. And that’s part of the creative process, too. The ability to think divergently and then convergently is the true poetry of the creative process and having a “helper” in that process allows greater tracking and reflection on the process.
I use Stable Diffusion most often because it’s open-source and highly controllable. I like that I can run it locally on my machine without paying for credits or cloud storage. Not having a paywall gives me the freedom to really dig deep. I can generate a hundred versions of an idea, explore unexpected paths, and move fast without overthinking costs.
There was one recent project that brought this all into focus. A client in Houston wanted a mural with about 25 different visual elements. Honestly, it was overwhelming. I started by asking AI to look for patterns in the list text, riffing with the chatbot to explore visual ways to combine the many items into one space. The ideas that chat suggested, but it did mention the word surrealism that made me think of Salvador Dali’s haunting landscapes. That was it! A landscape like Dali’s is a place where all the elements could exist together logically. A dreamscape, so to speak. That unlocked the whole thing. Without AI, I might’ve taken much longer to get there.
Most AI images aren’t good. I’d say maybe 10 percent are worth a second look. I know I’ve hit something useful when it meets my “GE” standard: Good Enough to move forward. I’m looking for strong composition and clear visual hierarchy. Everything else can be worked on later. But if the image holds space in a striking way, I’ll keep going.
AI for Process, Not Just Product: An Ethical Approach
People often think of AI as a tool for generating a final image. That’s not how I use it. For me, AI is most powerful when it supports the process. It helps me evaluate, brainstorm, and reflect. Sometimes, I upload a rough idea or a brain dump and use a chatbot to ask questions or poke holes in it. That outside perspective—fast, responsive, and nonjudgmental—is gold.
If I’m stuck, I might ask the AI to generate some moodboards or rough compositions. I’m not expecting polished work. I’m looking for sparks. A direction to follow or a problem to solve. It’s the same way I’d sketch a dozen thumbnails on paper.
And no, I don’t fall in love with the first cool thing AI spits out. That novelty wore off fast. I’ve trained myself to be curatorial. Most results don’t hit the mark. But knowing I can always make more helps me stay loose. I push ideas until they’re solid.
One piece that stands out was an illustration I made for a Torah portion about Pharaoh’s dream. I had drawn a grim-looking cow skull and was thinking of placing it in a field of wheat. Then, the AI surprised me. It created a cow skull made out of wheat. That twist was mysterious and totally unexpected—perfect for the surreal nature of a dream. I never would have gone there on my own. But once I saw it, I knew exactly where to take it.
AI Won’t Replace Me. It Will Refine Me.
As an educator, I’ve brought AI into the classroom not just as a tool but as a way to help students understand how creativity works. I show them how to use it to ideate, test ideas, and refine their thinking. We do in-class exercises where students generate images or prompts and then share their chatlogs with me. It gives me a real window into how they’re thinking—and how they’re growing.
Some students are skeptical at first. Others dive in headfirst and sometimes know more than I do about certain tools. The ones who get it quickly start using AI not to shortcut the work but to deepen it. They realize that it’s not about letting AI do the thinking. It’s about using AI to push your thinking further.
I’ve even used AI as part of a critique. One time, I had students feed their near-final projects into ChatGPT and ask for feedback. With the right prompts, the feedback was surprisingly thoughtful. Not perfect. But useful. It opened a door for them to reflect and iterate in ways they hadn’t before and to be critical of comments that didn’t in fact, help improve their work.
What AI reveals, I think, is that creativity isn’t just about generating something new. It’s about discovering connections, asking better questions, and recognizing what’s missing. AI isn’t great at originality on its own, but it’s fantastic at remixing and showing what’s possible. It’s like a mirror that reflects potential back at you. I’ve worked closely with a few creative development systems like the Double Diamond and SCAMPER and I can say with confidence: AI can support both divergent and convergent thinking, especially when used intentionally.
Originality: The Collage Conversation
This comes up a lot: “Isn’t AI just stealing?” And my response usually starts with this: what about collage?
We’ve accepted collage as a legitimate art form for over a century. Artists like Hannah Höch, Romare Bearden, and Robert Rauschenberg all used found images, many of them copyrighted. They cut, glued, layered, and remixed to create something new. If we call that art, why are we drawing the line at AI?
To me, AI-generated images are collage-like. The human prompts them with intention. The AI recombines things based on patterns it has learned. The process is digital, but the creative act is still there. Cutting and pasting by hand doesn’t make something inherently more authentic. It’s the idea behind the work that matters.
Now, I don’t ignore the legal and ethical side of this. Most major AI image models are trained on datasets built from scraped web images, and that’s a problem. I’ve been exploring more ethically sourced options. For example, Adobe’s Firefly and Shutterstock’s model are trained on licensed stock images. Even better, I recently started working with a model called PixelDust. It’s a rebuild of Stable Diffusion, but trained only on public domain and Creative Commons Zero (CC0) images—think Wikipedia, museum archives, and open repositories. While it’s the closest public domain model out there, it still is not 100% certain it is copyright-free.
I fine-tuned that model using 380 of my own original works. That means when I prompt it now, it generates images in my style using my visual language. It’s still collaborative, but it feels more personal. And the results have seriously improved my ideation speed and image quality.
There’s a difference between copying and remixing. Collage artists have done it forever. Musicians sample. Writers quote. AI might be new, but it fits within a long tradition of borrowing, blending, and transforming. What complicates the conversation is “style”. People think style is protected by copyright, but it’s not. Only specific works are. So, while artists may be known for their style, that alone doesn’t make it off-limits.
Yes, people have questioned the validity of my AI-assisted work. When that happens, I explain. I describe how I use AI for ideation, how I fine-tune models on my own work, and how that affects the output. Once people understand that I’m building on my own images and ideas, they usually come around.
Co-Creating with the Machine: How AI Refines My Process
Over time, I’ve started experimenting with creating a kind of AI version of myself. Not in a sci-fi clone kind of way, but as a tool trained to think and see more like me. I fine-tuned a model using 360 of my own artworks, each paired with carefully written prompts. That way, when I generate new images, they come out in my visual language, not someone else’s.
I also use a tool called ControlNet. It lets me upload sketches or basic compositions, and then the AI fills in the style and detail. This setup allows me to keep control over layout and flow while still tapping into the speed and surprise of the AI. It doesn’t always work the first time, and it can be a long back-and-forth, but the results are worth it.
Eventually, I’d love to have a copyright-safe, fully custom model that supports my entire process. The goal isn’t automation for the sake of ease. I want to hand off the repetitive, procedural stuff so I can stay focused on creativity, strategy, and ideas.
And no, I don’t want my AI self to be autonomous. That would defeat the point. I’m the creative leader here. The AI is my partner. It helps me explore, test, and refine, but I make the final call.
I’ve also made peace with the idea of my style being encoded. I’ve been an illustrator long enough to know that you don’t really “own” a style. My style is a blend of influences I’ve absorbed over the years, and it’s always evolving. As a professional, I’ve had to learn multiple styles just to stay competitive. So, no, I don’t see style as sacred. It’s the ideas and the content that matter most to me.
Rethinking and Remixing Creativity with AI
My relationship with AI has changed a lot since I started. At first, I believed the hype. I thought it would be a job killer that could replace me. But as I worked with it more, I realized its limitations. It isn’t a one-click creative solution. It’s a tool that depends on my input, my ideas, and my vision. It helps me move faster and reflect more deeply, but it doesn’t do the thinking or the feeling for me.
I’ve come to believe that AI isn’t replacing creativity. It’s revealing it. It shows us how we think, where we hesitate, and what we ignore. It challenges the old myths that artists work in isolation, drawing purely from inspiration or talent. That myth never held true for working designers and educators like me. And it definitely doesn’t reflect how creativity works in the real world.
Still, I respect the artists who are hesitant or resistant. I’ve listened to powerful critiques and concerns. The lawsuits over unauthorized dataset usage raise important ethical and legal questions. And they should. If we can’t build these tools on ethically sourced, copyright-free content, then we have no foundation to build from. But if we can create models trained on ethically gathered images, then we should be having different conversations. One would be about practice. Another about process. We’d also be talking about expanding what it means to be creative. Instead, we’re stuck in echo chamber-like debates with half-truths and misunderstandings.
AI is not a threat to purity in art because that purity never really existed. From collage to sampling to appropriation, art has always thrived on remix. This is what Benjamin meant when he spoke of the “aura” of artworks over a hundred years ago. Reproduction changes the way we relate to art, but it doesn’t remove its meaning. It shifts the space where meaning happens.
So I use AI not because it replaces me but because it helps me be more of who I already am. A generative artist. A question asker. A teacher. A remix thinker. A designer trained in collaboration, systems, and complexity. AI is now a part of that system. And I welcome it, carefully and critically, into my process.
The tech will keep evolving. But the core of creativity, being curiosity, play, rigor, surprise, and reflection, has not changed. AI just gives us more ways to explore it.