DJ Pablo - Our first step into AI integration

THE BIG BAD DJ

Working in animation today means facing a rapidly changing landscape. AI tools are becoming part of our workflows — sometimes in obvious ways, sometimes behind the scenes. But alongside the excitement about speed and possibilities, there’s a lot of unease and debate among artists, especially when it comes to the ethics of AI-generated work.

WHAT WE NEEDED

Recently, I ran into a small dilemma that highlights these tensions clearly.

In the interest of the ongoing construction of the Gopnikverse, I wanted to give my day one classic Big Bad DJ Pablo a makeover to align better with the style of my short film, Three Wings. The update needed a quick turnaround, so my resources and time were limited. One big challenge: I needed lots of background elements scrolling past a window — blurred and moving, so details wouldn’t be obvious, but the scene still needed variety and depth.

For Three Wings, I’d worked closely with a talented background artist who created incredible work under my direction. Her style was a key part of the film’s identity. But to commission dozens of new background assets for this update would have been time-consuming and expensive — far beyond the budget for what was essentially filler.

HOW WE GOT IT

So I tried an experiment. I used one of her backgrounds from Three Wings as a prompt for an AI image generator to create simple, minimal filler backgrounds. These assets weren’t meant to replace her art or be featured prominently — just to fill in practical gaps efficiently. The results were a little bit magoo, but totally serviceable. I set about putting them into the scene and soon enough I had a convincing variety of pubs, offies and bookies scrolling past the window. 

WHAT THIS MEANT

But I was cautious. She had previously expressed skepticism about AI, especially concerns around companies scraping artists’ work to train AI models without consent. These worries are widespread in the creative community, from illustrators to animators, and they’re valid. No one wants their work stolen or replicated without permission.

This raised a question I needed to answer: was my use of her art “training” the AI, or was it simply “prompting”? The distinction matters deeply for ethics and legality, and it wasn’t immediately clear to me.

So, I did some research.

TRAINING vs PROMPTING

Training an AI model involves feeding it massive datasets — sometimes millions of images — to teach it patterns, styles, and features. This process happens behind the scenes and can include artworks scraped from the internet, often without the original artists’ knowledge or permission. That’s the root of many artists’ anger and fear: their work is used as raw material for AI learning without any credit or control.

Prompting, by contrast, is different. It means taking a pre-trained AI model — one that has already learned from huge datasets — and giving it a specific input or “prompt” to guide its output. In my case, I used a single image of the background art as a prompt, asking the AI to generate new images inspired by it. The AI isn’t learning or storing her style from this one image to reproduce later; it’s generating brand-new images based on patterns it’s already internalized.

Understanding this distinction was a relief. My use of her art wasn’t feeding the AI’s training data or teaching it her style wholesale. It was more like showing the AI a single reference photo to influence the look of new work. That subtlety is crucial but often lost in conversations about AI ethics. It’s something all of us in animation and art need to understand better as we navigate these new tools.

This technical exploration also reminded me how rapidly AI is evolving, and how the legal and ethical frameworks are still playing catch-up. Many artists worry not just about theft but about the devaluation of skill and craft if AI-generated art floods the market without proper context or credit.

AI IN THE COMMUNITY

Meanwhile, figures like Hayao Miyazaki have spoken out about AI’s potential to erode the “human touch” that defines true artistry. Ghibli fans have seen a surge of AI-generated “Ghibli-style” art online — a strange mix of admiration and unease. It’s thrilling to see beloved aesthetics revisited, but also worrying to see those styles replicated without respect for the original artists.

These contrasting attitudes reveal a real divide in the creative community. On one side, the imperative to protect artists’ rights, their creative labor, and the uniqueness of their styles. On the other, the practical need for faster, cheaper production amid tight budgets and deadlines. AI promises to bridge that gap, but only if used thoughtfully, transparently, and ethically.

In my case, I chose to have an honest conversation with my background artist. I explained what I was doing and why, emphasizing that this was about efficiency for minor filler elements, not replacing her main work. I wanted her blessing before proceeding. Luckily, she was fine with it - Seeing as I’d directed her and paid her for the work, she insisted it was mine to do with as I pleased. Initially that seemed obvious and I wondered if I’d just let anxiety get the better of me. But I’m glad I asked. 

WHAT WE LEARNED

This exchange reinforced something vital: collaboration is more than contracts and rights. It’s about trust, respect, and clear communication — especially as new technology blurs the boundaries of authorship and creative labor.

For those of us working lean, AI can be a powerful ally. It can take care of tedious, repetitive tasks, speed up workflows, and free us to focus on the craft that only humans can bring. But making that partnership work means ensuring everyone involved understands and consents to how AI is used.

If you’re a director or producer, it means checking in with your collaborators before incorporating AI into their work. If you’re an artist, it means setting boundaries but also exploring how AI might support your creativity rather than replace it.

No one has all the answers yet. Technology is moving faster than the conversation around it. But if we want to keep making art that feels human, authentic, and respectful of its creators, transparency and dialogue are our best tools.

At the end of the day, creativity is fundamentally human — whether it comes from pencil and paint, pixels and code, or a blend of both.