The (Hidden) Cost of Creative AI

I spent years at the forefront of generative AI adoption, educating creative professionals as well as students at design and art schools, persuading them to at least try AI out and experiment with it playfully. But my enthusiasm often hit the hard wall of mistrust and negative attitude towards anything generated. For many, AI-generated workflows were cheating: inauthentic, lacking craft, and often unsatisfactory in quality. I understood that, and I never forced anyone to use AI if they felt that way.

Nowadays, many of those same people use it daily in their practice.

Generative AI has crept in gradually. What began as casual chats with models like ChatGPT turned into daily brainstorming, while tools like Adobe Creative Cloud quietly folded AI into their workflows. (Sure, you can opt out, but who wouldn’t use the AI-enhanced Remove tool to seamlessly retouch an image?).

It’s hard to resist generating a quick video from a still design, or letting AI speed up the tedious work of collaging and retouching. Even those who don’t intentionally co-create with it now use it in parts of their process—a shift driven largely by market pressure. The logic is brutal: keep up, or get left behind—run over by those who optimize.

As an early adopter and promoter, I however also can’t ignore how the business around generative AI is twisting our sense of what human creativity is truly about. Therefore I would like to take a step back here for a reflection on the myth that AI makes creativity easy and cheaper.

Errors, Explorers, and Early AI Art

Between 2017 and 2021, the creative uses of generative AI weren’t at the centre of attention. It was merely a small group of weirdos spending too much time at their computers—myself included.

It felt like the early days of the internet, or of personal computers: playing with software not designed for visual outcomes, breaking tools just to see what might happen, and treasuring those accidents like shiny little pieces of gold pulled from the mud.

Much of the AI art from this time—like the works of Mario Klingemann, Mike Tyka, Memo Akten, or even my Strange Attractions project—came out of errors and glitches. Artists aestheticized and elevated these accidents into art precisely because they were chosen and captured by human hands. They were moments of deep exploration and interventions into the evolving AI systems. Often working without proper interfaces, artists became explorers venturing into the dark, not knowing what they might encounter—or how it might change them forever.

Strange Attractions (Lenka Hamosova & Pavol Rusnak, 2019)

It’s hard to imagine that such a practice could ever be monetized. And of course it couldn’t. It thrived on disruption and the freedom to go beyond officially designed routes. By nature, it was wild, rough, imperfect, and free.

Back then, many of us dreamed of AI tools becoming more stable, more evolved, producing higher-quality results—maybe even offering better ways of interacting with them. And we got it: first the natural-language interface (prompts), then full UIs in the current generation of AI apps. Suddenly AI was much more approachable, and definitely less weird.

But with the interface came the monthly subscriptions. And with them came the overload of AI-generated content—along with prompt-engineering experts, AI consultants, AI businesses, and even AI automating AI businesses. The profit-oriented perspective of AI business does not influence only the market, but creeps into the minds of the users as well.

But how can creative AI practice stay wild, rough, and free when it’s being commodified alongside our attention spans and personal data?

The Wild West Without the Freedom

Take the strange, double-edged feature most generative AI tools have: the explore feed. On one side, it feels like an echo of the open-source spirit: sharing, community, and building on each other’s work. But in reality people just steal each other's ideas, copy each other's prompts, and claim the outcomes as their own.

If you want to keep your sketches private, you usually have to pay for the most expensive tier. This is often designed only for companies that can afford it. Freelancers and small studios are left with two choices: generate publicly, or invest in local pipelines with their own hardware. But setting up Stable Diffusion or Stream Diffusion locally isn’t cheap: a GPU in the creative sweet spot can easily cost €1,700—2,000, and the setup requires technical skill on top. Apple Silicon offers a slicker entry point—but at a premium price, and it still struggles with heavier generative models. Or you can run your workflows through cloud services like RunDiffusion or Paperspace, where you’re not locked into a single app subscription but you still pay by the hour or by GPU time. In the end, it’s a different kind of meter running—and it doesn’t always come out cheaper.

So what does this mean in practice? Most freelancers and small studios end up relying on online services and AI apps. And this is where the pressure really piles up. Creators often juggle multiple subscriptions—€30, €100, sometimes €300 a month—just for access. Their work is visible the moment it’s generated, traceable in history, free to be copied or remixed. The tools themselves shift under their hands, with no guarantee they’ll work tomorrow, or that files won’t vanish. It’s a Wild West situation, but with no room for the creators to be wild in return. You sign up, you pay regularly, you hand over your data for free—and too often, you’re sold something overhyped, glitchy, and unreliable.

Reality behind the Magic “Generate” Button

While nobody really forces us to use generative AI, our clients now do expect much faster turnarounds. They’ve read about the magic of a one-click ‘generate’ button—they want it and they want it fast.

It’s taken for granted that we’ll keep paying for these AI apps on subscription, yet our hourly rates remain the same. And while trying to keep up with market speed through AI workflows keeps many creators in precarious situations, the value of creative work keeps shrinking.

To outsiders encountering generative AI for the first time, the entire creative process seems too easy—and that perception spreads. The current influx of newcomers into the creative industries—thanks to generative AI’s democratization of media tools—disrupts the industry standards and creates whole new “fast-food” category of creative products. No, they won’t replace professionals with years of training and experience. But their hobby projects and social media content spread a distorted idea of how complex creative work really is, and that does more harm than good.

And as if that weren’t enough, there are other stressors tied to working with AI. It can pull you into hyperstimulation (too many options, too fast), trigger over-identification with outputs (ego highs and lows), and lead to dissociation from your artistic voice if you don’t approach it with a certain level of awareness. These aren’t economic pressures, but they take a real toll on mental and physical wellbeing.

Contrary to the widespread myth that generative AI makes things “easier,” meaningful integration into a creative practice is anything but instant. It requires time for reflection, for testing, for reworking workflows so they actually support an authentic voice. That’s why a proper use of AI shouldn’t be seen as a way to cheapen creative work, but as a premium layer—a carefully tailored, experimental and luxury add-on. When I teach AI integration into creative workflows, I always begin with deep introspection and a detailed audit of each individual practice before stacking AI on top of it. That takes time and dedication. And it’s precisely what the fast-paced AI hype cycle refuses to acknowledge.

Why Human Creativity Still Matters

This time also creates opportunities for reflection. Instead of going for quick wins, let us advocate for human creativity that is embodied, rooted in emotional states generated through the flesh of our bodies.

For centuries, society has fetishized the mind over the body, convinced of this duality. Now we find ourselves in a paradox: we have handed the mind to a synthetic realm, where it can exist on its own and even supersede us in cognitive tasks.

But our existence is not replicable in circuits and silicone. Yes, AI can perform creative tasks now, but never with the depth of human lived experience. A chatbot might mimic empathy, even keep up with the emotional rollercoasters we humans endure, but it will never know what it feels like. Unless it has a body that has lived on this planet with everything it brings, our joy, pain, fear, love encoded viscerally in its physical architecture, it will not create equally to a human.

Artists and creators tune into their environments and the needs of their clients—not only conceptually, but with their whole embodied existence. They alchemize their lived experiences stored in body memory: restoring, reliving, re-feeling, and surfacing them into forms that resonate here and now.

To create as a human is to create with the body—it is impossible otherwise. And that’s our true advantage over artificial intelligence. The familiarity of human concepts we share as individuals and as a society creates depth, infusing meaning into what might otherwise be shallow containers. AI outputs may amaze with polished surfaces, but behind them there is no relatable experience.

There’s a world of difference between a movie rooted in a real human story and one that’s completely generated. If the goal is just entertainment, the AI version might do fine. But if what you need is to be uplifted by someone’s story—to find hope, or to witness another human’s bravery—AI-generated content won’t do the trick.

The same goes for a song, a logo, an article. It’s the difference between soul-cleansing blues and elevator jazz. It’s a logo that carries the vibration of your brand versus a mediocre one that looks like thousands of others. It’s a human voice speaking between the lines versus neatly arranged tokens that offer facts but leave you with the angry sense of wasted time.

That’s what creatives should be valued and paid for: creating work with the potential to change us from the inside out—on a whole-body level, not just in the brain.

But meanwhile, the market keeps pushing forward the automation of creativity. Generative “AI magic” is sold like snake oil, promising instant results from one simple prompt. In reality, it is often just a recombination of pre-selected elements chosen by a model behind the scenes. And now we’re offered full automation—the “AI design agent.” Professionals know this has little in common with a genuine creative process, yet it grows harder to explain this difference to a public massaged by AI creativity propaganda.

Reclaiming the Value of Creativity

So what do we do? One option is watchful waiting—let clients realize for themselves that what they’re buying is low quality, then step back onto the dance floor with our skillset. But that risks long periods of unemployment, or working for unsustainable wages to compete with synthetic outputs. The other option is more proactive: raising awareness, and sharing what it really means to create.

That could mean telling our own stories of how we work. Not just polished craft vignettes on YouTube or Instagram, but poetic documentation of digital workflows, personal rituals, the moments of inspiration and the hours of struggle. Even inviting people to witness the pain behind creation, not just the glamour.

And for those of us co-creating with AI, this means showing that it can be part of a deeper creative practice—when approached consciously, with presence, with embodiment. Not as a shortcut or automation, but as a collaborator that still depends on human intention, intuition, and care.

Think of it as bread. Even in an era of pre-baked, processed baguettes, people still pay a premium for craft sourdough. Cheap bread can be thrown into the oven by anyone and called “baking.” But we know the difference. The same goes for creativity: whether you’re shaping something entirely by hand or co-creating with AI in a conscious, embodied way, what matters is the depth, the care, the human presence baked into it. That’s what makes it nourishing.

---

I’d love to hear how you’re practicing this in your own work—whether with AI, without it, or somewhere in between. Share your stories with me—the more we make these processes visible, the harder it will be for the world to confuse craft with cheap shortcuts.

💡 PS: If you’re interested in documenting and sharing your creative process, take a look at Qtures — a new platform built exactly for that, where workflows are open-sourced and shared among creatives. I love what they’re building, and it might be the right place for your experiments to live.