Where Aesthetics Meet the Algorithm
A conversation with creative director Gizem Akdağ on Turkish heritage, creative partnership, and style exploration.
By Nando Costa on November, 2025
If you’re paying attention to artists and designers that are deeply and thoroughly exploring the limits of what is possible with generative AI tools, it is likely that you have seen Gizem’s work. She religiously shares her process and output on social, helping people see the possibilities of the technology.
What stands out to me the most in her work is the ability to steer AI models into aesthetics with tactile qualities from paintings, to collage, film photography, and so on.
I hope you enjoy our conversation.
Turkey has this incredible artistic legacy. As a young designer working with AI tools, do you see any connection between these traditional artistic approaches and how you work with generative technology today?
Turkey has an incredibly rich cultural heritage: layers of history, aesthetics, and craftsmanship from many different eras all existing on top of each other. Growing up in a place that literally bridges Europe and Asia means you’re constantly exposed to this mix. Even if you don’t study it formally, it shapes your visual instincts.
I feel that influence in my work today, especially when I’m working with AI. I naturally gravitate toward that multilayered feeling, mixing textures, eras, and styles that normally wouldn’t sit together. And AI is the perfect playground for that. It lets me experiment fast, combine ideas freely, and push visuals into directions I wouldn’t have imagined on my own. I’m not afraid of trying things or blending aesthetics that don’t “traditionally” belong together.
So yes, there’s definitely a connection. It’s not a direct translation of Turkish heritage into my work, but more of an underlying influence, a quiet logic that shapes how I build and refine images.
As a country that sits at the crossroads between Europe and Asia, do you feel that this geographic and cultural position gives Turkish creatives a unique perspective on integrating the latest AI tools with traditional design principles?
Absolutely. Growing up in a country that literally sits between Europe and Asia gives you a built-in sensitivity to multiple aesthetics, multiple cultural references, and multiple visual languages.
In Turkey, you’re constantly surrounded by contrasts, modern and ancient, minimal and ornate, Western and Eastern, and I think that naturally shapes how Turkish creatives approach design.
Outside of your day-to-day experiments, and your design and art sphere, what is the relationship of local Turkish people with AI? Is it as widespread as it seems to be here in the U.S.?
In Turkey, there are actually some amazing AI creators, people producing high-quality, experimental work that I really enjoy following. But outside that art and design world, from what I see, many people already use AI in very practical ways, especially tools like ChatGPT. Even my parents have started asking health-related questions directly to it. So on a day-to-day level, AI is entering people’s lives more naturally than many might expect.
For visual content, the most active and creative adoption I notice is in sports. Many football clubs in Turkey are experimenting with AI for their social media, especially those attention-grabbing short videos. Sports is one of the easiest ways to reach large audiences, so AI spreads quickly in that environment. These striking visuals and quick, engaging edits definitely increase public interest in AI. Even if people aren’t creating with the tools themselves yet, they’re constantly exposed to AI-generated content, and that visibility plays a big role.
Is it fair to describe your design background as Brand Design? If so, how do you see that kind of work evolving in the coming years?
I studied architecture, but after graduating I moved fully into Brand Design, and that’s where I’ve built my entire career. I’ve always been interested in how brands express themselves visually, and creating logos, systems, and the stories around them has been one of the parts I enjoy the most.
What’s exciting now is how fast branding is evolving with AI. Before AI, brand identity was built mostly through more static assets, a logo, a color palette, a set of guidelines. The visual language was defined, but it was relatively fixed.
With AI, especially with Midjourney’s style references, we can now explore ‘worlds’. We’re not limited to one static style anymore. We can design richer, more flexible visual systems. Iteration is incredibly fast, and brands can shift their tone depending on the platform or context while still feeling consistent.
Right now, most of my work with brands focuses on this AI layer. I’ve spent almost two years exploring Midjourney styles, blending srefs, and developing custom visual languages for different brands. I build style systems and prompt formulas that allow teams to generate their own imagery in a consistent, recognizable way.
Brand design is becoming more dynamic, more fluid, and more collaborative, and it’s exciting to help shape what that next phase looks like.
Can you describe what you do at Perplexity?
I’m a Creative Ambassador at Perplexity, which means I get to collaborate closely with their incredible creative team. I create images and videos for their products, like Comet, as well as for projects such as PPLX Radio, and I work with the team to explore new visual styles, formats, and possibilities.
It’s a very collaborative process. We experiment with different aesthetics, push what’s possible with AI imagery and video, and look for fresh ways to express Perplexity’s identity. It’s one of those roles where you feel like you’re contributing to something genuinely forward-looking, which is what makes it exciting.
What are your thoughts on the fact that the models we all use today were built on billions of artworks and photographs created by people all around the world, some of which took years to create, and are now reduced to a miniscule data point within these tools?
It’s not something I can personally control, so I don’t believe that simply opposing it would change much. What I can control is how I work. For example, I avoid using the names and works of artists in my prompts. I don’t reference artists directly, and I prefer to build my own visual language instead of borrowing someone else’s.
And as you mentioned, within these large models each artwork becomes a miniscule data point. Unless you intentionally prompt a very direct reference, it’s almost impossible for a model to replicate a specific piece. That doesn’t diminish the value or importance of the original works. Those artworks remain unique.
Midjourney is quite an accessible tool, but would you please share more about your process? Are there nuances about how you prefer to work with the tool that you feel sets your work apart?
I create artworks with AI, I explore different tools constantly, and I also work with brands on AI imagery and video, so each area has its own workflow. But Midjourney, specifically, has a very special place in my process. It’s my favorite tool for pure creativity.
When it comes to Midjourney, I would describe myself as a style explorer. I love discovering different aesthetics, imagining new worlds, and building small narratives around them. For almost two years now, I’ve been deeply exploring Midjourney’s style references. I’ve built a huge sref library, but it’s not just a collection I scroll through. I study each style. I try to understand its DNA: its textures, color behavior, and composition tendencies. Once I understand a style well enough, I treat it almost like paint. That’s when the fun part starts. I mix styles together, experiment with blends, and try to create completely new aesthetics that didn’t exist before. After I develop a style, I explore what kinds of worlds, characters, or compositions it expresses best. I run many iterations to understand its strengths and boundaries.
So for me, it’s never “write a prompt, get an image.” It’s a much more layered, iterative process, closer to world-building than simple prompting. And I genuinely enjoy every part of it.
Your AI images often have a distinctive illustrative quality to them. At first glance, someone might be convinced it had been made by a person rather than an algorithm. From what I tell, there is intentionality behind them.
From your posts on Generative Art, you appear to describe AI as a "creative partner" rather than just a tool. What does that partnership actually look like in your day-to-day work? How do you decide when to lead and when to let the AI surprise you when making your images?
To be honest, when you work with AI, the final image is never only your creativity. Even with a very detailed prompt, the model always adds its own touch. Especially with Midjourney. It has a level of creativity that feels very different from other tools, and I actually enjoy working with that. It’s one of the reasons I describe it as a creative partner.
Over time, I’ve learned how to guide that creativity. Sometimes I use very simple prompts just to see where the model wants to go, and those unexpected directions can be the start of a whole new idea. In those moments, I let the AI lead, because surprise is part of the process.
But for specific projects where I need a clear, defined outcome, I do the opposite. I minimize the model’s creative freedom and structure the prompt in a very controlled way. In those cases, I lead completely, the AI is just helping me reach a precise visual goal.
So it’s a balance. Some days I follow the AI’s creativity, and other days I shape it tightly. That flexibility is what makes the partnership interesting.
Can you share with me ways in which your collaboration with other people has changed since you’ve begun creating more with AI?
My collaborations have changed quite a lot since I started working more with AI. Before, my focus in branding was mainly on logos and visual identity systems. Now I’m creating entire worlds, and that shift has opened up many new types of partnerships.
I work with companies on AI-generated imagery and video, building visual systems and concepts that didn’t really exist in my workflow before. At the same time, I’ve also started working more with artists, especially on projects like album covers, which require a more narrative or atmospheric approach.
Another new layer is the community aspect. On platforms like Instagram, I’ve been doing collaborations with other AI artists by blending our styles together. It’s something that wasn’t even possible a few years ago, and it’s genuinely inspiring to see how two visual languages can merge into something new.
And I think we’re still at the very beginning. There will be completely new forms of collaboration that we can’t even imagine yet. AI is creating new creative roles and new ways of working together, and that’s one of the most exciting parts of this whole space.
Do you think there will be a time when collaboration between people will be less common, and instead there will be more mediation between agents to get something done?
What about Creativity? After fully embracing and integrating AI into your workflow, what has changed about how you think about it?
AI has actually given me more space to be creative. Before, a lot of my time went into the labor-intensive parts of design. Now I can let AI handle those steps and focus my energy on exploration, experimentation, and building ideas. It allows me to work on more diverse projects and push concepts much further than I could before.
As for the future, it’s really hard to predict where collaboration will go. Maybe agents will take over some of the operational communication, and that’s fine, but I don’t think the human element disappears. The “human touch” will always matter. Our taste, our intuition, our ability to choose what feels right, that’s not something you can fully automate.
So even if AI systems start collaborating with each other in the background, people will still be at the center of defining the vision, the intent, and the emotional direction of the work.
What would you say to those that have held back their enthusiasm for AI-aided art?
I think if people try creating visuals with AI, they’ll realize it’s not as effortless as it might look from the outside. A lot of skepticism comes from not seeing the work behind it. The creative process is not easy.
We all have phones and we all take photos, but that doesn’t automatically make us photographers or artists. AI is similar. Making compelling work with it still requires a good eye, a strong sense of aesthetics, an understanding of the tools, and the ability to build a story. There’s craft involved , it’s just a different kind of craft.
So my message would be: try it before deciding what it is. You might be surprised by how much of yourself you actually need to bring into the process.
Have you considered how different the life of a future art, design or photography student might be? Do you have any advice for those starting a new career?
They’ll grow up with AI as a natural part of their toolbox, not something new or intimidating, but something as normal as Photoshop or a camera. That means they’ll be able to experiment faster, iterate more, and focus more on ideas rather than the slow technical steps.
But the core of creativity won’t change. Having a good eye, a sense of taste, curiosity, and the ability to tell a story will still matter just as much, maybe even more. The tools evolve, but the intention behind the work is still human.
My biggest advice is to stay open and not be afraid of exploring new tools. Don’t limit yourself to one discipline. Learn how to mix traditional skills with new technologies. The people who will thrive are the ones who treat AI as a collaborator, not competition, and who use it to expand their imagination rather than replace it.
Focus on developing your perspective, your point of view. Because in a world where everyone has access to powerful tools, taste becomes your real superpower.
Comments
Links