AI Artist KEZIAI Talks ‘Intentional AI,’ Impossible Cinema and the Coming “Imagination Age”

…I’ve always embraced new technology. Working with AI reminds me of creating video art in the 90s because of its experimental nature. I’ve shot on a massive variety of formats over the years: SVHS, VHS, DigiBeta, Beta, 8mm, 16mm, 35mm. I bring from traditional filmmaking a love of the craft and a deep understanding of film language: storytelling, performance, concept, visual imagination and technical curiosity. All of that translates directly into AI. I think adaptability, critical thinking and a hunger to keep learning are valuable skills today. This emerging technology is a revolution; the impacts it will have are bigger than the Industrial Revolution. In terms of creativity, it democratizes learning. With AI you can go down any rabbit hole, learn whatever you like, go deeper, move faster. It is a remarkable time to be a creative…

Wehad the pleasure of talking with Kezia Barnett aka KEZIAI, a woman who is currently standing at the messy, flickering intersection of art and the digital frontier. To look at Barnett’s life is to see a series of vivid frames, much like the film rolls she used to be given as a child for Christmas in New Zealand. Long before she was navigating the complex nodes of artificial intelligence, she was a girl growing up in a creative and literary family by the sea. Her childhood sounds like a scene from an indie film, centered around a garden shed that did double duty as a darkroom, an art studio, and a storage space for the costumes of Red Mole, an avant-garde theater company.

The roots of her artistic identity run deep through a family tree of creators. Her mother was an artist and lecturer in art and design who took up acting and modeling in her 70s. Her father was a book publisher, editor and artist, and her grandmother held the largest costume collection in the country. “That was very inspiring,” Barnett notes, looking back at a youth spent surrounded by the raw material of transformation. By the time the 90s rolled around, she was studying Fine Arts and majoring in Intermedia Studies. The term, coined by Fluxus artist Dick Higgins, refers to work that lives in the cracks between established mediums. It was a philosophy that would eventually define her entire career, though she didn’t know then that the medium she was destined for hadn’t even been invented yet.

Since 2004, Barnett lived the full-throttle life of a commercial director, traveling the globe, and calling the shots on massive sets. But the physical world eventually pushed back. A struggle with chronic illness took traditional production away from her, effectively ending the life she had known. The physical demands of a film set became an impossibility. For many, that would have been the final credits, but for Barnett, it was a redirection.

“I discovered AI in mid-2022 and instantly felt the potential and the freedom to create directly from imagination,” she says. For her, this technology wasn’t a replacement for the human touch; it was a rescue mission. “In a way, AI saved my life and enabled me to create again. It’s an immediate tap from your imagination straight through to creating.” She describes her current work as “AI cinema”. She views the technology as “a camera for the not-yet-imagined,” a medium that allows her to bypass the physical limitations of a “lens to capture what exists only as thoughts”.

Her process is not typing a prompt and walking away, but a high-tech collaboration. She uses node-based programs like ComfyUI, setting up intricate pipelines and workflows to maintain consistency in her characters. She calls this “Intentional AI.” She isn’t just looking for a cool image; she is looking for a way to communicate emotion through a machine. “You collaborate with the machine, but the essence or the artistry is channeled directly through the AI,” she explains. This approach allows her to create “impossible transitions,” like spinning 360 degrees while traveling through a body of water and into a character’s internal thoughts. In one piece, she directed a dance performance where every dancer’s head was on fire — a feat that would be a logistical nightmare in traditional filmmaking but becomes a playground in the AI space.

Barnett is quick to defend the medium against those who label AI art as mere “slop.” To her, the value of the work doesn’t come from how easy it can be to generate an image, but from the idea and the intent behind it. “I have worked in most mediums in my life, and to me, it’s a catalyst”. She sees the current era as a “creative renaissance” where the barriers between an idea and an image have been torn down. She compares the current state of AI to the early days of film, when directors simply pointed cameras at theater stages because they hadn’t yet learned the language of the new medium. “If the mechanical age let us reproduce the world, the AI age lets us imagine it,” she says.

Her work isn’t limited to the moving image. She spans the distance from the truly analog to the hyper-digital: synthography, AI cinema, music. For Barnett, AI cinema is the through-line, not the boundary. In 2024, her work was shown at the Saatchi Gallery, where she displayed AI-generated stills as archival metal prints, giving the digital files a depth that mimicked cibachrome slide prints. By upscaling the images multiple times, she created scenes within scenes. She also dives into AI music, drawing on the electronic music studies from her art school days to brief the machine on emotions and feelings, much like she once briefed human composers.

Despite the fears many artists have about being replaced by algorithms, Barnett remains a lifelong learner and a staunch optimist. She believes that traditional artists have an advantage in this new world because they understand the language of light, movement, and storytelling. “Adaptability is a valuable skill today,” she says, noting that this revolution is bigger than the Industrial Revolution. For her, the goal isn’t to compete with the machine, but to use it to “imagine more bravely than we ever have.”

When asked for the keys to a successful career, Barnett points to a philosophy of persistence. She believes in problem-solving, refusing to be boxed into one medium, and, perhaps most importantly, embracing failure. She quotes Samuel Beckett’s famous line: “Ever tried. Ever failed. No matter. Try again. Fail again. Fail better.” In the world of AI, where an unexpected result is just a reason to adjust and generate again, this “iterative failure” is the heartbeat of the work. She encourages new artists not to wait to create, and to make what they’ve always wanted to.

As our interview wound down, Barnett described it simply: “We are flying the plane as we build it.” The legalities of AI music and the ethics of digital models are still in a state of flux, but she remains focused on the human element. “I am interested in creating work with meaning that resonates with people and possesses humanity,” she concludes. For Barnett, the “Imagination Age” isn’t about the end of human creativity; it’s the moment where “the distance between the dreamed and the made collapses”.

Yitzi: Kezia Barnett, it’s such a delight to meet you. Before we dive deep and talk about your amazing work, our readers would love to learn about Kezia Barnett’s personal origin story. Can you share with us the story of your childhood and how you grew up, and particularly the seeds and genesis for all the amazing work that has come since then?

KEZIAI: Thank you so much for having me. I grew up in a creative and literary family by the sea in New Zealand, and I’ve taken photographs for as long as I can remember, excited to get film rolls for Christmas presents every year. In my teens, we had a dark room in the garden shed, which also housed Red Mole’s costumes. They were an avant-garde theater company. My mother is an artist; she became a senior lecturer in art and design and started modeling in her 70s. My father was a book editor and artist. My grandmother was a costume designer and had the largest costume collection in the country. That was very inspiring. I went on to study Fine Arts in the 90s, majoring in Intermedia Studies, which is a term coined by Dick Higgins, a Fluxus artist and co-founder in the 60s, for work that lives between the mediums.

Yitzi: Tell us the next chapter. Tell us how you evolved from there to being this avant-garde AI artist.

KEZIAI: After art school in the 90s, I was directing commercials, making photographs, and enjoying the zing of the zeitgeist. It all converged with AI. Intermedia is about the spaces between art forms, not quite belonging to any single medium. That’s still where I work. AI doesn’t resolve this instability; it radicalizes it.

Since 2004, I had been traveling around the world directing commercials when chronic illness took traditional production away; that life stopped. I discovered AI in mid-2022 and instantly felt the potential and the freedom to create directly from imagination. It removes the barriers between idea and image, allowing impossible shots from my dreams, or from your dreams. In a way, AI saved my life and enabled me to create again. It’s an immediate tap from your mind straight through to creating. I feel like almost everything I’ve done so far has been preparation for a medium that didn’t exist yet.

Yitzi: There’s a saying that no is not rejection, but redirection. Do you have a story about when you got no to an opportunity, but that led to an unexpected, even better opportunity or an unexpected blessing?

KEZIAI: That is very interesting, because I do feel like a whole lot of the creative process is about problem-solving. When you have restrictions, it brings out more or something different. I made a short film for Straight 8. You are given one roll of 8mm film and you shoot everything in-camera on that one roll. They develop it, you give them the sound and the film, and they put it together. The first time you see it is at the screening. You can’t see the footage, you can’t grade it, and you can’t do post-production. It actually led to a fascinating project. I thought, “Well, what can I do that can only be done once?”

Yitzi: Let’s talk about AI. Could you describe your unique body of work in your own words? When people hear AI art, they usually think of images made in ChatGPT. How would you describe your work?

KEZIAI: My work spans AI cinema and experimental art, but what defines both is intentionality: Intentional AI. Not generating images, directing them. I bring the same intentionality and instinct I’d bring to a live-action set — emotional arc, pacing, mise en scène.

I am interested in surreal and metaphorical thresholds: liminal states.

I think of AI as a camera for the not-yet-imagined, and cinematic direction for the imagination age. Prompting is coding with language, language as the syntax for an image. It’s a process of leaning into the AI, observing what it gives back, then guiding it toward something that comes directly from within. You collaborate with the machine, and the artistry is channeled directly through it. It’s a feedback loop, collaborative in the deepest sense: you direct, and the AI expands what’s possible. It requires a particular combination — creative vision and technical aptitude. You start with the feeling or the idea and the work tells you what it wants to be.

Yitzi: What I’m hearing you say is that you focus on surreal or non-realistic imagery and video. Therefore, you are able to go from your imagination to the medium in a way that normally couldn’t be done. Am I stating that correctly?

KEZIAI: Yes, and I also really enjoy the realistic and the cinematic. But what excites me most is what you can’t do with other mediums, because it’s an entirely new medium itself. We’re inventing a whole new language, so we get to express ourselves in a different way. Every medium finds what only it can do. AI is finding that now. I bring everything I’ve learned from previous mediums into AI — it feels like a convergence of everything. I think it’s a creative accelerator. Liberating. I am hoping that we have a creative renaissance based on this technology.

Yitzi: What I’m getting from what you’re saying is regarding the standard critique where people say AI is slop. My feeling is that they are usually referring to the uncanny valley. But when you try to make it specifically with an artistic flavor, like a cartoon or something surreal, they don’t say that it’s slop. They see the creativity in it. You’ve gone way past that issue because you’re genuinely creating art. It’s creativity, not an easy way out.

KEZIAI: If anybody can create anything, what are you going to create? It comes down to the meaning, the idea, and the intent behind the work. I have worked in most mediums in my life — to me this is a catalyst. That is the space I work in: collaborating with the technology, experimenting, finding ways to communicate in a human, emotional way.

Yitzi: Can you give an example of what it means to create a video that couldn’t be done with other means? What is an example of a shot that you could capture with AI that you couldn’t do otherwise?

KEZIAI: I like creating impossible transitions. For instance, you might be on somebody’s face, then transition to a wide shot of dancers in a way no traditional camera could achieve. Even with the most advanced technology, you couldn’t do it. Spinning 360 degrees while going into their eye, or traveling through a body of water into their internal thoughts and back out again. I love the idea that you can go internal and external seamlessly.

Yitzi: Are there other examples of scenes that you can create now that were impossible before?

KEZIAI: You can extend what you can create with visual effects, and take them much further. I created a dance piece where everyone’s heads were on fire. In traditional filmmaking, that would have been a huge undertaking. Things that were simple become hard, and things that were hard become simple with AI. It‘s a different medium entirely.

With photography we captured light, with film we captured movement, and with AI we capture imagination. Every medium begins by mimicking the last. When film was invented, people pointed the camera at the stage because they didn’t yet understand the language of cinema. Now we point AI at film for the same reason. Early film copied theater, and early AI copies cinema. If the mechanical age let us reproduce the world, the AI age lets us imagine it.

It is made for the curious and for people who want to learn and evolve constantly.

Yitzi: Are you able to make feature-length films with AI now? Something like 90 or 120 minutes?

KEZIAI: The technology is there. You can control the elements of consistency needed to establish a throughline and create feature films. I’ve produced pieces that are a couple of minutes long, and that could be extended to a feature film with the right resources.

Yitzi: Have you been able to solve the consistency problem? When I play around with it, it’s hard to keep the same character features across multiple iterations without them ending up looking different.

KEZIAI: You can do quite a lot with consistency now. I work with node-based programs like ComfyUI. You can set up pipelines and workflows that maintain consistency and use character references throughout. I made ‘Casting Anna’, at the cusp of being able to do that, exploring what it means to cast an AI character, the challenges of being ‘too digital’ or lacking consistency. Now you really can achieve it, using character reference and replacement techniques across different scenes. It is not a simple process yet.

Yitzi: I hear you use ComfyUI. Do you use that on your own computer rig, or do you use cloud-based setups?

KEZIAI: I have a private cloud-based setup with all my custom nodes and models so I can use the full range of tools and techniques. I work across various setups, ranging from private, ethical models to off-the-shelf options, choosing based on what the project needs, because every model gives you something different. Part of the artistry is knowing which techniques and models to use for which purpose. I often do what they call kitbashing — going from one model to the next to get exactly what I want out of the AI. Finding your unique way of speaking with the AI and establishing your workflow is where the artistry lives.

Yitzi: For someone who is starting out and doesn’t want to just use basic video or text fields, are there secret models you would recommend that they might not be aware of?

KEZIAI: I would recommend ComfyUI. It can be intimidating because it’s node-based, but it allows for control and real freedom. There are a lot more templates to start with now, making it more accessible. It reminds me of working with color grading and post-production, so I can translate those skills. I also work in DaVinci Resolve, which uses a node-based system as well.

I also like using off-the-shelf tools like Leonardo AI, which has a great range of tools that allow you to use a variety of models. I’m on their creative partner programme and I’ve been a beta tester for other companies.

Yitzi: Do you focus primarily on video, or do you also focus on still photography and music? Or all of the above?

KEZIAI: I do all of the above, and I love letting them inform one another. I work a lot with AI music. I studied electronic music at art school and have collaborated with many composers over the years, briefing them on music for my pieces. Being able to produce my own tracks is something else entirely. I can start with an emotion or a feeling, put that into words or images, and then create the music. Synthography is another part of my practice. I had an exhibition at the Saatchi Gallery, London, in 2024, as part of Pivotal Digitalism, featuring stills printed on archival metal. It was wonderful seeing these AI images blown up large. I had upscaled them multiple times so that if you looked closely, there were scenes within the scene. The metal prints gave them a depth that almost looked like a Cibachrome print — those slide prints with luminous depth. I like spanning from the truly analog right through to the very digital, finding the right form to get the concept across.

Yitzi: What fascinates me about you is that I’ve spoken to a lot of artists who hate AI or are terrified of it, thinking it will put them out of business. I’ve also spoken to AI artists who don’t really understand old-fashioned, analog methods of creating art. But you are a real expert in both traditional analog art and AI. Why is your perspective different from the many artists who are afraid of AI?

KEZIAI: I’ve always embraced new technology. Working with AI reminds me of creating video art in the 90s because of its experimental nature. I’ve shot on a massive variety of formats over the years: SVHS, VHS, DigiBeta, Beta, 8mm, 16mm, 35mm. I bring from traditional filmmaking a love of the craft and a deep understanding of film language: storytelling, performance, concept, visual imagination and technical curiosity. All of that translates directly into AI. I think adaptability, critical thinking and a hunger to keep learning are valuable skills today. This emerging technology is a revolution; the impacts it will have are bigger than the Industrial Revolution. In terms of creativity, it democratizes learning. With AI you can go down any rabbit hole, learn whatever you like, go deeper, move faster. It is a remarkable time to be a creative.

Yitzi: One thing that strikes me from what you’re saying is that your analog artistry actually made you a far better AI artist. By focusing on things that are impossible to do with traditional mediums, you understand where the limits are, and now you can create something truly fantastic. It sounds like traditional artists actually have a leg up. That is a great way to allay the fears of artists being replaced — if you embrace it, you could become an even better artist than before. Am I understanding that correctly?

KEZIAI: Yes, you can be a different artist than before, and it accelerates your ideation and thought process. I’ve been vibe-coding interactive artworks — using language to code. I did some coding at art school for interactive pieces, and AI really allows me to expand on those ideas now. The gap between your idea and something others can see has collapsed. Even if you’re just ideating to build with a team later, you can use AI to hone your ideas.

Yitzi: I read that some platforms like Spotify ban AI music. You either can’t upload AI music to their platform, or if you do, you cannot monetize it. That’s what I read, though I could be misinformed.

KEZIAI: All of this is in a state of flux; we are building the plane as we fly it because it’s moving so fast. Musicians have an established royalty system that filmmakers don’t. When I create AI music it genuinely comes from myself, just like any other creative process; I enjoy it deeply. I write the lyrics or the brief, refine it, edit it, and go through a similar creative workflow. You could just push “go” and let it automatically make things, but if the world is flooded with everything, what has meaning? What stands the test of time? What do people engage with? I am interested in creating work with meaning that resonates with people and possesses humanity.

Yitzi: Here is our signature question. Kezia, you’ve been blessed with a lot of success, and you must have learned a lot from your experiences. Looking back to when you first started, can you share five things that you need to create a highly successful career as an artist?

KEZIAI: Five things you need to be a successful artist. Okay.

• First, be curious. Curiosity is the engine. It leads you into new mediums, new tools, new ways of thinking. It’s what keeps the practice alive.
• Second, transcend boundaries. Don’t limit yourself to one medium. The world tries to box you in, but working across your interests, like how mine have all tied together, is exhilarating. Life and art are not linear progressions.
• Third, embrace failure as a methodology. There’s a great quote from Samuel Beckett we had on our wall growing up: “Ever tried. Ever failed. No matter. Try again. Fail again. Fail better.” I love that idea, and it ties perfectly into AI. Working with AI is iterative failure made literal. You generate, it misses, you adjust, you generate again. The practice of making the work is the practice of failing better. Enjoy the play.
• Fourth, imagination is the medium. Whatever you can imagine, or haven’t imagined yet, is now possible to visualise. Create from within, and reach outward.
• And fifth, don’t wait to create. Create in whatever way you feel compelled to create. It will lead to the next thing, and the next thing, a lifelong trajectory of creating in whatever medium fits the idea.

Yitzi: What would you recommend to AI artists? How can they monetize their work, turn it into a career, and support themselves?

KEZIAI: There are several avenues people take with AI art. Working as an AI artist or director on commercials and films. You can be an NFT artist and sell work digitally online; there is an interconnected world of AI artists in Web3 and on X. Some people focus on social media and monetize that way. I know people successfully creating across all of those different avenues. It’s about finding your niche, and following your passion.

Who are some of the most interesting people you have interacted with? What was that like? Do you have any stories?

Jacqui Kenny, known as The Agoraphobic Traveller, is a New Zealand artist based in London who makes photographs from Google Street View because agoraphobia keeps her from travelling in person. Twenty years ago she was a director’s assistant and visual researcher when I started out directing — she’s always had such an amazing eye. We’ve stayed friends over the years and curated my AI stills collection CHRONIC together. We both ended up working with technology to make personal work, decades after working together on commercials. It’s pretty magic.

Rebekah Tolley-Georgiou, who curated PIVOTAL: Digitalism at the Saatchi Gallery in London. Being part of that show was a turning point. She said of the work: “Her ability to merge the language of traditional cinematic storytelling with the innovative world of AI is unlike anything I’ve seen.” That kind of support from a curator of her calibre is wonderful.

Alejandro Cartagena, the Mexican photographer-artist. We’ve had ongoing conversations on X Spaces about how AI is changing the art landscape. With Andrea Itzel Rodríguez, he co-founded Fellowship — a gallery dedicated to digital artists working with new media. Their platform daily.xyz released my ‘Dance Interrupted’ series, which brought my AI work to a wider international audience.

Farrah Carbonell at Art and Vault — an NYC based curator and artist. She’s been a generous supporter of my work, curating it, and together we host conversations and interviews with other artists in the space. That kind of behind-the-scenes community-building is unique in any creative scene, and matters a lot in one still defining itself.

None of us are able to achieve success without some help along the way. Is there a particular person who you are grateful towards who helped get you to where you are? Can you share a story about that?

It would have to be Megan Kelly, my EP at Honor Society. We’ve been working together since 2010. With Megan it isn’t one moment — it’s the consistency. Megan backs women directors and brave storytelling, and now embraces AI cinema in commercial work.

To have someone in your corner that long, with that kind of openness, giving directors the space to do their best work — that’s rare. I’m excited to move into the Imagination Age together.

Yitzi: Because of the platform you’ve built and your enormous personal influence, if you could spread an idea or inspire a movement that would bring the most amount of good to the most amount of people, what would that be? You never know what your idea can spread.

KEZIAI: Be curious. Follow your interests and instincts; you never know where it will lead. It might surprise you. Embrace heuristic learning — learning through doing, through failing: that’s where the real discoveries are. Embrace change. Imagine more bravely than we ever have, with creativity and empathy. Curiosity leads to innovation.

Yitzi: Beautiful. Thank you so much for your time. I wish you continued success and good health, and I hope we can do this again next year.

KEZIAI: It has been really lovely speaking to you and talking about how AI is a pseudo-film set for what isn’t yet.

Yitzi: I love it. I love it so much. I could learn a lot from you. Thank you.

KEZIAI: Thank you, anytime. It was lovely speaking with you.

Kezia Barnett is represented commercially through Honor Society.