AN INTERVIEW WITH EMI KUSANO BY SCARLETT CLARKE
Emi Kusano is a Tokyo-based artist who began her career as a street photographer in Harajuku, capturing the niche local fashion — think Kawaii, mismatched prints, neon colour, punk and anime cosplay, all mashed together. Like most people, the artist is an amalgamation of everything that has ever inspired her: she was raised on 80s pop idols like Seiko Matsudo and the cyberpunk manga Ghost in the Shell. She also enjoys the ‘magic girl’ genre in Japanese fantasy. For Kusano, the 80s carried a charming ‘so-bad-it’s-good’ aesthetic, so it’s no surprise that her work is tinged with nostalgia. Her photography reads like technicolour memories with the saturation dialed up, or, as she calls it, a “playful artificiality”.
Artificiality is both the aesthetic and the medium, because her artistic collaborator is artificial intelligence. Kusano uses models like Midjourney and Stable Diffusion to generate images. Together, they tell stories about fictitious pasts and imagined futures. Her work has been exhibited internationally, including the Saatchi Gallery (London), M+ (Hong Kong) and Grand Palais Immersif (Paris). Kusano’s approach to art reflects the techno-animist culture that shaped her. Her photography often captures surreal portraits of people in futuristic worlds being physically swallowed by machinery, blurring the lines between human and tech.
Both her subjects and settings are uncanny, perhaps because of the subject matter, and perhaps because of AI. They are like science fiction stories made visual. For her, AI is not a disembodied tool but can possess “spirit, intention, or personality” — much like the cybernetic human Major Motoko Kusanagi, the protagonist of Ghost in the Shell.
Scarlett Clarke: You describe AI as your 'collaborator'. Where do you draw the line to make sure the work still feels unmistakably yours?
Emi Kusano: The boundary that keeps the work unmistakably mine lies in curation. AI is like a camera without a lens. Anyone can press the shutter, but authorship in photography is defined by what you choose to photograph, how you frame it, which image you select from hundreds or thousands of shots. I also use my own body and face as custom data. AI tends to output optimised images — what it assumes to be the most 'beautiful' or neutral result. Training models on my own face and physical presence results in something slightly uncanny that resists smooth perfection. AI assists with generation, but assigning meaning is where I draw the line.
SC: You belong to a generation that experienced both mass media and the rise of internet culture. How has the tension between these two eras influenced the way you think about images and truth in your work?
EK: As a teenager, I took street snapshots in Harajuku. At the time, Japan had countless fashion publications. Fashion was not yet optimised for global visibility; it was local, strange and niche. Over time, I watched that magazine culture decline. As fashion became increasingly mediated by platforms and algorithms, distinct styles were homogenised. Witnessing that process gave me a sense of loss, which later informed my work. For example, Neural Fad emerged from a desire to reconstruct pasts that could have existed but never did.
SC: In EGO in the Shell: Ghost Interrogation, you use AI to create an imagined version of your own childhood. How do you feel when you see these fictitious images of yourself now?
EK: For me, looking at these images brings an uncanny sense of nostalgia. It feels familiar, yet fundamentally wrong. My parents were also deeply shocked: although these images were simulations produced by AI, they looked uncannily similar to how I actually was as a child. That moment revealed how easily memory can slip between truth and fiction once it is interpreted through an algorithm. Through this discomfort, I hope viewers begin to sense how fragile and unstable our idea of the 'self' really is.
SC: How does working with fabricated memories change the way you think about trust and connection with other people?
EK: Confronting the reality that memory can be generated or altered so easily by AI — as depicted in EGO in the Shell — has led me, paradoxically, to place greater trust in physical embodiment as the most reliable foundation for human connection. Digital images and records now always carry the possibility of being fake, making the verification of truth increasingly difficult. A real photograph of me with Sam Altman was questioned and suspected of being AI-generated. This experience made me acutely aware of how fragile digital evidence has become. Because of this, I believe that embodied experience is now the last anchor through which we can confirm one another as human.
SC: The work is now being shown alongside original Ghost in the Shell drawings at TOKYO NODE. How does your installation change when it's placed in conversation with that legacy?
EK: It honestly feels surreal. Ghost in the Shell foresaw, even in the late 1980s, technological progression, integration with AI, the spread of fake news and the resulting social upheaval we now see today. Its influence extends far beyond manga and anime, shaping films like The Matrix and many other works in global cinema. To have my installation shown alongside such an iconic and philosophically rich legacy is incredibly moving. In this context, EGO in the Shell is not just a contemporary AI project, but part of a longer conversation about identity, technology and consciousness. I found myself returning to the exhibition multiple times, simply to stand inside that dialogue.
SC: You've suggested we may be the "last generation" to experience nostalgia collectively. Is this something you find comforting, or does it reveal an anxiety about what's ahead?
EK: It gives me a very ambivalent feeling — both unsettling and comforting. Television and magazines used to create shared references: people listened to the same songs, admired the same fashion and absorbed the same images. Today, we each live inside different echo chambers. The loss of a common cultural language is, for me, a deep source of anxiety about the future. At the same time, nostalgia functions as a kind of shelter. Earlier visions of technology and the future carried a sense of excitement and hope that feels largely absent now, but that lost energy continues to resonate with me.
SC: What kind of future do you actually want to see?
EK: The future I want to see is not an escape into the past. We are the first generation to coexist with a form of intelligence that has no body, no pain, no mortality — something closer to an alien presence than a human one. Instead of treating it as an enemy, I want to engage with it as a companion and rebuild the kind of exciting future we once believed in.
SC: You've said Western-trained AI models distort Japanese cultural memory, yet you keep these distortions as part of your work. What does this look like in practice?
EK: In my project Neural Fad, I once tried to generate images of the Takenoko-zoku, a real street culture that existed in Japan during the 1980s. Because major AI models such as Midjourney and Stable Diffusion are trained primarily on Western internet data, they failed to understand this very specific and local cultural context. Instead of accurate representations, the models produced images that were completely off — including what AI researchers often call 'hallucinations'. Rather than treating these results as errors to be corrected, I chose to embrace them. I see these 'hallucinations' as something closer to alternative memories. By accepting the bias of techno-orientalism embedded in Western-trained models, I deliberately work through that distorted lens to fabricate 'pasts that could have existed'. I want to show how fragile, subjective and rewritable cultural memory actually is.
Images courtesy of Emi Kusano
