Tahn is a Korean media artist building living chronotopes from postwar memory, structural perception, closed-dataset AI, biosignals, and Korean symbolic systems. His work does not use technology to decorate tradition; it exposes how memory survives, mutates, and returns under pressure.
Only works that clarify Tahn's core system are placed here.
Before the algorithm, there is the hand: 101 original drawings become a self-contained dataset, then a custom LoRA system, then a generative memory field. This method protects authorship while keeping the visual language rooted in Tahn’s own structural perception rather than generic prompt aesthetics.
101 hand-drawn works form the artist-owned source material. No external image scraping, no internet-sourced training data, no third-party visual material.
Gestural memory, structural perception, fracture, surface pressure, and spatial rhythm are trained into a controlled model that carries Tahn's perceptual grammar.
Live news, sound, EEG, breath, and attention signals disturb the trained field, so memory appears as a living structure rewritten by the present rather than a static image output.
Artist-owned drawings establish the visual corpus and ethical boundary.
The model learns structure, density, grain, fracture, and temporal rhythm.
TouchDesigner and live signals turn the trained field into a time-based system.
Selected works can be configured as single-channel moving image, multi-screen projection, sensor-responsive installation, public media façade, or research-based participatory system. TouchDesigner-based logic, local playback, fallback modes, and site-specific projection mapping keep the work adaptable without confusing proposal language with finished documentation.
This AI-generated spatial visualization study is not documentation of a completed installation. It functions as a scenario model for scale, atmosphere, projection logic, visitor movement, and site-specific adaptation.
Open on YouTubeI was born into a country that had just emerged from the devastation of war. My parents' generation fought for survival and reconstruction; my generation inherited a different war, one of identity, adaptation, and attention in an accelerated technological era.
Because of my red-green color deficiency, I do not build images around chromatic certainty. I read the world through structure, density, texture, fracture, rhythm, and spatial pressure. This perceptual condition became the conceptual engine of my media installations.
My work brings Korean symbolic systems such as Ilwolobongdo, moon jars, folk talismans, and ritual architectures into contact with AI vision models, EEG, breath, environmental signals, and generative code. I do not use technology to decorate tradition. I use it to test how tradition survives under pressure.
In my installations, time is not a neutral background. It is a material. Memory is not an archive. It is a living organism. The audience is not a visitor outside the work. They become a force inside its field.
War is treated not as illustration but as a migrating condition: from rubble to labor, from identity to data, from national recovery to private perception.
Color is not the primary grammar. Form, density, grain, interval, fracture, luminance, and pressure become the visual language of the work.
Custom AI systems are trained from Tahn's original works, making the model an extension of the artist's visual DNA rather than an anonymous image machine.
EEG, breath, movement, dwell time, and environmental data do not decorate the screen. They reorganize memory, rhythm, and spatial behavior.
Selected artist, International Symposium on Electronic Art, Dubai, United Arab Emirates.
Finalist, Top 3. International jury; ceremony presented by the Director of Guggenheim Abu Dhabi.
Watch event videoVisual Arts Commission, Ministry of Culture, JAX District, Riyadh, Saudi Arabia.
Official media facade screening within Riyadh Art public art context, Saudi Arabia.
Data and AI, Imaginary Landscapes, Fictional Narratives, three-person exhibition.
Platform exhibition, collaborative project with Pozamong.
PDF materials are provided for curators, juries, residency panels, commissioners, and press review.
Focused inquiries are welcome from museums, biennials, public art programs, cultural foundations, universities, curators, collectors, and commissioning bodies seeking media works grounded in time, memory, attention, and place.
Single-channel moving image, multi-screen installation, AI-generative time-based installation, and sensor-responsive environments for museums, festivals, and curated programs.
Site-specific public media rituals for façades, civic screens, airports, cultural plazas, and architectural surfaces where Korean symbolic systems meet contemporary urban memory.
Inquiries regarding editioned media works, institutional acquisition, closed-dataset AI works, provenance documentation, and long-term preservation materials.
Artist talks and workshops on chronotopic media art, structural perception beyond color, closed-dataset AI authorship, biosignal interaction, and public media practice.
For site-specific commissions, acquisitions, exhibitions, public media projects, or research collaborations, please outline the institutional context and technical conditions. This form is configured for Netlify Forms after deployment; direct email remains available below.
For urgent deadlines, please include the institution name, project format, venue or platform, technical requirements, budget or support range, and expected timeline in the first message.
For curatorial review, commission proposals, public media projects, artwork acquisition, artist talks, and workshop invitations, please contact Tahn directly.
tahn.artist@gmail.com