U Ai? (2025)
The Project: A complete music video generated entirely through AI tools, from lyrics and composition to visual generation and lip-synced animation. The workflow integrated Claude for lyrical content, Suno for music generation, Flux for character imagery, and multiple image-to-video models (HuMo, Wan 2.2) for animation, with voice extraction via Stemroller and phoneme recognition through Whisper and Wav2vec2 for lip-sync accuracy.
The Development Process: This required orchestrating a complex pipeline where each AI tool's output became another's input. The technical challenge wasn't individual tool operation, but understanding how stem separation, phoneme detection, and video generation models interacted. When lip-sync accuracy failed, I needed to diagnose whether the problem stemmed from vocal isolation quality, phoneme timing precision, or the image-to-video model's limitations. The three-day timeline forced pragmatic decisions, using detailed text prompts for visual consistency rather than training a LoRA model, which would have been technically superior but time-prohibitive.
Success required understanding the entire generative pipeline: how audio waveforms translate to phoneme timing, how diffusion models interpret visual prompts, and why consistency across video frames requires either model fine-tuning or precise prompt engineering. ComfyUI provided the workflow interface, but I needed to comprehend data flow between nodes, troubleshoot where generation quality degraded, and make informed trade-offs between speed and accuracy. The result demonstrated that AI-assisted creative production isn't just about automated generation, understanding each system's capabilities and constraints well enough to guide the process meaningfully.
BLØFFEREN2k (2025)
The Game: All players view an AI-genereated image simultaneously for ten seconds, then two players, a truthful "witness" and a deceptive "liar", describe the image while investigators vote on who's being honest. The game tests memory, persuasion, and deception through visual evidence.
The Development Process: Creating this multiplayer web application required a conversational back-and-forth where I articulated game mechanics while Claude handled technical implementation. The dialogue wasn't linear, each version (v2.0, v2.1, v2.2) emerged from testing and discovering new problems.
Success depended on hybrid competency: I needed enough technical literacy to identify where problems occurred (client-side timing vs. server-side state) and evaluate whether proposed solutions made architectural sense. Claude provided syntactically correct code, but I maintained oversight of game logic, user experience, and whether mechanics actually worked as intended.
LUREREN2k (2025)
I developed this social deduction game, where one player unknowingly lacks the secret word through iterative dialogue with Claude. The process required me to understand web architecture well enough to articulate what I wanted, evaluate solutions critically, and spot problems when they arose. Claude handled the technical implementation, but I had to maintain the overview: game mechanics, user experience, and how everything fit together.
This isn't about typing prompts and getting finished products. It's more like working with a skilled collaborator who's excellent at syntax and technical details, while I focus on design logic and whether things actually work as intended. The result was a functional multiplayer game in under two days - something that would normally take weeks - but only because I could bridge the gap between "here's what the game should do" and "here's why the database isn't updating correctly."
PUBLIC ENEMY: ART BY LAILA BERTHEUSSEN (2023)
AI-graphics and visual profiling to exhibition during Arendalsuka, Norway.
XSILO (2022)
AI generated art exhibition at Agder Kunstsenter, Norway. Exploring the value of visual art, and goes into dialogue with the newly complete modern art museum in Kristiansand, Norway, KUNSTSILO.
Part of Traavik.info and Wastefullness comission project
KUNSTENS BETYDNING? (2021)
Anthology by research group “Art and young people”, University of Agder.
Co-author on an anthology looking at expanded views on children and young peoples exposure to art, and arts place in society.
Available as an open access book, and also as print on demand. Most chapters are written in Norwegian, with English abstracts.
PARTY SONGS (2021)
The project is investigating the possibilities of pop music as an instrument of control and liberation that brings together idolatry, hero-worship and still darker strains of mass behavior.
St. Petersburg, Russia 2021
EN MASSE (2020)
Photographic exhibition at SørVeiv2020 - exploring the many live-experiences that took place in the region over the past 10 years. Focus on the collective gatherings, and the individual component that builds the crowds we once were used to - and perhaps one day will see once more.
Kristiansand, Norway 2020
Hjerneteppe (2020)
Exploring whats inside our heads through drawing. A play on word often mistaken in Norwegian, participants would fill in their thoughts on brain-shaped pieces of paper, creating a “hjerneteppe”. Interactive, collective art during Forskningsdagene at Universitet i Agder.
Kristiansand, Norway 2020
Thanks to MFA student Lilly Bjørk for additional help with practical arrangements and concept!
The liberation of Israel (2019)
Photography: Solo exhibition, Tel Aviv 2019
Liberation Days (2019)
Jean “Valnoir” Simoulin | Morten Traavik
Book: Photographic contribution 2019
War of Art (2019)
Film: Stills photography 2019
Forræderens guide til Nord-Korea (2018)
Book: Photographic contribution 2018
Preuss Museum juried exhibition (2018)
Preuss national photo museum
Horten, Norway 2018
Photography: Joint exhibition
Forbidden whispers. Laibach and North Korea. (2018)
Liberation Day (2016)
Film: Stills photography 2016
Yes we love this country (2012)
The Narrows (2008)
Film: Digital artist.
Visual effects (particles, compositing) 2008
Zombie Prom (2006)
Film: Digital artist.
Visual effects (Compositing, rotoscoping) 2006