Deepest Unknown project is an artistic collaboration with AI, generating speculative stories and images that illuminate the mysteries and potential losses of remote underwater environments. The project envisions life beyond our limitations while exploring our connection with water and the ecological consequences of human actions. Through digital tools and mechanical techniques, the project transports viewers into alternative realities, expanding our comprehension of deep ocean ecologies. The analysis of individual sea creatures prompts a reassessment of our relationship with water and recognizes technology's impact on our perception of nature.

The Deepest Unknown v. 2 has evolved into an immersive VR film experience, showcasing a digitized deep-sea landscape populated by AI-generated creatures. Using customized datasets and advanced techniques like the StableDiffusion plugin enhances the realism of the virtual world. The soundscape is inspired by sound propagation in deep water, piecing together AI-generated sound samples and hydrophone field recording data from MBARI and NOAA. The film script utilizes ChatGPT to analyze a variety of text inputs including conversations, a grant proposal, speculative dialogues, and project descriptions, which are then read by an AI-voice named Carla. The installation layout is inspired by deep-water currents, offering a 3D aquatic landscape from the two individual perspectives of the VR viewers, which the public can follow on the separate screens.

-
Deepest Unknown v.2 is realized in collaboration with Colette Aliman (deep-sea soundscape design), Loden Rietveld (deep-sea soundscape engineering), Taegwon Bae (film Art Directing, 3D Art), Changju Yoon(UnrealEngine engineering), Seungmin Lee(film editing, VFX post processing), YoungJae Nah(animator), and Marsha van Leersum (assistant). Deepest Unknown v.1 is realised by collaboration with Ahnjili Zhuparris (Customising StableDiffusion AI), Colette Aliman (Deep-sea soundscape design), Tiiu Meiner (Co-writing GPT-2 video script), Tivon Rice (‘models for environmental literacy’ GPT-2 dataset), Taegwon Bae(Film Art Directing , 3D Art), Changju Yoon(UnrealEngine engineering), Seungmin Lee(film editing, VFX post processing), YoungJae Nah(animator), and Marsha van Leersum(Assistant).