The Process of AI-Generated Simulation

I recently finished an exhibition project, collaborated with Nero Chenxuan He, presented as part of the 2025 Osaka Expo under the unit theme “Raising Awareness and Engaging in SDGs through Sustainable Art”.

The installation explores the speculative relationships among natural organisms through digitally simulated growth, hybrid materialities, and spatial choreography.

Our aim was to reimagine an ecosystem where orchids, mosses, fungi, and vines cohabit in mutualistic wonder—each element simulated with procedural modeling and dynamic animation in Houdini and integrated into a unified scene in Cinema 4D. These hybrid species evolve in synchronized motions, suggesting a future where digital fabrication, ecological intelligence, and synthetic biology converge.

Through this work, we celebrate the aesthetic complexity and structural intelligence of natural forms, while also reflecting on the digital afterlives of materiality—a Garden of Eden reborn through computation, evoking speculative symbiosis between nature and technology.

Installation on Display at the 2025 Osaka Expo – Japan Pavilion

Throughout the development of the simulation system, I was responsible for building the core logic and structural behavior in Houdini, focusing on dynamic growth, procedural folds, and material response. We actively integrated GPT-4o into our workflow—not merely as a tool, but as a co-creative agent. By feeding reference imagery and sharing screens during iterations, we were able to receive feedback on node linking strategies, write and debug VEX code, and refine attribute controls. This experimental approach helped us navigate technical challenges quickly while achieving highly intricate and responsive simulations.