AI 3D Environments Help Designers See Through Non-Human Eyes
Navigable worlds from animal traces let designers reflect in seconds.
Aung Pyae's exploratory study tackles the epistemic challenge of more-than-human design: designers cannot directly access non-human experience. The research leverages a text-to-3D world generation platform that creates navigable environments from non-human traces (e.g., animal sensor data). Five participants from engineering and sustainability backgrounds explored these worlds, finding that instant navigation (within seconds) fosters reflection-in-action — a dynamic, iterative process distinct from evaluating static 2D or 3D representations. This approach helped surface anthropocentric assumptions that typically go unnoticed in traditional design workflows.
Key findings reveal that designers' epistemic stances oscillate: some treat AI outputs as generative provocations for speculation, while others see them as authoritative depictions of non-human reality. This oscillation creates what the author calls "technologically-amplified backtalk" and "productive provisionality" — frameworks for understanding how AI-generated 3D environments can mediate more-than-human design. The study suggests that rapid, explorable 3D worlds can help designers challenge human-centered biases and incorporate non-human stakeholders into the design process, though further research with larger, more diverse participant groups is needed.
- Study uses a text-to-3D world generator to create navigable environments from non-human traces (e.g., animal sensor data).
- Instant navigation (seconds) enables reflection-in-action, unlike static representations like photos or blueprints.
- Designers oscillate between seeing AI outputs as speculative provocations or authoritative representations, revealing anthropocentric biases.
Why It Matters
Brings non-human stakeholders into design, challenging anthropocentric assumptions with rapid AI-generated environments.