AI + design: Turning Taste Into Data

Generative AI commoditizes image creation. What becomes scarce is creative direction and visual coherence. I treat generative AI as a creative system rather than a tool, translating art direction, storytelling, and cinematic language into structured frameworks that machines can interpret while maintaining human creative judgment. By breaking visual ideas into primitives—lighting, composition, color contrast, motion cues, and emotional tone—I turn subjective direction like “make it cinematic” into repeatable visual logic that can guide generative models.

I also develop sketch-driven frameworks that embed these creative rules directly into the drawing. The sketch defines composition, gesture, lighting direction, spatial depth, and narrative focus, allowing AI to generate variations while preserving the original storytelling intent. This approach combines traditional visual development with generative iteration, enabling rapid exploration while maintaining a consistent visual language across outputs.