Back to Research
AISOMA: What Choreography Teaches Us About AI Fitness
Inspiration5 min2026-03-02

AISOMA: What Choreography Teaches Us About AI Fitness

Wayne McGregor's Google Arts & Culture project proves that AI and human movement can have a creative conversation — and why that matters for fitness technology.

AI as a Creative Catalyst

"AI is a tool just like a pencil or a camera. It's a way of being able to develop new thoughts in choreography." Wayne McGregor's words frame the most important insight about AI in movement: it's not the end point, it's a catalyst.

AISOMA is a choreographic tool built with Google Arts & Culture. You dance a phrase of movement in front of a camera. Using both pose detection and phrase detection from McGregor's archive of choreography over three decades, the AI generates a new phrase added onto yours. You can then learn that combined phrase as a dance.

The Creativity Parameter

The most interesting feature of AISOMA is the creativity slider — how much creative freedom you give the AI relative to the original phrase. It can stick very closely to your original movement, or it can push that phrase into a dimension which is radically far from the original.

This is the exact same design challenge in AI fitness coaching. How much should the AI deviate from the "textbook" form? A rigid system that only accepts perfect poses discourages beginners. A system that's too lenient provides no value. The creativity parameter in AISOMA is the UX challenge of every AI fitness product, reframed as art.

Human in the Loop

McGregor describes the interaction: "When you have this idea of the human in the loop of an interaction with an AI, you create a similar kind of conversations." The AI offers things up in the same way that a dancer in the studio works with you in real time.

This is the design principle that separates good AI fitness products from bad ones. The AI should feel like a training partner — responsive, adaptive, offering suggestions — not a judge handing down verdicts. The goal is conversation, not correction.

From Dance to Fitness

AISOMA proves something that most fitness tech companies haven't internalized: the most compelling use of pose detection isn't clinical analysis — it's creative interaction. "Your dance journey might just be starting now with AI Soma. It's a tool that might encourage more people to try things with their bodies."

The fitness industry is obsessed with accuracy metrics and joint-angle deviations. AISOMA suggests the real opportunity is using AI to make movement joyful, exploratory, and personal — with accuracy as a foundation, not a feature.

Want to go deeper?

Explore our audience-specific research for investors, studio owners, or builders.