This study explores the meaning of AI-generated audience responses for a choreographer in dance creation. Focusing on one completed dance work, the study retrospectively constructed AI audience responses from archived creative materials—choreographic and rehearsal notes, movement research videos, an early-stage recording, a final performance video, and a published review—and examined the choreographer’s reflective interpretation of them through a semi-structured interview.
The participant received the AI responses through active acceptance, selective acceptance, and critical distancing, and experienced them as a “creative mirror,” an “external viewpoint,” and “mechanical interference.” Building on these findings, the study proposes the concept of AI audiencehood: a complementary, interpretive audience function distinct from human audiencehood, defined by its core (mirroring), mechanism (externalizing the imagined audience’s gaze into reviewable text), conditions (repeatability and low relational burden), and boundary (the absence of embodied empathy). As a single instrumental case, the study aims not at generalization but at the heuristic generation of this concept.