Touching Wires: tactility and a quilted musical interface for human-AI musical co-creation

Abstract

Interactions with computers have traditionally been mediated by rigid materials, but as technology evolves, there is increasing potential to rethink these relationships. This paper explores how a soft, textile-based interface can reshape human-AI interaction, particularly in musical co-creation. We introduce a textile-based human-AI system used both for musical performance and public interaction. This system enables embodied, tactile engagement with an AI agent, offering users a more unique and participatory experience in human-AI musical co-creation. We aim to examine the potential for soft materiality to mediate more dynamic human-AI interactions. Our findings reveal that users’ choices when interacting with novel systems are informed by their expectations and biases, that embodied learning is built iteratively on layered multi-sensory experiences, and that there is a desire for familiarity and understanding when interacting with AI systems. We found that the materiality of our textile human-AI interfaces influenced how users choose to interact, and that users sought clarity in the AI’s role in collaborative creation. This work contributes to our understanding of how entanglement, embodiment, and materiality impact our relationships in human-AI collaborations. Show more