Pinyao Liu

The rapid adoption of conversational AI has established language as the dominant interface between humans and artificial intelligence. Through conversational systems, interaction with machines is structured as a dialogue of questions and answers, positioning AI as a source of authority. In doing so, these systems reshape how humans relate to knowledge, authority, and each other.

SOMA is a multisensory robotic sculpture that moves and breathes with the visitor’s movement, engaging bodily memory as an alternative mode of human–AI communication. The work invites visitors to re-enact a remembered movement from personal memory through a wearable sensing device. As the visitor moves, the machine surfaces the qualitative aspects of that movement through its own mechanical choreography—reorchestrating the narrative fabric of memory through its motion. As it listens, interprets, and responds, SOMA forms an intimate, nonverbal dialogue between human and robotic bodies.

This project investigates somatic intelligence at a time when human–AI interaction is largely dominated by conversational agents. As contemporary AI systems increasingly privilege text, instruction, and optimization, the body is often reduced to an input device—or excluded altogether

SOMA proposes a speculative interface in which human and machine engage through the tacit intelligence of the body. It seeks to reclaim the body as a site of communication with artificial intelligence, positioning somatic memory as both metaphor and method. It asks what it means to relate to artificial beings in a time when machines are increasingly instrumentalized and the body is marginalized in human–AI interaction.