Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

ReTracing: An Archaeological Approach Through Body, Machine, and Generative Systems

Yitong Wang, Yue Yao

TL;DR

ReTracing tackles how generative AI shapes embodied action by treating AI systems as active agents in an archaeological workflow. The method converts literary excerpts into prompt-based choreographies, realized through diffusion-based video generation for humans and scripted robotic actions, then captured and reconstructed as 3D motion traces on a mirrored stage. This multi-agent pipeline reveals how AI internalizes social biases and control logic within movement and preserves it as an open digital archive. The work contributes a concrete, auditable workflow for examining AI embodiment at the intersection of art, robotics, and critical technology studies, with implications for bias analysis and responsible data governance.

Abstract

We present ReTracing, a multi-agent embodied performance art that adopts an archaeological approach to examine how artificial intelligence shapes, constrains, and produces bodily movement. Drawing from science-fiction novels, the project extracts sentences that describe human-machine interaction. We use large language models (LLMs) to generate paired prompts "what to do" and "what not to do" for each excerpt. A diffusion-based text-to-video model transforms these prompts into choreographic guides for a human performer and motor commands for a quadruped robot. Both agents enact the actions on a mirrored floor, captured by multi-camera motion tracking and reconstructed into 3D point clouds and motion trails, forming a digital archive of motion traces. Through this process, ReTracing serves as a novel approach to reveal how generative systems encode socio-cultural biases through choreographed movements. Through an immersive interplay of AI, human, and robot, ReTracing confronts a critical question of our time: What does it mean to be human among AIs that also move, think, and leave traces behind?

ReTracing: An Archaeological Approach Through Body, Machine, and Generative Systems

TL;DR

ReTracing tackles how generative AI shapes embodied action by treating AI systems as active agents in an archaeological workflow. The method converts literary excerpts into prompt-based choreographies, realized through diffusion-based video generation for humans and scripted robotic actions, then captured and reconstructed as 3D motion traces on a mirrored stage. This multi-agent pipeline reveals how AI internalizes social biases and control logic within movement and preserves it as an open digital archive. The work contributes a concrete, auditable workflow for examining AI embodiment at the intersection of art, robotics, and critical technology studies, with implications for bias analysis and responsible data governance.

Abstract

We present ReTracing, a multi-agent embodied performance art that adopts an archaeological approach to examine how artificial intelligence shapes, constrains, and produces bodily movement. Drawing from science-fiction novels, the project extracts sentences that describe human-machine interaction. We use large language models (LLMs) to generate paired prompts "what to do" and "what not to do" for each excerpt. A diffusion-based text-to-video model transforms these prompts into choreographic guides for a human performer and motor commands for a quadruped robot. Both agents enact the actions on a mirrored floor, captured by multi-camera motion tracking and reconstructed into 3D point clouds and motion trails, forming a digital archive of motion traces. Through this process, ReTracing serves as a novel approach to reveal how generative systems encode socio-cultural biases through choreographed movements. Through an immersive interplay of AI, human, and robot, ReTracing confronts a critical question of our time: What does it mean to be human among AIs that also move, think, and leave traces behind?
Paper Structure (5 sections, 7 figures, 4 tables)

This paper contains 5 sections, 7 figures, 4 tables.

Table of Contents

  1. Introduction
  2. ReTracing Framework
  3. Conclusion
  4. Ethical Implication
  5. Appendix / supplemental material

Figures (7)

  • Figure 1: Overview of the ReTracing framework.
  • Figure 2: A human performer and a quadruped robot enact AI-generated prompts, derived from literary depictions of human–machine interactions, within a mirrored environment.
  • Figure 3: Installation setup for the human–robot performance. (a) Side view of the mirrored installation with the human performer and quadruped robot. (b)Camera angle capturing the scene for motion trace reconstruction.
  • Figure 4: Samples of choreography video generated by a diffusion-based text-to-video model wan2.
  • Figure 5: Multi-view 3D motion tracking of human and robotic agents. Colored trajectories represent the temporal evolution of tracked keypoints.
  • ...and 2 more figures