Mercury from illegal gold mining spreads through Yanomami rivers, soil, and bodies, while the ground already carries the tremors of the machines causing this harm. This work reappropriates seismic sensing to detect those subsonic vibrations, using open-source data and machine learning to turn them into sound. The signals animate soil-filled containers that release golden dust—traces of extraction transformed into resistance. As the dust settles on surfaces and visitors’ shoes, a decolonized geological perception emerges: one that listens to the ground and senses what lies beyond visibility.