Traces is a multimedia exploration of storytelling, cultural preservation, and machine learning, rooted in the spatial-temporal philosophies of the Gullah people from Pawley’s Island, South Carolina. This work engages dance, photography, and AI to trace the shape of ancestral knowledge through embodied gestures.

At the heart of this project is a hand-drawn visualization of the Gullah understanding of time, space, life, death, and the divine—an abstract composition of overlapping loops that signifies an interconnected cosmology. This drawing serves as both a conceptual map and a kinetic score, informing a series of movement-based gestures that translate these ideas into a physical language.

Through performance, my body echoes these sacred forms, dancing the geometry of Gullah temporality into existence. The choreography tells a deeply personal narrative, reflecting the discrete laws of life within a Gullah family lineage. Still images from the dance capture key gestural moments, preserving them as visual artifacts of movement.

Beyond its performative dimension, the project integrates artificial intelligence to engage with storytelling in a multimodal capacity. A machine learning algorithm is trained to interpret these gestures, encoding them with elements of the narrative, effectively allowing AI to 'read' and respond to a story through motion. In live performances, this process extends storytelling beyond the body—activating AI as both a witness and a participant in the narrative.

This work challenges conventional AI training methodologies by centering cultural specificity, asking: How can AI be trained to understand and process the world through a Gullah cosmological lens? The project is a meditation on the ethics of data and the potential for AI to engage with non-Western epistemologies, offering a model for encoding cultural knowledge in ways that respect and extend community values.