Matt Lindsey

ABSTRACT This paper explores intersections between Buddhist wisdom traditions and artificial intelligence through the “Dharma Setu” framework. Combining Sanskrit dharma (teachings) and setu (bridge), this creates a technological bridge connecting ancient wisdom with modern computational capabilities. Drawing on 10 billion tokens from the Buddhist Digital Resource Center, we propose a tripartite methodology: (1) Data Resources – Buddhist textual corpora; (2) Computer Science – computational methodologies transforming text into structured representations; (3) Mathematics – frameworks deriving meaning from outputs. The framework employs vector addition, interpolation, and semantic similarity operations on high-dimensional embeddings within an integrated LLM + vector database + graph database architecture. We examine three dimensions: epistemological parallels between vector operations and Buddhist pramāṇavāda traditions; AI systems as vehicles for dharmic transmission via Tibetan sensory liberation practices; and technical architectures for Buddhist-informed AI systems. The framework functions as both a sophisticated retrieval system for organizing Buddhist teachings across the corpus and a generative mathematical system performing operations on dharmic concepts. While acknowledging limitations in representing contemplative traditions computationally, we suggest mathematical operations on dharmic vectors could generate new teachings maintaining authentic connection to traditional sources, enabling dharma to evolve for contemporary contexts.