AI and the narrative of the everyday life: machine learning applied to the social mapping of rural music festivals.

Given the growing ubiquity of AI and the consequent valorization of data in territorial representation, it becomes crucial to analyze how multimodal algorithms interpret spatial narratives and community dynamics. This research studies the effects of algorithmic automation on social mappings. Using “Balboa Observa” project, a collaborative web mapping initiative that documents Observatorio Music Festival in Spain, the study explores the interaction between collective cartography, ethnographic analysis, and AI-driven data processing. Unlike conventional AI practices, the prioritizes analyses that avoid imposing structured categories on everyday narratives, allowing inclusion of sensitive information for deeper festival impact understanding. Through multimodal algorithms like CLIP and ImageBind analyzing images, texts, and audio recordings, the research reveals how AI-generated spatial configurations differ from human interpretations and identifies biases from training data and algorithmic processes. The study highlights community participation and data ownership importance to mitigate biases and advocates for transparent, adaptable AI tools for social mapping.