Experiential AI: A transdisciplinary framework for legibility and agency in AI

Explainable AI has significant limitations: The paper argues that current approaches to making AI "explainable" fall short in helping people truly understand how AI systems work and affect their lives, suggesting that technical explanations alone are insufficient for meaningful public comprehension.

Artists and creative practitioners have essential roles to play: By bringing artistic methods into AI research and development, creators can help bridge the gap between technical complexity and human understanding, offering perspectives that scientists alone cannot provide.

Experience-based learning transforms understanding: Rather than relying solely on technical documentation or abstract explanations, the "Experiential AI" framework proposes that people learn about AI systems most effectively through direct, creative engagement that transforms how they think about human-machine relationships.

AI systems require transdisciplinary approaches: The paper emphasizes that addressing responsible AI development requires collaboration across multiple fields—not just computer science—to better account for social, cultural, and ethical dimensions of AI technology.

Public engagement shapes AI development: By creating rich, interactive experiences that help diverse audiences (scientists, practitioners, and the general public) understand AI differently, we can collectively influence how these systems are designed and deployed in society.