Presented by: Monojit Choudhury
Abstract: Any use case or output of AI systems carry deep cultural connotations, yet today’s systems remain predominantly Western and Anglo-centric and risk stereotyping and cultural homogenization—often erasing underrepresented (indeed majority) cultures. Paradoxically, defining “culture” for AI can itself flatten diversity, even as AI enables cultural analysis at unprecedented scale. This talk examines this duality and argues that they are two sides of one coin: culture-aware AI systems can simultaneously mitigate bias and expand rigorous, respectful cultural scholarship.