PhD candidate, Faculty of English, University of Cambridge

Nichole Anderson Ravindran’s doctoral research focuses on how to demonstrate an anticipatory storylistening method by using science fiction as alternative future realities - thought experiments - to study the complexity of the AI alignment problem within the human ecosystem. One of the central goals of her project is to understand how advanced AIs could potentially learn, understand, and evaluate the human values, principles, ideals, and goals set out for them within the larger learning environment of the human ecosystem. The research aims to integrate and synthesize different types of knowledge to generate more holistic insights and understandings on the AI alignment problem. It demonstrates the value of stories in providing the cognitive detachment necessary to assess real-time, real-world and future-oriented societal issues with longterm-nearterm dynamics. Finally, it offers translation from conceptual and theoretical understandings to practical ways of breaking down the alignment problem, which can eventually develop into beneficial AI strategies and policies.