This project explores AI safety, risks, and alignment through the lens of popoular TV shows and movies which most people will already be familiar with. The goal is to use these existing references as familiar, accessible, and engaging examples in order to broaden awareness, understanding, and debate around increasingly powerful AI.
This is an area where science fiction has generally had a very negative impact on cultural connotations, focusing on extreme existential scenarios like Skynet and Ultron. While these risks are important and definitely need representation, in practice, modern AI development and safety is much more nuanced. So to better understand that nuance, I wanted to explore it through a fun and engaging, albeit heavily biased lens.
These example scenearios are not meant to be exact predictions or concrete evidence of AI risk but rather useful analogies for understanding real AI risk scenarios and topics in AI safety. In doing so, we also naturally explore how AI is portrayed in popular media, how grounded these portrayals are, and how well they transfer to modern frontier AI development in practice.
Oh also, there’s a strong “for shits & giggles” aspect to this project. eg, I just thought it’d be a lot of fun 😉
This data is released publicly under a permissive CC0 Public Domain license.