Summary
Nageswaran argues that open-weight AI models, while widely diffused, pose significant dangers, including the potential to create biological weapons. He advocates for a global control apparatus similar to those established against nuclear threats to manage these risks.