Shrey Jain, an applied scientist at Microsoft Research Special Projects, speaks with Matt Prewitt on a very timely and topical subject: AI and – more specifically – the dangers it poses to the nature of natural human communication (“context collapse”). They take a deep dive into the current threats to privacy by expanding beyond the often discussed cryptographic sense into “privacy as contextual integrity”, and the immediate opportunity to embed ethical guardrails into this ever-changing realm of generative AI through possible solutions of designated verified signatures in “plural publics”. Shrey’s recently published paper co-authored with Divya Siddarth and E. Glen Weyl “Plural Publics” is linked in the episode notes.