The defense against AI is ZK. Artificial intelligence is total surveillance, while zero knowledge is complete privacy. Cryptographer Uma Roy joins the podcast to explain why AI detection is a dead end and how zero knowledge proofs can verify reality. If these ideas interest you, we're building startup societies at Network School. You can apply online at https://ns.com. OUTLINE 0:22 - ZK is a powerful cryptographic technique 6:58 - AI makes everything fake 8:49 - AI detection is not robust 12:51 - No public undisclosed AI 22:17 - AI detection is a dead end 26:21 - Crypto makes it real again 44:26 - If the news is fake, imagine history 52:29 - Combating negative externalities of AI REFERENCES * Principia Mathematica: https://www.amazon.com/Principia-Mathematica-Alfred-North-Whitehead/dp/1603864377/ * How the spinach, Popeye and iron decimal point error myth was finally bust: https://irep.ntu.ac.uk/id/eprint/30230/1/7987_Sutton.pdf * Balaji on the impossibility of mechanically banning "lying" on social media: https://x.com/balajis/status/1198727092835364864 * "Truegle", Gödel incompleteness, and limits of formal truth-checking: https://x.com/balajis/status/1198729374004060165 * Jurassic Ballpark as a metaphor for approximating reality with fictional priors: https://x.com/balajis/status/1441256189178761221 SOCIAL: https://ns.com https://x.com/pumatheuma https://x.com/SuccinctLabs https://x.com/ns