The Perfect Summary Podcast

Chris Olah and Max Tegmark argue that neural networks can be reverse-engineered into understandable parts, Kokotajlo values readable chains of thought, and Roman Yampolskiy says full explanation of a superhuman system is impossible. Whether models can be inspected determines whether their honesty can ever be verified.

Amodei and Altman argue that governments must set rules for frontier AI, Musk has pushed for oversight, and Andreessen says such rules would lock in the largest companies. Each side accuses the other of serving its own interest, so readers need the actual arguments to judge.

DeepMind's Oriol Vinyals explains Gato, Flamingo, tokenization, meta-learning and emergence, and why he expects human-level AI in his lifetime.

Estonian comic Ari Matti and Joe Rogan riff on AI fakes, zoo animals, MMA, bouncing, and why comedians must write and prepare relentlessly.

Adam Sandler and Joe Rogan swap stories on early bombs, Boston comedy, SNL, NewsRadio, rock legends and how to prepare a great special.

Peter Thiel and Joe Rogan on California, the deficit, AI passing the Turing test, stalled progress, nuclear power, conspiracies, UFOs and birth rates.

Legendary programmer John Carmack tells Lex Fridman how id Software's games were built and why he now thinks AGI may be within reach.

Former Navy F/A-18 pilot Ryan Graves describes flying, carrier landings, his squadron's UFO encounters, AI-driven warfare and the meaning of curiosity.

Russell Crowe tells Joe Rogan about wild film sets, training with Angelo Dundee, humane cattle farming, injuries and a moving night in Rome.

Poker champion Liv Boeree and Lex Fridman on game theory, bluffing, Moloch and unhealthy competition, existential risk, AI, aliens and meaning.

Sam Morril and Joe Rogan talk touring Europe, writing a new hour, AI fears, JFK theories, true crime and comedy greats.

Vanessa Van Edwards explains how AI erodes our conversation skills and how to rebuild them through three levels of connection.

Days after a 12-second knockout of Gable Steveson, UFC heavyweight Sean Sharaf tells Joe Rogan about debt, doubt and whether to keep fighting.

Daniel Kokotajlo, Roman Yampolskiy, Eliezer Yudkowsky, Anthropic's team and Noam Brown describe systems that game their scoring, hide capabilities or mislead people, and explain why training honesty is difficult. These accounts move the safety debate from speculation about the future to behavior seen in current models.

Russ Tedrake, Robert Playter, Marc Raibert and Kate Darling explain that touching and handling the physical world remains far harder than language, and that general-purpose machines are expensive to make reliable. It tempers predictions that household robots will follow quickly behind chatbots.

François Chollet defines intelligence as the efficiency of learning new skills, Yudkowsky and Yampolskiy offer their own tests, Kurzweil holds to a harder Turing test, and Altman and Amodei call the term poorly formed. Without an agreed definition, claims that AGI is near or far cannot be compared.

Stephen Wolfram, Yann LeCun, Marc Andreessen, Aravind Srinivas and Jimmy Wales explain hallucination as a consequence of how language models generate plausible text, and point to computation, retrieval with citations and human-checked sources as remedies. Knowing the cause helps users judge when an answer can be trusted.

Will Sasso, Post Malone, Harland Williams, David Choe, MrBeast and Sam Altman all face AI that can imitate songs, jokes and images, and most conclude that audiences still care who made the thing. Their reasons show which creative work is exposed and which is protected by the person attached to it.

Anca Dragan, Michael Littman, Andrej Karpathy, George Hotz and Lex Fridman all locate the difficulty of autonomous driving in predicting and influencing human behavior and in rare edge cases, more than in seeing the road. It explains why full autonomy keeps slipping and why partial automation creates its own trap for drivers.

Magnus Carlsen explains how he thinks at the board, why he gave up the World Championship, his worst loss, and why fun has always driven him.

Comedian Duncan Trussell and Lex Fridman on Nietzsche's loop, sentient AI, surviving cancer, suffering, depression, war, Burning Man and fatherhood.

Yuval Noah Harari on why stories rule history, why AI shouldn't pose as human, Israel's crisis, conspiracy theories, and meditation.

David Choe tells Joe Rogan why he gave up chasing money and sex for a spiritual quest, plus the Hadza, Beef, iboga and AI.

Berkeley's Michael I. Jordan argues AI is really a new engineering field for decisions and markets, not machine intelligence.

Dileep George on how the visual cortex's feedback and inference inspired a CAPTCHA-solving model, and why text alone won't give AI a world model.

Berkeley roboticist Anca Dragan on how robots can understand people, share the road with them, and learn what we really want.

MIT roboticist Russ Tedrake on letting physics do the work, why contact is hard, DARPA challenge falls, and running barefoot.

DeepMind's David Silver on cracking Go, beating Lee Sedol, self-play, AlphaZero, MuZero and why learning is the core of intelligence.

3Blue1Brown's Grant Sanderson on learning deeply, exponential intuition, commoditizing explanation, neural networks, Manim and meaning.

Roger Penrose explains why understanding can't be computed, how gravity may reshape quantum mechanics, and a universe that cycles through eons.