The Smartest People Alive Cannot Agree Whether AI Ends Us. Read Every Side’s Best Argument Yourself.
Nate Soares says if anyone builds it, everyone dies. Roman Yampolskiy has spent a career proving why control may be mathematically impossible. Tom Griffiths studies the one intelligence we actually understand, ours, and what it predicts about the machines. Terence Tao, the greatest living mathematician, is already using AI to do mathematics. They cannot all be right. I sit with each of them for hours, push on the weakest joint in every argument, and refuse to tell you what to think. I’m not asking you to believe anyone. I’m giving you the papers, the impossibility proofs, and the rebuttals, so you can decide.
ENTER THE DEBATE
What you get:
- 📄 The AI Crossfire Reading List: the strongest published case from every camp, Soares and Yampolskiy’s warnings, Griffiths on minds, Tao on machine mathematics, and Yann LeCun’s steerable-by-design rebuttal
- 📝 Full transcripts + show notes from every AI episode
- 📖 Flatland, delivered instantly
- 📬 The Monday M.A.G.I.C. Message: one short email a week on the biggest questions in science
🌠 Have a .edu email and live in the USA? You automatically win a real meteorite. Sign up with your .edu address, and it’s yours. I mention this in the episode, and I mean it.
I respect your privacy. Unsubscribe at any time.
Meteorites ship to USA addresses only. Address optional now — winners are contacted to confirm shipping.
MORE FROM BRIAN
"The only way to discover the limits of the possible is to go beyond them: into the impossible." — Sir Arthur C. Clarke
Professor Brian Keating · UC San Diego