PBS Space Time PBS Space Time

Did AI Invalidate Our Proton Model WRONG?

Oct 30, 2026 · 21m

Summary

This episode explores the complex internal structure of protons, explaining how high-energy scattering experiments reveal a "quark sea" rather than just three static quarks. It discusses the long-standing mystery of "intrinsic charm," where heavy charm quarks appear inside protons despite being more massive than the proton itself. The host details how the NNPDF collaboration used artificial intelligence to test hundreds of thousands of theoretical models simultaneously, providing tentative evidence that protons may sometimes consist of five quarks.

Topics discussed

Intro: Bull spend and LinkedIn ad Sponsor: Google Chrome and Gemini Protons, quarks, and the mystery of the proton interior The proton's importance and the discovery of quarks Scattering experiments: How we see the subatomic world Rutherford's gold foil experiment and electron microscopes Electron wavelength and probing the proton's interior SLAC experiments and the discovery of three quarks The standard proton model: Valence quarks and gluons Sponsor: LinkedIn Hiring Pro Sponsor: Google Chrome and Gemini Sponsor: Cuvassier Cognac Higher energy accelerators reveal complex proton structure The Quark Sea: Gluons and virtual quark-antiquark pairs Valence quarks vs. the transient Quark Sea The mystery of the charm quark in protons Intrinsic vs. extrinsic particles: E=mc^2 in collisions Distinguishing intrinsic charm from collision-created particles Early evidence for intrinsic charm and the energy paradox Quantum Chromodynamics (QCD) and calculation difficulties Sponsor: LinkedIn feed and office gossip Sponsor: LinkedIn tagline Sponsor: Google Chrome and Gemini Sponsor: Cuvassier Cognac Perturbation theory: Why high-energy QCD is easier The difficulty of modeling low-energy proton interiors Heisenberg uncertainty principle and borrowing energy Momentary charm quark pairs and the 5-quark state Proton mass and the contribution of transient quarks Stanley Brodsky's 1980 theory of intrinsic charm Model ambiguity: Why previous tests were inconclusive AI enters the scene: Testing thousands of models at once NNPDF collaboration and neural network training The punchline: AI finds a better model with intrinsic charm Statistical significance: The three-sigma result Sponsor: LinkedIn Hiring Pro for small businesses Sponsor: Google Chrome and Gemini Sponsor: Cuvassier Cognac Five sigma standard and the future of this discovery The rapid pace of AI in experimental physics Conclusion: Cooperation of AI and human intelligence Outro: StarTalk promotion and comment responses Sponsor: LinkedIn career goals and office smells Sponsor: Cuvassier Cognac Sponsor: McDonald's Sausage McMuffin deal
Listen ad-free on Castria

Episode summaries and topics are generated from the audio by Castria — a free podcast app that skips ads automatically. IP Geolocation by DB-IP