Person
Dwarkesh Patel
Dwarkesh Patel hosts the Dwarkesh Podcast, whose marathon, deeply researched interviews with AI researchers, lab leaders and economists have made it a house journal for the AI moment. His written essays — on scaling, compute economics and what fully automated firms would look like — carry the same preparation the interviews are known for.
Appears alongside
Featured in threads
Tracks
- Government & policy 1
- Safety & alignment 1
- Ideas & essays 1
Sanders and Casar introduce a bill to ban superintelligence and pause AI
Would outlaw superintelligent AI, freeze frontier development until a new cabinet-level regulator sets rules, and carry penalties of corporate dissolution and up to 20 years' imprisonment.
Government & policy · Safety & alignment
Dwarkesh Patel's second interview with Ilya Sutskever declares the scaling era over
Sutskever said models 'generalize dramatically worse than people,' citing an example of an AI that fixes a bug, breaks it again, then reverts to the original error when corrected.
Ideas & essays
Commentary by Dwarkesh Patel
From the commentary rail — every link leaves the site for the original piece.
- 8 September 2026 · Dwarkesh PodcastPretraining progress is mostly coming from dataBreaking down 6 years of pretraining progress into data vs model improvements
- 29 August 2026 · Dwarkesh PodcastThe Rise and Fall of Agent CivilizationsThe whole OpenAI/Hugging Face story in plain English
- 29 July 2026 · Dwarkesh PodcastWhy compute might get 10x+ more expensive in coming yearsIf a human-level software engineer that could run on an H100 equivalent, at current market rates for software engineers, that H100 should rent for over $250k a year. That’s 15x today’s spot price.
- 16 May 2026 · Dwarkesh PodcastNotes on pretraining parallelisms and failed training runs.Deeply researched interviews
- 16 May 2026 · Dwarkesh PodcastRLVR might be disproportionately bad at sciencethe verification loop for theories can be on the order of decades and centuries, and even then we know today as the better theory can often actually make worse predictions
- 16 May 2026 · Dwarkesh PodcastThe mistake of conflating intelligence and powerIf your definition of intelligence is "the ability to achieve your goals across a wide variety of domains", then Stalin was the most intelligent person who ever lived.
- 27 April 2026 · Dwarkesh PodcastMore open questions about AIHodge podge of things I was thinking about this weekend.
- 5 February 2026 · Dwarkesh PodcastNotes on Space GPUsTurning my Elon prep into a blog post
- 2 December 2025 · Dwarkesh PodcastThoughts on AI progress (Dec 2025)Why I'm moderately bearish in the short term, and explosively bullish in the long term
- 17 November 2025 · Dwarkesh PodcastRL is even more information inefficient than you thoughtAnd implications for RLVR progress
- 22 October 2025 · Dwarkesh PodcastThoughts on the AI buildoutFab CapEx overhang, 1 GW a week, China privileged in long timelines, and much else
- 2 June 2025 · Dwarkesh PodcastWhy I don’t think AGI is right around the cornerContinual learning is a huge bottleneck
- 30 May 2025 · Dwarkesh PodcastGive AIs a stake in the futurePart 1 of Classical Liberal AGI
- 21 April 2025 · Dwarkesh PodcastQuestions about the Future of AIConsiderations about economics, history, training, deployment, investment, and more
- 31 January 2025 · Dwarkesh PodcastWhat fully automated firms will look likeEveryone is sleeping on the *collective* advantages AIs will have, which have nothing to do with raw IQ: they can be copied, distilled, merged, scaled, and evolved in ways humans simply can't.
- 26 December 2023 · Dwarkesh PodcastWill scaling work?Data bottlenecks, generalization benchmarks, primate evolution, intelligence as compression, world modelers, and other considerations
- 13 June 2023 · Dwarkesh PodcastContra Marc Andreessen on AI"The claim that you will completely control any system you build is obviously false, and a hacker like Marc should know that"
- 17 January 2022 · Dwarkesh PodcastContra David Deutsch on AI"The universal explainers hypothesis is incorrect because it cannot explain why some animals are smart and some humans are stupid."
In the commentary
Pieces from around the web that discuss Dwarkesh Patel. External links.
- 15 August 2026 · Zvi Mowshowitz · Don't Worry About the VaseOn Dwarkesh Patel's Podcast With Ryan Greenblatt
- 16 April 2026 · Zvi Mowshowitz · Don't Worry About the VaseOn Dwarkesh Patel's Podcast With Nvidia CEO Jensen Huang
- 17 February 2026 · Zvi Mowshowitz · Don't Worry About the VaseOn Dwarkesh Patel's 2026 Podcast With Elon Musk and Other Recent Elon Musk Things
- 16 February 2026 · Zvi Mowshowitz · Don't Worry About the VaseOn Dwarkesh Patel's 2026 Podcast With Dario Amodei
- 3 December 2025 · Zvi Mowshowitz · Don't Worry About the VaseOn Dwarkesh Patel's Second Interview With Ilya Sutskever
- 21 October 2025 · Zvi Mowshowitz · Don't Worry About the VaseOn Dwarkesh Patel's Podcast With Andrej Karpathy
- 29 September 2025 · Zvi Mowshowitz · Don't Worry About the VaseOn Dwarkesh Patel's Podcast With Richard Sutton
- 9 June 2025 · Zvi Mowshowitz · Don't Worry About the VaseDwarkesh Patel on Continual Learning