ABAB Pedia

Dwarkesh Patel

Dwarkesh Patel is a podcast host and author who hosts the long-form interview show Dwarkesh Podcast, covering fields such as artificial intelligence, technology, history, and economics. He began making the podcast in 2020 while in college, was named to the TIME100 AI list in 2024, and compiled and published The Scaling Era: An Oral History of AI, 2019–2025 with Gavin Leech.

Contents13 sections
Key facts

Early Life and Education

Dwarkesh Patel lived in India in his early years and came to the United States with his family as a child, living in several states before enrolling at the University of Texas at Austin to study computer science.⁠[1]

In college, he broadened his interests by reading books on economics and history and began reaching out to authors. After reading Bryan CaplanBryan CaplanBryan Caplan (Bryan D. Caplan) is an American economist and professor of economics at George Mason University whose fields include public economics, public choice and psychology and economics; he has written several books on economics and public policy.Open the full entry’s The Case Against Education, he emailed Caplan to express his interest in the book and then invited him for an interview.⁠[1]

Launching The Lunar Society Podcast

On May 22, 2020, Patel released an interview with the economist Bryan Caplan, which the episode page marks as the podcast’s first episode. At the time, he was producing the show, called The Lunar Society, from his college dorm, discussing education, economics, and social issues.⁠[2][1]

Continuing the Podcast After Graduation

In December 2021, Patel graduated from the University of Texas at Austin with a degree in computer science. At graduation, he was still looking for a long-term career direction.⁠[1][4]

He later wrote in a review of his work that an early listener, Anil Varanasi, gave him a $10,000 grant that allowed him to work full time on the podcast and writing for six months; he later received another small grant. Even while continuing the show, he at one point considered ending the podcast to start a company.⁠[4]

Interviewing Ilya Sutskever

Patel said in a later interview that, to prepare for this conversation, he implemented a Transformer himself. He regards extended reading and technical preparation as an important part of his interviewing method.⁠[1]

Interviewing Mark Zuckerberg and Growing the Show as a Business

On April 18, 2024, Patel released an interview with Meta CEO Mark ZuckerbergMark ZuckerbergMark Zuckerberg is the founder of Facebook, Chairman and CEO of Meta, and a co-founder of the Chan Zuckerberg Initiative. His work spans social networking, virtual reality, and artificial intelligence.Open the full entry, centered on Llama 3, open models, custom chips, synthetic data, and the energy constraints on scaling compute.⁠[6]

Looking back on this interview in 2025, Patel said that negotiating ads around the episode made him realize that revenue could support full-time editors and that the podcast could become a long-term business.⁠[7]

Explaining His Topic Selection and Research Method

In August 2024, Patel wrote in “Dwarkesh Podcast Progress Update” that when choosing guests he cares more about whether he is willing to spend two weeks researching their field than about whether the guest is well known. He cited his episodes with the historian Sarah Paine as an example showing that specialist scholars can also attract wide attention.⁠[4]

He also discussed how advertising, editing, and coordinating recordings took time away from research, and hoped to free up more time for reading and writing by dividing up the work. He believes that continuous learning and forming his own questions and judgments are the foundation for improving the show’s quality.⁠[4]

Named to the TIME100 AI

Publishing The Scaling Era

On March 25, 2025, Stripe Press announced the ebook release of The Scaling Era: An Oral History of AI, 2019–2025 and opened preorders for the print edition. Compiled by Patel and Gavin Leech, the book draws on the podcast’s interviews to present discussions of AI development during this period.⁠[9]

In a Q&A episode from the same period, Patel said that when choosing guests he considers whether he is willing to spend one to two weeks reading their work and understanding their research, rather than considering only their influence.⁠[7]

Discussing Continual Learning as a Limit on AI

On June 2, 2025, Patel published “Why I don’t think AGI is right around the corner,” drawing on his own attempts to use language models for podcast post-production work to argue that a lack of continual learning ability would limit models’ capacity to take on complex work.⁠[11]

He distinguished between completing a single task and the ability to absorb feedback and accumulate work experience over time, and acknowledged that reasoning models and coding tools had improved. He framed the timeline forecasts in the piece as personal judgments with great uncertainty.⁠[11]

Interviewing Jensen Huang

The two also discussed selling AI chips to China. Patel said he did not yet have a settled position and raised counterarguments from the perspective of national security risks, while Huang laid out his position in favor of having developers worldwide use the American technology stack.⁠[12]

Continuing to Interview AI Researchers

On September 11, 2026, Patel discussed recursive self-improvement, generalization, and research automation with John Schulman, Beren Millidge, and Charlie O’Neill. The episode used a multi-person conversation to present the researchers’ judgments and disagreements on these questions.⁠[13]

Public Accounts and Works

His representative works include the long-form interview show Dwarkesh Podcast and The Scaling Era: An Oral History of AI, 2019–2025, created with Gavin Leech.⁠[10]

Sources

  1. The future belongs to those who prepare like Dwarkesh Patel (opens in a new window)
  2. Bryan Caplan — Nurturing Orphaned Ideas (opens in a new window)
  3. Tyler Cowen — The Great Reset (opens in a new window)
  4. Dwarkesh Podcast Progress Update (opens in a new window)
  5. Ilya Sutskever — Why next-token prediction could surpass human intelligence (opens in a new window)
  6. Mark Zuckerberg — Llama 3, $10B models, Caesar Augustus, & 1 GW datacenters (opens in a new window)
  7. AMA: career advice given AGI, how I research ft. Sholto & Trenton (opens in a new window)
  8. Dwarkesh Patel: The 100 Most Influential People in AI 2024 (opens in a new window)
  9. Preorder The Scaling Era: An Oral History of AI, 2019–2025 (opens in a new window)
  10. The Scaling Era: An Oral History of AI, 2019–2025 (opens in a new window)
  11. Why I don’t think AGI is right around the corner (opens in a new window)
  12. Jensen Huang — TPU competition, why we should sell chips to China, & Nvidia’s supply chain moat (opens in a new window)
  13. AI researchers debate how close we are to recursive self-improvement (opens in a new window)
  14. Noam Brown — Agent swarms, alignment, & recursive self-improvement (opens in a new window)
  15. About — Dwarkesh Podcast (opens in a new window)
  16. Dwarkesh Patel (@dwarkesh_sp) (opens in a new window)