Wiki / People
Dwarkesh Patel
Dwarkesh Patel is a writer and podcast host known for long, carefully prepared interviews with AI researchers, founders, scientists, and historians. His main job is running the Dwarkesh Podcast: research a topic deeply, then sit with the world expert and ask the questions that matter. In 2025 he co-authored The Scaling Era, an oral history of the AI boom, with Stripe Press.
Quick facts
- Born August 19, 2000, Vadodara, Gujarat, India; moved to the U.S. at age 8
- Education Computer science, University of Texas at Austin (graduated 2021)
- Job Full-time podcast host, writer, and interviewer (formerly The Lunar Society)
- Home base dwarkesh.com (Substack, YouTube, Apple Podcasts, Spotify)
- Book The Scaling Era: An Oral History of AI, 2019-2025 (with Gavin Leech)
- Known for Getting frontier AI leaders to talk in public with unusual technical depth
What he does
Patel’s contribution is not building models. It is building the record of how the people who do build them think. He prepares like a student preparing for an oral exam, then runs interviews that mix scaling laws, alignment, compute, economics, and history.
The show started in college in 2020 as The Lunar Society, named after the 18th-century Birmingham club of inventors and scientists. He later renamed it Dwarkesh Podcast after listeners kept confusing “Lunar” with crypto slang. Early guests said yes because the questions were serious. Later guests said yes because earlier ones told them to.
Outside the mic, he writes essays on compute, AGI timelines, and how intelligence and power get confused. The podcast and writing sit in the same project: make the AI era legible while it is still happening.
Books
- The Scaling Era: An Oral History of AI, 2019-2025 (2025) - with Gavin Leech, published by Stripe Press. Curated and annotated conversations from the podcast, plus definitions, charts, classic essays, and some unpublished interviews. Guests and material include voices from Anthropic, DeepMind, OpenAI, Meta, MIRI, and Open Philanthropy.
As of 2026, this is his published book. The rest of his corpus is podcast episodes and essays on dwarkesh.com.
Podcast
- Site / archive dwarkesh.com · full archive
- YouTube search “Dwarkesh Podcast” (primary video home)
- Apple Podcasts / Spotify linked from the Substack home page
- Style multi-hour, deeply researched, often technical; AI-heavy but also history, math, science, and ideas
Orbit and relationships
Patel’s closest working relationships are with the people who keep showing up in his feed: AI lab leaders, safety researchers, and public intellectuals who treat long-form conversation as real work. He is not a lab employee. His access comes from reputation for preparation. Co-author Gavin Leech (Arb Research) shaped The Scaling Era with him. The Stripe Press relationship puts the oral history next to other progress-focused books.
Guests often recommend other guests. That network effect is the social graph of the show: Anthropic, DeepMind, OpenAI/SSI, Nvidia, Meta, and independent researchers all sit in overlapping conversations about scaling, alignment, and what comes after cheap intelligence.
Selected guests (with sites)
A sample of notable guests, linked to their org or personal site (not the episode page). Full list lives in the podcast archive.
- Dario Amodei - CEO, Anthropic
- Demis Hassabis - CEO, Google DeepMind
- Ilya Sutskever - Safe Superintelligence (OpenAI cofounder)
- Andrej Karpathy - AI researcher and educator
- Jensen Huang - CEO, Nvidia
- Mark Zuckerberg - CEO, Meta
- Sam Altman - CEO, OpenAI
- Satya Nadella - CEO, Microsoft
- Eliezer Yudkowsky - MIRI; AI alignment writer
- Ajeya Cotra - Open Philanthropy (appears in book material)
- Grant Sanderson - 3Blue1Brown
- Terence Tao - mathematician, UCLA
- Michael Nielsen - scientist and writer on research progress
- David Reich - geneticist, Harvard
- Ada Palmer - historian and novelist
- Tony Blair - former UK PM; Tony Blair Institute
Why he is interesting
- He treats interviewing as a research method, not celebrity access.
- He sits between technical AI and public understanding without dumbing either side down.
- The Scaling Era freezes a chaotic half-decade of AI into a single oral history while the people who lived it are still talking.
- The guest graph is a map of who currently shapes frontier AI and adjacent science.