Dwarkesh Podcast

Terence Tao, UCLA mathematician and one of the era’s top problem-solvers, explores Kepler’s surprising path to planetary motion and why great theories can look worse before they win. He gets into AI flooding science with ideas, the bottleneck of verification, why breadth beats depth for now, and how writing, persuasion, serendipity, and human-AI teamwork shape discovery.

Dylan Patel, founder and chief analyst at SemiAnalysis, maps the real choke points behind AI compute growth. He gets into why old H100s can get more valuable, how Nvidia locked in TSMC capacity early, why memory may be the nastiest crunch ahead, and why ASML could become the limiting factor by 2030. They also touch on power buildouts, China timelines, robots, and Taiwan risk.

A deep look at the Anthropic–Pentagon dispute and why it matters for AI’s role in society. Discussion of how AI makes mass surveillance cheap and ubiquitous. Examination of government levers like supply-chain designations and procurement power. Debate over who AI should be aligned to and whether regulation hands power to a despot.

Ada Palmer, Renaissance historian, novelist, and composer at the University of Chicago. She explains how Italian city-republics and Roman cosplay shaped politics and learning. She traces printing’s real revolution from books to pamphlets and how distribution made Gutenberg fail. She links libraries, networks, and inquisitorial labs to the rise of shared scientific practices.

Dario Amodei, AI researcher and CEO of Anthropic, renowned for work on large-scale models and AI risk. He discusses why scaling and task-specific RL may generalize, how AI could diffuse through the economy, timelines for near-AGI-like capabilities, Anthropic’s compute and profitability choices, and governance and geopolitical risks tied to powerful models.

Elon Musk, entrepreneur and CEO of SpaceX and Tesla, leading large-scale projects in space, energy and AI. He discusses why space could become the cheapest place for AI within 36 months. He explains power limits on Earth and scaling solar and launches. He covers scaling humanoid robots, xAI’s approach to alignment and business, and chip and memory bottlenecks for massive compute.

Adam Marblestone, CEO of Convergent Research and former Google DeepMind scientist, delves into the fascinating intersection of neuroscience and AI. He explains how the brain learns efficiently from minimal data and argues that the secret lies in complex reward functions shaped by evolution. Marblestone highlights how the brain’s predictive cortex contrasts with AI's next-token predictions, and discusses the cellular diversity linked to specific behaviors. The conversation also touches on the implications of this understanding for improving AI systems, including using insights from AI to unlock neural mechanisms.

The discussion delves into the complexities of AI progress and the limitations of current robotics. It highlights skepticism around automated researchers and the challenges of achieving human-like continual learning. The concept of scaling in reinforcement learning is examined, alongside the significant compute needs for advancements. Predictions for the future include the potential for brain-like intelligences and the need for efficient training methods. Lastly, the importance of competition in driving innovation is emphasized.

Sarah Paine, a political scientist and historian specializing in Russian affairs, discusses the multifaceted reasons behind the Soviet Union’s collapse. She explores Reagan's military strategies, the impact of the Helsinki Accords on Eastern Bloc dissidents, and the Sino-Soviet split. Paine highlights internal issues like economic failures and political reforms under Gorbachev. She also delves into Boris Yeltsin's pivotal role in the dissolution of the USSR. With a nod to current geopolitical tensions, she offers insights on navigating potential new conflicts.

Join Ilya Sutskever, co-founder of OpenAI and Stability AI, as he dives into the world of AI and machine learning. He discusses the intriguing concept of model jaggedness, explaining why AI sometimes behaves inconsistently. Ilya contrasts pre-training and reinforcement learning, emphasizing the importance of generalization and the barriers faced by AI. He also explores how emotions can serve as value functions and proposes new strategies for ensuring AGI is aligned with human values. Insights on continual learning and the future of superintelligence add depth to this fascinating conversation.