Podcasts

Machine Learning Street Talk

The video argues that every historical model of the brain—from hydraulic pumps to modern computers—represents a "fallacy of misplaced concreteness" where useful technological metaphors are mistaken for literal biological reality, advocating instead for epistemic humility regarding whether nature is truly simple or merely intelligible through necessary human simplifications.

The Pomp Podcast

Jordi Visser argues that AI-driven deflation will crush abundant digital assets (like Big Tech software) while dramatically increasing the value of scarce assets like Bitcoin, which he views as the ultimate hedge against technological abundance rather than just inflation.

Latent Space

Brex CTO James Reggio details how the company is executing a three-pillar AI strategy—corporate adoption, operational automation, and product innovation—to transform into an agentic finance platform, achieving 5x growth with 99% burn reduction by deploying a hybrid architecture that pairs internal LLM infrastructure with a custom multi-agent framework.

My First Million

A.G. Sulzberger recounts stewarding The New York Times through existential digital disruption and unprecedented political attacks, detailing how radical transparency about business threats, a leaked innovation report, and clarifying unchanging core values enabled cultural transformation while maintaining editorial independence.

Stanford Online

This seminar introduces a robot motion planning framework that embeds physics-based PDE priors (specifically the Eikonal equation) into neural networks, achieving real-time inference in high-dimensional spaces while eliminating the need for weeks of expert data gathering required by conventional data-driven methods.

Interesting Times with Ross Douthat

Francisco Segovia, executive director of Minneapolis nonprofit COPAL, describes training hundreds of residents as 'constitutional observers' to legally document ICE arrests without obstructing agents, operating amid a tense street environment where residents use whistles to warn communities and armed encounters have recently turned fatal.

Stanford Online

Drawing on the Computers Are Social Actors (CASA) framework, this seminar explores how large language models function as uniquely anthropomorphic social agents, examining user privacy behaviors with AI companions and arguing for HCI interventions that address the asymmetrical risks of these corporate-owned yet socially-intimate relationships.

Huberman Lab

Dr. Wendy Suzuki reveals how aerobic exercise creates a 'neurochemical bubble bath' that stimulates BDNF, grows new hippocampal neurons, and sharpens prefrontal function, with research showing that just 2-3 weekly cardio sessions can significantly enhance memory and potentially add nearly a decade of healthy cognition later in life.

Conversations with Tyler (Tyler Cowen)

Renuka Sane explains India's transition from unsustainable defined-benefit pensions to market-linked defined-contribution schemes, highlighting how design flaws in the National Pension System (NPS) and political pressure from vocal government employee unions have driven a recent reversion toward hybrid models that risk repeating past fiscal mistakes.

My First Million

Stanford GSB founders working in ed-tech reveal a spectrum of cautious optimism about AI in education (rating its potential 3-7/10), emphasizing that while AI can scale personalized support and translation, it remains a "force multiplier" rather than a replacement for human connection, with significant risks around implementation gaps, equity divides, and regulatory fragmentation requiring deliberate policy intervention.

Cognitive Revolution

Snowflake is deploying enterprise AI by bringing models directly to governed data rather than exporting sensitive information, leveraging reasoning models to unlock the 80-90% of unstructured enterprise data previously trapped in documents and enabling reliable natural language analytics for non-technical business users.

AI Engineer

Auth0/Okta leaders Patrick Riley and Carlos Galan unveil new AI identity infrastructure including Token Vault for secure credential management and Async OAuth for human approvals, presenting a four-pillar framework to authenticate users and authorize autonomous agent actions across enterprise applications.