In recent years there’s been an influx of theoretical physicists into the leading AI labs. Do they have unique capabilities suited to studying large models or is it just herd behavior? To find out, we talked to our former AI Fellow (and now OpenAI researcher) Dan Roberts.

Martin and Spielman discuss the potential for source-grounded AI, and how they created such delightful AI podcast hosts.

Sridhar discusses Snowflake's advancements in AI, democratizing AI usage, reliable data applications, and their efforts to achieve high accuracy in AI solutions.

Noam and his research team describe project Strawberry, inference-time compute and the launch of OpenAI o1 which teaches LLMs how to reason better by thinking longer.

Vlad and Tudor founded Harmonic with a mission both lofty—mathematical superintelligence—and imminently practical, verifying all safety-critical software.

Jim has spent his career pursuing embodied intelligence. At Nvidia he leads the group with the audacious task of training foundation models that span real-world robotics and virtual world agents.

In 2022 Eric realized that AGI was closer than he had previously thought and started Magic to automate the software engineering necessary to get there.

Clay describes how companies can capture their brand voice, values and internal processes to create AI agents that are delightful and truly represent the business.

Jim discusses the challenges of AI readiness in industrial settings and how we have to build on top of the control systems of the 70s and 80s to achieve the promise of the Fourth Industrial Revolution.

Fireworks Founder and CEO Lin Qiao, previously leader of the PyTorch team at Meta, discusses how fast inference and small models will benefit startups and enterprises.

GitHub CEO Thomas Dohmke discusses how a small team at GitHub built Copilot on top of GPT-3 and quickly created a product that developers love—and can’t live without.

Joe talks about the launch of Llama 3.1 405B and tells us why he thinks even frontier models will ultimately commoditize—and why that’s a good thing for the startup ecosystem.

Co-founder and CEO Sebastian Siemiatkowski tells how Klarna shipped its customer service assistant and how experimenting with AI is transforming the company.

Misha Laskin, co-founder of Reflection AI, talks about what we can learn from AlphaGo and Gemini to train the most reliable models for developers building agentic workflows.

Microsoft CTO Kevin Scott discusses the shift across the ecosystem to more inference compute as the frontier models continue to improve, serving wider and more reliable use cases.

Zapier co-founder and head of AI Mike Knoop talks about ARC Prize, a $1M+ competition to solve ARC-AGI, a benchmark that measures the ability to efficiently acquire new skills.

Factory’s Matan Grinberg and Eno Reyes are building a fleet of purpose-built agents designed to accomplish different tasks in the software development lifecycle, like code review or testing.

LangChain’s Harrison Chase explains custom cognitive architectures that allow agents to improve performance and find traction in the sweet spot of autonomy.