Show Notes
François Chollet has spent years asking a different question than most of the AI world. Instead of scaling what already works, he’s trying to understand what intelligence actually is—and how to build it from first principles. In this episode of Lightcone, he traces that path from his early work on deep learning to the creation of the ARC prize, and the launch of ARC V3, a new benchmark designed to measure something deeper than performance: the ability to learn, adapt, and reason efficiently in entirely new environments. He explains why today’s systems may be hitting limits, what recent breakthroughs really mean, and why reaching true general intelligence may require a fundamentally different approach.00:00 - AGI by 2030?00:31 - Introducing Ndea: A New Path Beyond Deep Learning01:08 - A New ML Paradigm 01:30 - Replacing neural nets with compact symbolic programs03:04 - Why Ndea Isn’t Competing With Coding Agents05:20 - Why Everyone Might Be Wrong About Scaling LLMs07:22 - Why Coding Agents Suddenly Work So Well08:50 - The Limits of LLMs in Non-Verifiable Domains10:48 - What AGI Actually Means (And Why Most Definitions Are Wrong)13:30 - Why Deep Learning Hits a Wall 14:00 - ARC’s Origin Story18:20 - ARC Benchmarks Explained: From V1 to V322:49 - The RL Loop Powering Coding Agents Today27:03 - ARC-AGI V3: Measuring “Agentic Intelligence”31:14 - Inside the ARC Game Studio35:31 - Could AGI Fit in 10,000 Lines of Code?44:01 - Building Ndea: From Idea to Compounding Research Stack46:46 - The Future of ARC: Benchmarks That Evolve With AI47:21 - Why There’s Still Huge Opportunity for New AI Paradigms53:37 - How to Build a Breakout Open Source Project - Lessons From Kera56:39 - Advice For How To Think About AIApply to Y Combinator: https://www.ycombinator.com/applyWork at a startup: https://www.ycombinator.com/jobs
More from Y Combinator Startup Podcast
Magic Patterns (YC W23) is one of the fastest-growing design tools in AI, helping product teams turn ideas into interactive, code-backed prototypes in minutes. Alex Danilowicz and his co-founder, both front-end engineers, built Magic Patterns to solve a problem they had themselves: finding a better way to turn ideas into something visual. They were among the first teams building prompt-to-design tools back in 2023.In this episode of Design Review, Alex joins YC’s Aaron Epstein to demo Magic Patterns and its new MagicOS canvas, where teams can bring together prompts, images, PRDs, and other context to explore ideas visually. Then they review a handful of startup websites to break down what makes a site feel “vibe coded,” where founders go wrong, and how to make AI-generated design feel more intentional.Chapters:00:00 - What Magic Patterns is03:01 - Nine months of pivots05:59 - Using the MCP from Cursor08:21 - The end stage is compressed now11:23 - Demo: postcards for every meetup city15:35 - The RollerCoaster Tycoon chocolate theme park18:17 - Site 1: Ego21:39 - Inputs you can't type into23:59 - Site 2: Sentient OS27:57 - "Free" should be giant30:20 - Instrument Serif and the em dash32:01 - Site 3: CeraNovus36:59 - Site 4: Slashy38:36 - Nobody clicks your tabs43:08 - Getting better at design as an engineer46:07 - What's next: Magic OSApply to Y Combinator: https://www.ycombinator.com/applyWork at a startup: https://www.ycombinator.com/jobs
One of the biggest surprises in AI over the last few years has been how well coding agents generalize beyond software. In a recent essay, MIT professor Philip Isola argued that we may be entering the era of robot-use agents: general-purpose models that can control different robots, write policies, and learn new physical tasks with little or no robot-specific training.In this episode of Decoded, we're joined by the founders of Waddle Labs and RoboCurve, two of the startups whose work helped drive this realization. They're working at the frontier of using general-purpose models to control robots, and their recent demos helped inspire the growing conversation around robot-use agents. Together, we dig into the research behind that idea, from code-as-policies and vision-language-action models to the harnesses and evals needed to make these systems work in the real world.
YC works with thousands of founders every year, which gives us an early look at how startups are changing. Right now, the shift is striking: startups are moving from bits to atoms, nearly one in five YC companies has a solo founder, and companies are reaching meaningful revenue faster than ever.In this episode of The Lightcone, Garry, Jared, Diana, and Harj dig into what’s driving these changes and what they mean for founders. They discuss how AI is making it possible for smaller teams to take on more ambitious problems, why experienced founders are having a resurgence, and why knowing what to build is becoming more important than simply knowing how to build it.
The best outbound starts before you automate anything. Founders need to learn who actually has the problem, what gets their attention, and what makes them reply. YC Visiting Partner Christina Gilbert recommends doing at least your first 100 outreaches by hand so you can learn those things yourself.
Ollama (YC W21) is used by 9 million developers and 85% of the Fortune 500, giving co-founder and CEO Jeffrey Morgan a unique view into which AI models people are actually using and how that’s changing.Right now, the biggest shift he sees is toward open models, driven by coding agents, falling costs, and capabilities that are rapidly catching up to the frontier labs. On Ollama Cloud, that shift has driven a 150x increase in token usage since the start of the year.In this episode of the Lightcone, Jeff joins us to talk about the future of open models and the story behind Ollama, from two years of searching for the right idea to building one of the most widely used AI developer tools in the world.