We help founders make something people want. The Y Combinator Podcast is where builders talk about building. From the earliest days of an idea to scaling a company that changes the world, YC partners and founders share real stories, lessons, and tactics from the frontlines.
Claude Code creator Boris Cherny details the paradigm shift in building agentic AI products, emphasizing that modern models like Opus 5 require an empirical, "unhobbling" approach rather than rigid, over-specified scaffolding. By deleting legacy system prompts and allowing models to operate autonomously, developers can...
NVIDIA CEO Jensen Huang details the company's evolution from a flawed 3D graphics startup to a global leader in accelerated computing. Success stems from a relentless commitment to first-principles thinking, where the focus remains on accelerating algorithmic domains rather than just building chips. Huang emphasizes th...
Sample efficiency remains a critical bottleneck in artificial intelligence, as current models require vast datasets compared to the rapid skill acquisition observed in humans. World models offer a promising solution by enabling agents to simulate environments and predict future states, effectively reducing the need for...
Aggregate user metrics often obscure individual behavior, leading founders to misinterpret product health. Dot plots solve this by mapping individual users across time periods, using specific events that signify value creation rather than superficial engagement. This visualization technique reveals granular patterns—su...
Bryant Chou, co-founder and CTO of Webflow, introduces Ploy, an AI-driven platform designed to democratize digital marketing and web development for small businesses. Ploy functions as a "company brain" by integrating with existing systems to automate SEO, content generation, and design consistency. By "slurping" legac...
Founders often struggle with "idea paralysis," waiting for the perfect startup concept before committing. Meaningful progress requires abandoning the search for perfection and instead selecting one idea to pursue with singular focus. "Going deep" involves immersing oneself in the domain until one can effectively run th...
AI integration requires a fundamental shift in business architecture where CEOs act as chief AI officers, prioritizing agentic workflows over traditional software development. Successful AI adoption hinges on treating agents as virtual employees capable of self-bootstrapping through tools and markdown configurations, r...
AI-native service companies represent a trillion-dollar opportunity to rebuild traditional industries like tax, audit, and law by utilizing AI to deliver outcomes rather than merely providing software tools. Ideal markets for these ventures exhibit low task-level judgment, high intelligence requirements, and regulatory...
Building superintelligence within an organization requires transitioning from using AI as a mere copilot to treating it as the foundational layer for all operations. By centralizing internal data and context, companies empower employees to encode complex workflows through natural language rather than rigid, determinist...
Modern corporate governance often forces founders into a cycle of value destruction by prioritizing short-term shareholder returns over long-term mission. This "shareholder primacy" is a recent, non-legislated normative consensus rather than a legal requirement, yet it frequently leads to the ousting of visionary leade...
Ambitious startup founders should move to Silicon Valley to access superior peer networks, benefit from high-frequency serendipitous encounters, and experience a unique "pay-it-forward" culture that accelerates decision-making. While local ecosystems often lack this density, spending time in a global hub provides a nec...
Garry Tan, CEO of Y Combinator, details his return to software engineering through the use of agentic AI systems like Claude Code and his self-developed GStack. This approach, termed "tokenmaxxing," involves aggressively utilizing AI to handle complex coding, research, and testing tasks, enabling a single developer to ...
Recursion offers a path to improve AI reasoning performance by enabling iterative computation at inference time rather than relying solely on increasing model size. Traditional LLMs operate as one-shot feed-forward processes, which limits their ability to solve incompressible problems like sorting or complex puzzles. H...
Artificial General Intelligence (AGI) development hinges on solving fundamental challenges in continual learning, long-term reasoning, and memory, with agent-based systems serving as the primary architecture for progress. While current large-scale models provide a foundation, achieving AGI requires moving beyond brute-...
Replit empowers individuals to build production-ready software through natural language, effectively bridging the gap between complex developer tools and consumer-friendly creation. By abstracting away technical infrastructure, the platform allows domain experts—from physical therapists to operations managers—to develo...
AI is fundamentally transforming startup operations, shifting from a mere productivity tool to a core operating system. Startups must transition from "open loop" systems to "closed loops" where every process is captured, analyzed, and improved by an intelligence layer. This requires making the entire organization "quer...
Stripe’s Head of Design, Katie Dill, details the strategic redesign of the company’s homepage, transitioning from a 2020 version to a sophisticated platform reflecting Stripe's evolution into a multinational financial infrastructure provider. The new site utilizes a "bento box" layout with progressive disclosure modals...
Physical Intelligence co-founder Quan Vuong details the shift toward general-purpose robotics, driven by the emergence of foundation models capable of controlling diverse hardware. By decoupling high-level semantic planning from low-level control, the industry is moving away from rigid, vertically integrated systems to...
BillionToOne's journey from a prenatal genetic testing startup to a multi-billion dollar company is explored, highlighting their innovative approach to DNA diagnostics. The company's co-founders, Oguzhan Atay and David Tsao, discuss their unique technology that uses synthetic DNA to reduce noise in blood samples, enabl...
Variance is coming out of stealth with a $21 million Series A to build purpose-built AI agents for risk and compliance, automating content, fraud, and identity reviews at scale. Karine Mellata, co-founder of Variance, explains that their AI agents help companies like GoFundMe verify fundraisers, ensuring money goes to ...
Follow this podcast in Podwise
Sign in to get AI summaries, transcripts and mind maps for any episode, including new ones.
