The End of the Designer–Engineer Divide

Why Domain Experts Are Winning In The Age Of AI

Jun 19, 202642 minSummary

In this episode, the hosts are joined by Bryant Cho, co-founder and CTO of Webflow, to discuss his latest venture, Ploy. The conversation centers on how experienced domain experts are leveraging artificial intelligence to build transformative, high-quality products in an increasingly crowded market. Ploy is presented as an AI-driven marketing platform that goes beyond simple website creation. By integrating design, coding, and marketing logic, the tool helps businesses build professional websites and manage growth on autopilot. A key theme of the discussion is the concept of Founder-Market Fit; Cho explains how he has channeled his decade of experience in web design and growth strategy into the platform to solve the complex, often arduous tasks that founders typically struggle with, such as SEO optimization and data-driven marketing. The episode explores the shift from human-built digital infrastructure to AI-assisted development, with the hosts noting that these models act as factories for human creativity. Cho emphasizes that while AI provides the "boundless intelligence," the real competitive advantage lies in having the expertise to steer those models toward world-class, coherent, and bespoke outcomes, ultimately helping founders better tell their stories and connect with customers.

How To Pick A Startup Idea

Jun 17, 202611 minSummary

In this episode of the Y Combinator Startup Podcast, the host addresses a common hurdle for entrepreneurs: the paralysis of choosing a startup idea. Many founders waste time waiting for the perfect concept or juggling multiple projects simultaneously, failing to realize that true clarity only comes from engaging with reality and customer feedback. The episode emphasizes that overthinking is a major pitfall. Instead of seeking the perfect idea or obsessing over founder-market fit, the host advises founders to pick an idea they are curious about and commit to it fully. This process requires burning all other bridges and developing such deep domain expertise that one could effectively run a customer’s business or teach a class on the subject. By doing so, founders gain the actionable data needed to determine if their concept is viable. Furthermore, the discussion highlights that in the AI era, the most promising ideas are often those that verticalize—selling outcomes rather than just software—and aim for the most ambitious version of the solution. Ultimately, even if an initial idea fails, the deep work invested positions the founder to discover a more significant, hidden opportunity underneath.

"The CEO Must Be the Chief AI Officer"

Jun 10, 202654 minSummary

In this episode, the hosts are joined by Pedro Franceschi, co-founder and CEO of Brex, to discuss why the modern CEO must function as the Chief AI Officer. Franceschi emphasizes that AI integration is not merely a task for engineering or product teams, but a fundamental shift in company identity that requires leaders to intimately understand the boundaries and capabilities of the technology. The conversation explores the evolution of AI tools, with Franceschi describing the arrival of advanced reasoning models as a transformative moment comparable to the invention of electricity. He shares his insights on building "harnesses"—the frameworks and agentic loops that allow AI to perform complex tasks autonomously. A key highlight is the discussion of Brex’s approach to security, specifically how they utilize network-layer proxies and LLMs as judges to safely enable AI agents to operate within sensitive financial environments. Throughout the discussion, Franceschi warns against the "curse of knowledge" and reminds founders that while AI excels at execution, the essential human task remains identifying the right problems to solve. He advocates for maintaining a minimalist surface area in early-stage products, using AI to deepen insights rather than to avoid the difficult work of customer discovery.

How to Build an AI-Native Services Company

Jun 3, 202611 minSummary

The podcast episode explores the burgeoning trend of AI-native services companies, arguing that the next generation of massive businesses will not be traditional software companies, but service providers—such as law firms, tax agencies, and insurance carriers—rebuilt from the ground up using artificial intelligence. Unlike software that provides tools for internal use, these companies deliver specific outcomes to customers, effectively displacing legacy vendors. The hosts outline a strategic playbook for founders, emphasizing the importance of identifying markets with specific traits: low trust in existing processes, automatable tasks that require minimal human judgment, high intelligence requirements, and strict regulatory environments. Key success factors include maintaining operational rigor, prioritizing throughput and cycle time as core product metrics, and ensuring AI operating leverage improves margins over time. A major warning is issued against the early demand trap, where founders sign too many pilot customers before establishing the necessary infrastructure, often leading to a reliance on human labor to mask product shortcomings. By focusing on building scalable processes rather than just software, founders can aim for the high margins typical of software startups within the massive scale of the professional services market.

How To Build Superintelligence Inside Your Company

May 27, 202646 minSummary

In this episode of the Y Combinator Startup Podcast, host Garry Tan and YC partner Pete Kuhn discuss the internal journey of transforming YC into an AI-native organization. Moving beyond the "copilot" model, they explore how to leverage AI as a fundamental building layer for an entire company, creating what they describe as a shared organizational brain. The conversation centers on the development of internal agent infrastructure, which began as a way to streamline finance workflows and evolved into a powerful suite of over 350 tools. A key insight is the importance of consolidating organizational data into a single, accessible context layer, such as a unified database. By granting agents access to this information, employees can pose complex, high-level questions that were previously too time-consuming to answer manually. The hosts also emphasize the necessity of building internal tool and skill registries, enabling agents to self-improve through feedback loops and meta-prompting. They argue that by defaulting to internal transparency regarding AI interactions, companies can foster a high-trust environment that balances security with the immense power of intelligent automation. Ultimately, they suggest that this approach allows organizations to synthesize the collective knowledge and instincts of their teams, making every individual more effective.

How The Best Companies Defend Against Mediocrity And Rot

May 25, 202650 minSummary

In this episode of the Y Combinator Startup Podcast, Eric Ries, author of The Lean Startup, joins the hosts to discuss the themes of his book, Incorruptible: Why Good Companies Go Bad and How Great Companies Stay Great. The conversation centers on the vulnerability of successful startups and the systemic risks posed by modern corporate governance practices. Ries argues that many founders are led astray by "best practices" that prioritize short-term shareholder value at the expense of long-term mission and product quality. He highlights how concepts like shareholder primacy often result in mission-driven founders being ousted, leaving their companies to lose their innovative edge or succumb to short-sighted management. Through historical case studies—such as the rise of Costco and the cautionary tale of Polaroid—Ries illustrates how companies can defend against this "rot" by establishing a "governance fortress." He advocates for structural changes, such as adopting Public Benefit Corporation (PBC) status, to ensure a company’s mission remains the primary fiduciary duty, ultimately suggesting that founders must build organizations that are mission-controlled rather than exclusively investor-controlled to ensure long-term, sustainable impact.

The AI Agent Economy Is Here

Inside Claude Code With Its Creator Boris Cherny

The New Way To Build A Startup

OpenClaw And The Future Of Personal AI Agents