AI + a16z

Martin Casado speaks with Ankur Goyal, founder and CEO of Braintrust, about where engineering actually matters in AI and where it doesn't. They cover the open source vs closed source model cycle, why Chinese models are gaining ground faster than spending suggests, whether AI demand will eventually saturate, and the Bash vs SQL benchmark that challenges the "just give it a computer" approach to agents.

Christopher Mims and Tim Higgins of the Wall Street Journal sit down with a16z General Partner Martin Casado on WSJ’s Bold Names to ask whether the AI spending boom is a bubble waiting to burst. Martin explains why the fundamentals differ dramatically from the dot-com era—when WorldCom had $40 billion in debt versus today's tech giants with hundreds of billions on their balance sheets—and why a speculative valuation correction shouldn't be confused with systemic collapse. They also discuss where a16z sees opportunity in the "long tail" of AI companies beyond the state-of-the-art large language models.

Mintlify is a documentation platform built by cofounders Han Wang and Hahnbee Lee to help teams create and maintain developer docs. In this episode, Andreessen Horowitz general partners Jennifer Li and Yoko Li speak with Han and Hahnbee about how coding agents are changing what “good docs” mean, shifting documentation from a human-only resource into infrastructure that powers AI tools, support agents, and internal knowledge workflows. They share Mintlify’s early journey, including eight pivots, the two-day prototype that landed their first customer, and the “do things that don’t scale” sales motion that helped them win early traction. The conversation also covers why docs go out of date, what “self-healing” documentation requires to actually work, and how serving fast-moving customers has shaped both their product priorities and their pace.

In this episode, Chief Legal & Policy Officer Jai Ramaswamy, Matt Perault, and a16z General Partner Martin Casado take a first-principles look at AI regulation, arguing that policymakers should focus on regulating harmful uses rather than model development until AI’s marginal risks are better understood. Drawing on decades of software governance debates, from encryption to cybersecurity, they contend that development-level rules are hard to define, easy to loophole, and likely to become obsolete in a fast-moving field where even the definition of “AI” remains unstable. They also unpack how regulatory uncertainty already creates real-world consequences by chilling open-source releases, advantaging well-resourced incumbents over startups, and pushing the next generation of builders and researchers toward Chinese open models, making the case for evidence-based, technology-neutral policy that protects against bad behavior without stifling innovation.

Nvidia’s $5 billion investment in Intel is one of the biggest surprises in semiconductors in years. Two longtime rivals are now teaming up, and the ripple effects could reshape AI, cloud, and the global chip race. To make sense of it all, Erik Torenberg is joined by Dylan Patel, chief analyst at SemiAnalysis, joins Sarah Wang, general partner at a16z, and Guido Appenzeller, a16z partner and former CTO of Intel’s Data Center and AI business unit. Together, they dig into what the deal means for Nvidia, Intel, AMD, ARM, and Huawei; the state of US-China tech bans; Nvidia’s moat and Jensen Huang’s leadership; and the future of GPUs, mega data centers, and AI infrastructure.

What if the next leap in artificial intelligence isn’t about better language—but better understanding of space? In this episode, a16z General Partner Erik Torenberg moderates a conversation with Fei-Fei Li, cofounder and CEO of World Labs, and a16z General Partner Martin Casado, an early investor in the company. Together, they dive into the concept of world models—AI systems that can understand and reason about the 3D, physical world, not just generate text. Often called the “godmother of AI,” Fei-Fei explains why spatial intelligence is a fundamental and still-missing piece of today’s AI—and why she’s building an entire company to solve it. Martin shares how he and Fei-Fei aligned on this vision long before it became fashionable, and why it could reshape the future of robotics, creativity, and computational interfaces. From the limits of LLMs to the promise of embodied intelligence, this conversation blends personal stories with deep technical insights—exploring what it really means to build AI that understands the real (and virtual) world.

Matt Knight spent five years as OpenAI’s CISO. Now he runs what colleagues call “the most interesting job at the company”: leading Aardvark, an AI agent that finds security vulnerabilities the way a human researcher would—by reading code, writing tests, and proposing patches. It recently found a memory corruption bug in OpenSSH, one of the most heavily audited codebases in existence.