a16z Podcast

→ WHAT IT COVERS OpenAI's Ari Weinstein and James Sun join a16z's Sima Amble and Fabrizio Serafini to examine how computer-using agents now operate faster than average humans, where enterprise and consumer automation is headed, and how companies can deploy agents across legacy software while maintaining audit controls and human oversight. → KEY INSIGHTS - **Speed threshold:** Computer-using agents have surpassed average human speed and are approaching expert-level performance.

→ WHAT IT COVERS Box CEO Aaron Levie explains how AI agents give 3-to-10-person startups the output capacity of 50-to-200-person teams, why strategy and taste now matter more than execution speed, and why distribution remains the one part of the startup playbook that AI has not fundamentally changed. → KEY INSIGHTS - **Agent-as-leverage model:** Founders who shift from writing code to reviewing and orchestrating AI-generated code can achieve 3x-to-10x engineering output.

→ WHAT IT COVERS a16z co-founder Ben Horowitz and Andrew Chen examine what AI changes and what it doesn't about building companies. They cover AI's unique adoption advantages over the internet era, defensibility without network effects, why small teams can reach billion-dollar scale, and the psychological demands that technology cannot automate for founders.

→ WHAT IT COVERS Amazon CEO Andy Jassy explains why Amazon commits $220 billion in 2026 CapEx to AI infrastructure, why AWS could reach a $1 trillion annual run rate, how inference costs must drop an order of magnitude to unlock enterprise adoption, and how AI is forcing Amazon to rebuild products, teams, and organizational structures from scratch. → KEY INSIGHTS - **AI demand vs. supply reality:** Amazon's $220 billion 2026 CapEx still leaves meaningful demand unmet.

→ WHAT IT COVERS a16z announces the expansion of its crypto fund into "a16z Frontier," broadening its mandate to cover emerging technologies beyond crypto. Marc Andreessen, Ben Horowitz, and Chris Dixon explain why AI is creating the largest capability gap between early adopters and incumbents they have ever observed, and why founders should now ask "is this ambitious enough?" rather than "is this too ambitious?

→ WHAT IT COVERS Reducto CEO Adit Abraham and a16z's Jennifer Lee examine why document processing remains unsolved infrastructure for AI agents. Reducto's R1 model processes over 7 billion pages monthly at one cent per page, targeting healthcare, legal, finance, and insurance workflows where OCR errors carry real-world consequences. → KEY INSIGHTS - **Benchmark reliability:** Frontier models like GPT-4 outperform specialized document parsers on precision but fail critically on recall, silently...

→ WHAT IT COVERS Karan Singhal, who leads health AI research at OpenAI, explains how ChatGPT serves 300 million weekly health users, why medical expertise is embedded into general-purpose models rather than separate systems, and how AI can act as a proactive quarterback navigating fragmented healthcare for patients and clinicians alike. → KEY INSIGHTS - **General vs.

→ WHAT IT COVERS a16z's Alex Rampell interviews Jean Denis Greyes, former Plaid CTO and Town founder, about building AI assistants for workplace workflows. They cover Town's pivot from SMB tax software, why model neutrality matters for enterprise buyers, how agents will reshape commerce and software distribution, and why AI adoption remains concentrated among technical tinkerers rather than spreading broadly across businesses.

→ WHAT IT COVERS EliseAI's head of applied research Mario Martone, a former theoretical physicist with 15 years studying black holes and string theory, explains why deploying AI in housing and healthcare at scale — handling 5 million phone calls annually — demands fundamentally different standards than academic benchmarks suggest. → KEY INSIGHTS - **Benchmark vs. Production Gap:** A 90% accuracy rate that earns praise in academic papers translates to millions of failures at production scale.

→ WHAT IT COVERS ZocDoc CEO Oliver Kharraz explains how the company built scheduling infrastructure connecting patients to doctors across 175+ EHR systems, why national appointment wait times have doubled to 31 days since ZocDoc launched, and how AI agents and payer partnerships like Blue Shield of California are expanding healthcare access. → KEY INSIGHTS - **Provider Directory Accuracy:** Static provider directories fail because they function like Encyclopedia Britannica — outdated the moment...

→ WHAT IT COVERS a16z General Partner Martin Casado and Grepr CEO Jad Naous examine how AI transforms software engineering, from Naous's 10-person team shipping in days what once took months, to the harder unsolved frontier of AI-driven operations. They debate whether production systems—with unpredictable workloads and near-infinite failure states—can ever be managed the way code is generated.

→ WHAT IT COVERS A16z general partner Alex Rampell and VGS chief strategy officer Terry Angelos examine how AI agents are shifting from passive recommenders to active buyers, the infrastructure enabling this transition, and how new agentic payment tokens from Visa and Mastercard are replacing ad-hoc workarounds to preserve consumer protections during AI-initiated transactions.

→ WHAT IT COVERS a16z partner Lisha Li interviews AI researcher Nathan Lambert, founder of nonprofit lab Trillium, examining whether open-weight models can keep pace with closed frontier labs like OpenAI and Anthropic, analyzing the roles of compute, distillation, post-training, and recursive self-improvement in determining who leads AI development. → KEY INSIGHTS - **Compute as the decisive gap:** Chinese AI labs match US frontier labs in technical talent but are severely compute-constrained —...

→ WHAT IT COVERS Stanford cryptographer Dan Boneh joins a16z's Eddie Lazarin and Justin Thaler to assess whether OpenAI's recent mathematical breakthroughs—spanning roughly 700 problems across computer science—pose a genuine threat to elliptic curve cryptography securing banking systems, private messages, and blockchain wallets, and what institutions should actually do right now. → KEY INSIGHTS - **Cryptographic assumptions vs.

→ WHAT IT COVERS A16Z general partner Martin Casado interviews MakerMaker.ai cofounders Daval Ajita and Owen He on true Recursive Self-Improvement (RSI), distinguishing it from autocatalytic AI development. They cover closed-loop system design, measuring genuine capability gains versus resource costs, and early empirical evidence that self-improving systems can simultaneously improve performance and efficiency.

→ WHAT IT COVERS Nick Bostrom, founder of the Macro Strategy Research Initiative and author of Superintelligence, joins the a16z podcast to examine AI alignment risks, optimal timing for superintelligence development, open source model dangers, AI consciousness, and why delaying advanced AI carries measurable human costs measured in lost life-years. → KEY INSIGHTS - **Optimal Timing Calculus:** Delaying superintelligence development has a concrete human cost: roughly 60 million people die...

→ WHAT IT COVERS Ben Horowitz, co-founder of Andreessen Horowitz, speaks with James Norman on the inaugural Vintage x Vintage episode about evaluating founders on exceptional strengths rather than absent weaknesses, building a16z as an entrepreneur-first firm, and why AI represents a larger shift than the entire internet era. → KEY INSIGHTS - **Founder Evaluation Framework:** When assessing entrepreneurs, ignore the checklist of weaknesses — every successful founder has critical gaps.

→ WHAT IT COVERS Playground Global General Partner Pat Gelsinger joins a16z's Raghu Raghuram and Guido Appenzeller to map the next bottlenecks in AI infrastructure — chip manufacturing timelines, memory bandwidth limitations, power capacity constraints, optical networking transitions, and how virtualization principles apply to agent-based computing systems. → KEY INSIGHTS - **Chip Design vs.

→ WHAT IT COVERS AWS CEO Matt Garman speaks with a16z about the infrastructure demands of an AI-driven world, covering how AWS allocates scarce GPU capacity across frontier labs and startups, the shift to agentic cloud architecture, Trainium chip performance, $220 billion in 2026 capital expenditure, and how agents are reshaping internal software development velocity.

→ WHAT IT COVERS Valon co-founders Linda Du and Andrew Wang explain how they rebuilt mortgage servicing — a $13 trillion market running on pre-internet infrastructure — by becoming a licensed servicer first, encoding 50 states of regulation into software, achieving 3x operational efficiency, and now deploying AI agents across enterprise customers including a 4-million-loan transfer from NewRez.