Latest Interviews
Showing 1–15 of 79 transcripts.
Clear all filters- Y Combinator56 min
How To Design In The Agent Era
Led by Stephen Haney, Paper is an AI-native design platform currently ranking as the third most-used tool behind Figma and Sketch, utilizing a native HTML and CSS rendering engine to facilitate precise agent collaboration. The application distinguishes itself by treating code as the single source of truth, offering features like direct React component handoff and an MCP server that allows autonomous agents to manipulate local repositories and design canvases. While the team prioritizes human curation to maintain design quality against common AI aesthetic pitfalls, the platform is rapidly evolving into a visual coordination layer for the broader agent ecosystem by integrating tools like Paper Shaders and multi-model image generation.
- a16z13 min
The Nuclear Renaissance - Radiant | a16z American Dynamism
Tori Shivanandan, Doug Bernauer
Radiant Energy is developing the transportable Kaleidos micro-reactor, a modular nuclear power unit designed to replace aging infrastructure and support applications ranging from AI data centers to deep space exploration. Founded by former SpaceX engineer Doug Kim, the company leverages a factory-built assembly line model to deliver plug-and-play energy modules with a five-year fuel cycle, aiming to conduct its first full-power test at Idaho National Laboratory's Dome by 2026. This initiative seeks to restore U.S. energy independence and accelerate nuclear innovation by overcoming historical regulatory stagnation through a commitment to aggressive safety testing and mass production.
Inside the Factory Behind Base Power's $1B Series D
Base Power recently secured a $1 billion Series D funding round led by Ribbit and JPMorgan Chase, establishing a $13 billion valuation to accelerate production of its high-capacity BaseCore home battery system. The vertically integrated manufacturer is scaling operations at its Austin facility to 4,000 units weekly, featuring a modular design that enables full-home backup power in under two hours per unit. This growth supports a strategic rollout of both direct consumer sales and utility partnership models across key regional markets as the company prepares to open a second factory near the Austin airport.
- All-In Podcast50 min
Former Intel CEO on What Went Wrong, What's Next + Lovable CEO on the Real Promise of Vibe Coding
Pat Gelsinger, Jason, Anton Osika
Intel CEO Pat Gelsinger diagnosed the company's strategic decline as a fifteen-year period of non-technical leadership that diverted $100 billion to shareholders rather than manufacturing infrastructure, allowing competitors like NVIDIA and TSMC to seize market dominance. He further warned of acute geopolitical risks to the semiconductor supply chain in Taiwan while projecting that AI growth will be constrained by energy capacity rather than a speculative bubble. Concurrently, Lovable founder Anton Osicka outlined a business model achieving $600 million in revenue by leveraging AI to enable non-technical users to build production-ready software, fundamentally reducing development costs and accelerating enterprise tool creation.
- Y Combinator31 min
YC's Head of Design Shows You How To Design With AI
Eve Bouffard, Aaron Epstein, Ev Bufar
Designer Ev Bufar utilizes a voice-first workflow with tools like Aqua, Conductor, and Paper.design to rapidly produce high-fidelity assets for diverse projects, including the Paxyl transcript analyzer and the Soda Zine. Her recent work at Startup School demonstrates an agent-driven production model that generated visual identities for 6,000 attendees through custom shaders and automated speaker assets, while Paxyl analyzes coding agent interactions to reveal developer patterns. By prioritizing massive context density and disposable internal models, Bufar accelerates the design iteration cycle to handle complex branding and interactive web experiences that explicitly cater to both human users and AI agents.
- Y Combinator32 min
Solving the Blank Canvas Problem: Gusto's AI Co-Founder
Launched in just ten weeks by a five-person team including co-founder Eddie Kim, Gusto Co-founder is a new AI product that transforms Gusto's payroll and HR data into proactive, agent-based automations for small businesses. The platform has already onboarded 500 paying customers by replacing traditional dashboards with text-based interfaces that execute complex workflows, such as full payroll processing, weather-based customer notifications, and tax credit application. This initiative validates a strategic pivot toward becoming a proactive business partner capable of lowering entry barriers for new entrepreneurs while automating compliance and growth opportunities through a "vibe coding" development methodology.
Inside Harvey AI: CEO Winston Weinberg on How a 4-Year-Old Company Is Taking on the Foundation Labs
Winston Weinberg, Molly O'Shea
Harvey AI has scaled to nearly $300 million in ARR and over 2,000 enterprise clients by migrating to cloud agents and deploying specialized legal models that outperform general frontier AI. CEO Sam Saltman drives a culture of rapid reinvention and meritocracy, where over 200 legal professionals transition into product and engineering roles to build vertical-specific solutions for high-growth sectors like finance and pharmaceuticals. The company differentiates itself from traditional legal tech and generic AI labs by prioritizing granular ROI proof per token and rejecting legacy technology acquisitions in favor of acquiring high-performing teams to fuel its current 12-to-13 trillion token monthly usage projections.
- Y Combinator17 min
Conductor CEO Charlie Holtz Walks Us Through His AI Coding Setup
Charlie, co-founder of Conductor, leads a team developing a macOS application that orchestrates multiple coding agents to generate software primarily in TypeScript and Rust. The platform enforces a strict non-terminal workflow where AI handles code generation while humans intervene only through specific review modes to maintain quality and prevent error cycles. With a strategic focus on high-efficiency model selection and a future roadmap emphasizing malleable, modifiable agents, Conductor positions itself as a tool that shifts software value from code itself to human-authored prompts and descriptions.
Tom Mueller: SpaceX's #1 Employee Built the Merlin Engine. Now He's Building Impulse Space
Former SpaceX founding employee Tom Mueller founded Impulse Space to revolutionize intra-orbital transport with its fleet of advanced spacecraft, including the orbital transfer vehicle Mira and the heavy-lift Helios. The company leverages proprietary Nova Light alloy and additive manufacturing to produce the Deneb hydrocarbon engine and lightweight composite tanks, achieving performance metrics that could significantly boost payload delivery to the Moon and Mars. Currently advancing lunar infrastructure efforts through successful rendezvous tests and an active lander contract bid, Impulse Space combines in-house manufacturing with AI-driven design to support a future lunar economy.
Inside PsiQuantum: The Photonics Chip Never Seen Before
Leveraging Molecular Beam Epitaxy within a Class 1,000 cleanroom, engineers fabricate high-precision Barium Titanate on Strontium Titanate films atop 300mm silicon wafers to support single-photon quantum computing architectures. The ultra-high vacuum process employs rigorous environmental controls and automated transfer systems to ensure crystalline integrity while facilitating immediate integration into high-volume manufacturing lines with partners like GlobalFoundries. Ongoing characterization using RHEED and optical interference validates these substrates for future dual utility as qubits or storage vectors within scalable quantum systems.
- 80,000 Hours36 min
Will AI cause mass unemployment? Maybe not.
Amid widespread fear of AI-driven job displacement, the event analyzes historical automation trends to reveal that wages may rise significantly until 2037 before facing a decline if full task automation is achieved without human oversight. Experts identify four categories of skills likely to increase in value, including complex physical tasks, AI deployment capabilities, and roles producing scalable goods, while warning against exclusive specialization in routine knowledge work. The presentation concludes by advising professionals to cultivate transferable leadership and technical abilities that bridge human decision-making with AI efficiency to navigate a shifting labor market.
- Jane Street16 min
Dwarkesh Goes Inside Jane Street's Latest AI Data Center
Dwarkesh, Ron Minsky, Daniel Pontecorvo, Mark Mirchandani
Jane Street transformed a Texas data center into a high-density liquid-cooled facility housing 4,032 GPUs to execute large language model training and custom trading architectures. The retrofit replaces legacy air-cooling with an 18°C fluid distribution system that manages 140 kW per cabinet while utilizing proprietary software to dynamically redistribute power and prevent breaker trips during peak loads. Engineering safeguards now focus on mitigating new liquid-cooling risks like biological growth and leaks, enabling sub-100-nanosecond latency required for modern algorithmic trading compared to the millisecond scales of its historical "Hive" cluster.
Figure's First Full HQ Tour: From the Lab to the Factory Floor
Brett Adcock, Molly O'Shea, Moritz
Figure AI is a Bay Area-based pure-play manufacturer with a 500-person engineering workforce deploying its flagship Figure 3 humanoid robot, which operates autonomously using the locally running Helix AI policy. The company has achieved record production output at its "Baku" facility while reducing unit costs by 90% compared to previous models, enabling deployments in settings ranging from the White House to BMW automotive assembly lines. Looking ahead, Figure aims to commercialize a lease-based home model and advance toward its "iPhone moment" with the upcoming Figure 4, which targets radical improvements in cost and usability while pursuing physical interaction-based AGI.
Skydio HQ Tour: The $3.5B Bet on American Drone Manufacturing with CEO Adam Bry
An autonomous drone network delivers 24/7 public safety operations by deploying units from charging docks to locate missing persons, track suspects, and inspect infrastructure without human pilots. The system utilizes Beyond Visual Line of Sight waivers and Pathfinder AI to execute complex missions across diverse hardware, ranging from the indoor R10 tactical unit to the high-speed F10 fixed-wing prototype. Through a transparency portal and continuous real-world stress testing, the company has integrated these force-multiplier tools into law enforcement culture while scaling production for imminent general availability.
- The Economist9 min
AI has got better at hacking—how big a risk is it?
Anthropic withheld its latest model, Mythos, from public release after it demonstrated the ability to autonomously identify and exploit critical software vulnerabilities, such as a 27-year-old OpenBSD flaw, with minimal human intervention. This decision follows evidence that AI scaling laws now enable automated systems to accelerate the vulnerability exploitation cycle from years to mere hours, drastically outpacing current defensive capabilities. While Anthropic employed responsible disclosure by patching the specific bug before sharing details, the incident highlights a growing risk that open-source models will soon democratize high-level cyberattacks.