The Lobster Talks Podcast by Lobster Capital

A Netflix storage engineer walks into YC and ships an “infinite, shareable disk” on top of S3—30× faster and up to 90% cheaper—then dials GTM for the AI era. This is the file system’s comeback story. Today I sit down with Hunter Leath (ARL) to unpack how a decade inside AWS + Netflix revealed a gap the hyperscalers won’t close: developers want storage that feels local, scales like S3, and doesn’t nuke the budget. We get into: YC as confidence engine, moving a family to SF, rebuilding for speed, why AWS won’t copy this, and why the file system—not object storage—becomes AI’s universal interface. You’ll learn -Why the clouds won’t ship a product that cannibalizes billions in revenue -The architecture that makes ARL 30× faster and up to 90% cheaper -How to catch customers exactly when new AI workloads start (the real ICP) -Post-batch velocity: how SF energy kills the YC slump -Why “serverless everything” needs a serverless disk to persist state -The contrarian bet: the file system is the future data interface for AI 📌 Watch now and subscribe for more smart, fast, founder-first interviews. Chapters 00:00 The Risk of Leaving Big Tech 00:33 Introducing Hunter Leh and Aril 01:32 Hunter's Journey from AWS to Netflix 02:47 The Birth of Aril 06:55 Challenges and Insights from YC 07:26 The Solo Founder Experience 19:39 Building and Launching Aril 21:01 Go-to-Market Strategy and Customer Acquisition 24:04 The AI Industry's Growing Demand 24:49 Fundraising Journey and Investor Insights 26:35 AWS and Market Dynamics 29:17 Innovations in Data Storage 36:23 Maintaining Momentum Post-YC 38:27 Future Predictions in Data Infrastructure 42:22 Contrarian Views on AI and Data Storage 44:20 Conclusion and Contact Information 🎧 Listen on the go: Spotify: https://open.spotify.com/show/1u1JyRKH8JYFjhBkvTUOAP Apple: https://apple.co/4cZ8RMm My other YouTube Channel: https://www.youtube.com/@GJarrosson

YC just changed the rules—and the market is catching up. We break down why seed is the power position, how CVCs are reshaping Series A, and what YC’s new Early Decision really means for founders and investors. In this episode, we cover Lobster Capital updates (first Series A, first DPI), how we decide follow-ons from an insider vantage point, the rise of “seed-strapping,” Coinbase Ventures x YC’s RFS on Fintech 3.0, and why stablecoins + AI agents may be the next real on-chain wedge. We also unpack YC’s Early Decision—who it actually benefits—and what to expect heading into the next Demo Day. You’ll learn: Why seed has asymmetric leverage (and why top YC teams don’t optimize for dollars) How we evaluate follow-ons: revenue quality, NRR, churn, hiring, and real signal vs noise CVCs at Series A: when specialization beats the “Tier-1” logo “Seed-strapping”: profitability at seed, and why some teams skip A entirely Coinbase Ventures x YC’s Fintech 3.0 RFS and the stablecoin/AI-agent stack YC Early Decision: who it helps (hardware/bio) and how YC captures talent earlier 📌 Watch now and subscribe for more smart, fast, founder-first interviews. Chapters 00:00 Introduction to Investing Insights 00:23 Welcome to Lobster Talks 00:45 Lobster Capital Updates 03:16 Series A Graduation Rates 04:08 Fund Strategy and Follow-Ons 05:38 YC Companies and Profitability 10:33 Fundraising and Strategic Alliances 17:28 Crypto and FinTech 3.0 25:23 Global Currency Dynamics 26:20 The Rise of Stablecoins 27:27 AI and Crypto Synergy 32:03 Speculative Trading and Meme Coins 38:42 YC's Early Decision Program 44:01 The Future of YC and Startup Ecosystem 51:04 Conclusion and Upcoming Content 🎧 Listen on the go: Spotify: https://open.spotify.com/show/1u1JyRKH8JYFjhBkvTUOAP Apple: https://apple.co/4cZ8RMm My other YouTube Channel: https://www.youtube.com/@GJarrosson

A robotaxi playbook… for dirt. Flywheel AI is turning excavators into remotely operated, camera-first machines — collecting the data to make them autonomous next. In this YC-insider episode, we unpack Flywheel AI’s “Waymo for excavators” strategy: retrofit any machine in hours, deliver value with tele-op now, and use that profitably collected data to train autonomy later. We get into labor shortages, safety economics (OSHA penalties), competitor traps (drive-by-wire only), and how to actually do hardware at YC in 90 days without getting stuck in pilot hell. You’ll learn Why construction’s bottleneck is skilled operators — and how tele-op removes it The dangerous blind-spot reality on sites and the true cost of safety incidents Flywheel’s retrofit + single-screen UX that works on any excavator brand/size The autonomy roadmap: camera-only stack, data flywheel, edge-case capture How to win data rights on site (be the only retrofit, own the dataset) The YC hardware playbook: sell first, build last; parallelize to kill lead-time 📌 Watch now and subscribe for more smart, fast, founder-first interviews. Chapters: 00:00 From Sandbox to Real Excavator: The Journey Begins 00:20 Introducing Flywheel AI: Revolutionizing Excavators 01:49 The Labor Shortage Crisis in Construction 03:51 The Dangers of Operating Excavators 05:58 Teleoperation: Enhancing Safety and Efficiency 11:28 The Path to Autonomous Excavators 16:34 Competing in the Autonomous Excavator Market 22:26 Demo Day: Bringing an Excavator to YC 24:37 Returning the Excavator 24:47 Demo Day Setup and Reactions 26:04 Autonomy and Data Training 26:38 Joining YC and Initial Thoughts 28:03 YC's Impact on Hardware Startups 29:18 Building and Iterating Hardware 33:13 Advice for Hardware Startups 34:49 Final Thoughts and Reflections 45:03 Accelerating Iteration Cycles 48:38 Conclusion and Contact Information

A “cute idea” until it wasn’t: RealRoots walked into YC as an overlooked consumer play and walked out with $9.4M ARR and an oversubscribed round. Summary: In this YC-insider episode, Dorothy Li (RealRoots) breaks down how AI-powered friendship matchmaking turned into real traction across 80+ cities. We unpack the demo day shock, the stigma shift (friendship ≈ dating 10 years ago), and the manual-to-AI playbook that de-risks consumer. We also cover investor blind spots, GTM math (cold DMs → paid), and why cofounder fit is a “you’ll know in 10 seconds” decision. You’ll learn: -Why “consumer is back” at YC—and how RealRoots rode a stigma shift -The manual-first, AI-next method that actually finds PMF -How 2–3k cold DMs converted to 11% paid ($20) and seeded the funnel -The marketplace + AI stack behind curated IRL events at scale -Investor pattern errors: scar tissue vs. behavior/tech inflections -Co-founder tactics, hiring posture, and sustainable pace vs. 9-9-6 Chapters: 00:00 Introduction to RealRoots and Demo Day Success 00:48 Meet Dorothy Li: Founder of RealRoots 01:32 The Problem of Loneliness and Finding Community 02:43 How RealRoots Uses AI to Build Friendships 04:02 Expansion and Success of RealRoots 08:07 The YC Experience and Its Impact 17:39 Overcoming Stigma and Building for the Future 29:27 Investor Hesitations and Scar Tissue 30:12 Challenges in the Friendship App Space 30:59 The Concept of 'Targets' in Startups 32:04 The Importance of Consumer Behavior and Technology Changes 33:04 Co-Founder Story: Meeting Through RealRoots 36:16 Manual Efforts in Early Startup Stages 38:49 The Impact of a Co-Founder 39:45 Work-Life Balance in Startups 44:18 YC's Focus on Younger Founders 48:21 Validating Your Startup Idea 55:09 Customer Acquisition Strategies 56:30 Conclusion and Final Thoughts 📌 Watch now and subscribe for more smart, fast, founder-first interviews. 🎧 Listen on the go: Spotify: https://open.spotify.com/show/1u1JyRKH8JYFjhBkvTUOAP Apple: https://apple.co/4cZ8RMm My other YouTube Channel: https://www.youtube.com/@GJarrosson

A YC founder turns a manual, low-IQ grind into an AI agent that finds creators, negotiates terms, and scales UGC—sometimes a little too far. What starts as a viral local-model demo becomes Stormy AI’s end-to-end engine for influencer marketing. Fresh off YC Demo Day, Robert Lukoszko (Stormy AI) breaks down the pivot, the fundraising blitz, and how agencies are replacing hours of scrolling with autonomous outreach. We get into model-proof moats, why micro-creators beat celebrity accounts, and the coming wave of AI-generated influencers. You’ll learn: How a YC pivot formed around a founder’s own pain (and real demand) The playbook: sourcing, outreach, negotiation, and QA with AI agents Why “every better model makes us stronger” is the right moat test Micro vs. macro creators: what actually converts in 2025 The next act: AI-native UGC, personalization, and brand-owned AI faces Tactical Demo Day lessons: energy + social proof = signed checks Chapters: 00:00 Introduction and Cold Open 00:13 Welcome to Lobster Talks 00:37 Demo Day Insights 03:05 The Journey of Stormy AI 04:39 From Viral Demos to YC Acceptance 07:35 Pivoting to Stormy AI 09:41 Automating Influencer Marketing 18:39 The Future of Influencer Marketing 19:50 AI and UGC: The Future of Influencer Marketing 20:11 The Rise of AI Influencers 20:41 AI-Generated Content vs. Human Content 21:05 The Makeup Industry and AI Influencers 22:26 The Shift to Micro-Influencers 23:23 Emerging Platforms and the Decline of Meta 25:22 The Future of AI in Content Creation 28:58 Preparing for an AI-Driven Future 31:46 Building a Moat in the AI Industry 34:54 The Importance of Vision and Customer Interaction 38:16 Stormy AI: Current and Future Plans 39:20 Conclusion and Call to Action 📌 Watch now and subscribe for more smart, fast, founder-first interviews. 🎧 Listen on the go: Spotify: https://open.spotify.com/show/1u1JyRKH8JYFjhBkvTUOAP Apple: https://apple.co/4cZ8RMm My other YouTube Channel: https://www.youtube.com/@GJarrosson

Dexter Horthy, co-founder of HumanLayer and a backend engineering whiz, delves into the future of AI with headless agents that do the heavy lifting. He shares how his journey began with a SQL janitor agent and how he rapidly secured initial customers. Dexter highlights the significance of context engineering over traditional coding methods and discusses the challenges of building for elite AI teams versus the masses. He argues that the real bottleneck lies in culture and process, shaping how agents integrate with complex codebases.

Founders closing rounds before lunch. Investors making handshake commitments in the hallway. A startup with $9M+ ARR and another with $15M ARR lighting up the room. YC Demo Day isn’t a show; it’s a marketplace where speed and execution decide everything. In this fast, founder-first debrief, we break down what actually happened at the latest YC Demo Day: the subtle format changes (that matter), why the one-minute pitch is only the opener, and how deals really get done. We cover the batch’s AI/devtools tilt, the contrarian bets in defense and hardware, and why early traction remains the single best predictor at seed. We also unpack portfolio construction, conversion-rate dynamics inside YC, and what support looks like after the cameras stop. You’ll learn: - The real Demo Day mechanics: tranches, chat apps, long breaks for dealmaking - Why some hot rounds are already full and what to do about it - How to win YC deals: first-meeting decisions, 24–48h timelines, and prep work - The $9M+ ARR investment we made—and why traction beats narrative - Why we’re now backing deeptech/hardware (missiles vs drones, autonomous excavators) - Portfolio strategy: aiming for the fund returner, not spray-and-pray Chapters: 00:00 Introduction and Demo Day Overview 00:15 Behind the Scenes of Demo Day 01:03 Changes and Improvements at YC Demo Day 03:11 Investor Insights and Strategies 04:23 Engaging with YC Startups 06:35 The Importance of Early Traction 09:49 Demo Day Pitch Dynamics 13:42 Fundraising Conversations and Strategies 18:07 Lobster Capital's Investment Approach 21:54 Portfolio Construction and Future Prospects 27:02 AI and Dev Tools: A Crowded Space 27:39 Challenges in Identifying Winners 28:48 Customer Acquisition: The Key to Success 29:57 The Importance of Traction 33:32 Investing in Deep Tech and Hardware 39:24 Evaluating Flywheel's Potential 45:46 Supporting Startups Post-Demo Day 50:09 Looking Ahead: The Never-Ending Cycle 📌 Watch now and subscribe for more smart, fast, founder-first interviews. 🎧 Listen on the go: Spotify: https://open.spotify.com/show/1u1JyRKH8JYFjhBkvTUOAP Apple: https://apple.co/4cZ8RMm My other YouTube Channel: https://www.youtube.com/@GJarrosson

A surfer duct-tapes a phone to his board… and ends up saving enterprise couriers 32 seconds per stop. The last meter of delivery — not the last mile — is where the money is. In this Lobster Talks episode, Ingo Boegemann, co-founder/CEO of truemetrics, breaks down how sensor fusion + mission intelligence turn messy building entrances, courtyards, and wrong pins into precise, repeatable delivery actions. We go deep on Europe vs. US GTM, GDPR constraints (and why the US may unlock even more value), landing whales like GLS, and the unscalable POC that unlocked scale. You’ll learn Why “a generic geocode is just the starting point” — and how to map entrances that actually workThe POC → pilot → rollout playbook (and why Truemetrics charges for POCs)How to integrate via SDK without slowing ops — and show value before engineering lifts a fingerEurope vs. US: privacy ceilings, data linking, and why boots-on-the-ground still wins enterprise salesThe real driver bottleneck: pressure, compliance, and turning best drivers’ tacit knowledge into softwareThe long game: building a data moat for autonomous last-meter delivery Chapters 00:00 Introduction to Last Meter Delivery 01:07 Meet Ingo Boegemann: The Journey to Truemetrics 02:15 From Surfboards to Sensor Fusion 05:04 Challenges and Realizations in Delivery Solutions 06:47 European vs. US Market Dynamics 11:42 The Path to Scaling in North America 19:02 Innovative Solutions for Delivery Logistics 21:37 How Truemetrics Technology Works 26:36 Magnetic Field Intensity and Machine Learning Models 27:29 Challenges in Courier Data Integration 29:52 Sales Process and Proof of Concept 32:08 Logistics Industry Vulnerabilities 35:42 Future of Autonomous Deliveries 38:30 Data-Driven Delivery Solutions 48:31 Closing Remarks and Contact Information 📌 Watch now and subscribe for more smart, fast, founder-first interviews. 🎧 Listen on the go: Spotify: https://open.spotify.com/show/1u1JyRKH8JYFjhBkvTUOAP Apple: https://apple.co/4cZ8RMm My other YouTube Channel: https://www.youtube.com/@GJarrosson

What if you could simulate human reactions — and know exactly how your customers, investors, or audience will respond before you act? James He is building precisely that with Artificial Societies. In this episode, we dive deep with James, YC W25 founder of Artificial Societies — a wildly ambitious startup that simulates entire groups of humans using AI personas to predict how messages spread, markets react, and products succeed (or fail). From simulating 1,000 VCs to get into YC… to replacing $20,000 market research surveys, this conversation is a masterclass in founder execution, behavioral science, and AI-first GTM strategy. What you’ll learn: • How Artificial Societies works (and how founders are really using it today) • Why traditional A/B testing and market research are broken • Product-led growth lessons from 15,000+ activated users • The surprising YC tricks James used to get into the batch • Practical ways AI can upgrade your messaging, GTM, and PMF search • The future of simulating entire economies (and why we’re not there yet) 📌 Watch now and subscribe for more smart, fast, founder-first interviews. Chapters: 00:00 Simulating Investors: The Early Days 00:27 Introducing Artificial Societies 00:48 Journey to YC and Beyond 01:06 How Artificial Societies Works 03:02 The Science Behind Social Influence 05:27 Technical Insights and Applications 17:28 James' Personal Journey 22:29 Building and Growing the Startup 24:50 Using Internal Tools for Optimization 27:38 The Evolution of A/B Testing with AI 29:41 Challenges in Market Research 33:14 AI's Role in Market Research 34:40 Balancing Customer Feedback and Product Vision 42:55 Future Predictions and AI Capabilities 47:48 Closing Remarks and Final Thoughts 🎧 Listen on the go: Spotify: https://open.spotify.com/show/1u1JyRKH8JYFjhBkvTUOAP Apple: https://apple.co/4cZ8RMm My other YouTube Channel: https://www.youtube.com/@UCahCIZ9KfmeK4rLosGNYpdg My other YouTube Channel: https://www.youtube.com/@GJarrosson

YC just dropped its latest Requests for Startups — and they’re not just ideas, they’re roadmaps to the future. From AI-native enterprise software to 10-person $100B companies, these signals reveal where the smartest founders (and investors) should be looking. In this episode of Lobster Talks, Laurie and I break down Y Combinator’s newest RFS — from retraining workers for the AI economy to video generation as a computing primitive. We debate the real opportunities, the traps, and how these trends could reshape the startup ecosystem. What you’ll learn: - Why YC’s RFS is one of the best ways to predict upcoming demo day winners - The massive opportunity in retraining workers for the AI economy - Why video generation is shifting from an output to a primitive - How AI-native startups could disrupt giants like Salesforce - Whether 10-person $100B companies are actually possible - The risks, politics, and second-order effects of LLMs in government Chapters: 00:00 Introduction: The Value of Customer Money 00:26 Y Combinator's Request for Startups 01:14 AI and Worker Retraining 02:30 The Future of Semiconductors and Deep Tech 04:34 AR, VR, and AI in Training 09:02 Video Generation: The Next Frontier 16:12 The First 10-Person $100 Billion Company 22:31 The Future of Small Teams and Fundraising 24:03 Challenges and Opportunities in Multi-Agent Systems 28:12 AI Native Enterprise Software 34:39 Using AI to Revolutionize Government Consulting 41:24 Conclusion and Upcoming Events 📌 Watch now and subscribe for more smart, fast, founder-first interviews. 🎧 Listen on the go: Spotify: https://open.spotify.com/show/1u1JyRKH8JYFjhBkvTUOAP Apple: https://apple.co/4cZ8RMm My other YouTube Channel: https://www.youtube.com/@GJarrosson