April 3, 2026
Marc Andreessen frames AI as an "80-year overnight success," detailing the PI/OpenClaw agent architecture and why "this time is different."
April 2, 2026
Moonlake champions causal world models built on symbolic reasoning for true interactivity and efficiency, challenging the scale-is-all approach.
March 30, 2026
Mistral's Guillaume Lample & Pavan Kumar Reddy launch Voxtral TTS, a 3B model with novel flow matching. Plus, their open source vision & Forge.
March 24, 2026
MIT's Heather Kulik explains why there's no "AlphaFold for Materials," citing data scarcity, chemical complexity, and the limits of current AI.
March 20, 2026
David Singleton introduces Dreamer, a personal agent OS empowering everyone to discover, build, and use AI agents, no tech background needed.
March 17, 2026
Felix Rieseberg on Claude Cowork: A sandboxed AI in a local VM, designed to be a powerful, safe, and user-friendly personal coworker.
March 12, 2026
Turbopuffer's Simon Eskildsen on building an S3-native search DB for AI. He discusses hybrid search, agent workloads, and hiring P99 engineers.
March 10, 2026
NVIDIA's AI engineers on scaling agent inference with Dynamo, the Brev acquisition story, and the 'Speed of Light' culture driving innovation.
March 6, 2026
Cursor's Third Era introduces Cloud Agents: AI developers in a VM that test their code and create video demos, moving coding to a higher level.
March 5, 2026
Box CEO Aaron Levie explains why every AI agent needs a "box"—a secure, governed workspace to tackle the enterprise challenges of agent identity.
February 27, 2026
METR's Joel Becker discusses their "Time Horizon" eval for AI capabilities, threat models, and the continuous but astonishing pace of AI progress.
February 26, 2026
Anthropic's distillation 'attacks' highlight AI's geopolitical tensions, while the death of SWE-Bench reveals the crisis in AI evaluation.
February 25, 2026
Prof. Max Welling of CuspAI discusses using AI to program 'nature as a computer' to accelerate materials science and solve climate change.
February 24, 2026
SemiAnalysis's Doug O'Laughlin on his "Claude Code psychosis," the coming global memory shortage, and why AI agents are changing finance.
February 23, 2026
OpenAI's Frontier Evals team declares SWE-Bench Verified dead due to saturation and contamination, advocating for a move to SWE-Bench Pro.
February 19, 2026
a16z's Martin Casado & Sarah Wang dissect AI's $10B+ capital flywheel, where dollars translate directly to capabilities, blurring old VC rules.
February 12, 2026
Google's Jeff Dean on owning the AI Pareto Frontier through a full-stack strategy of co-designing hardware, software, and models.
Boltz founders discuss their journey from AlphaFold's breakthrough to open-sourcing AI for drug discovery, democratizing science and innovation.
February 6, 2026
Goodfire AI pioneers production-grade mechanistic interpretability, using techniques like model steering to design, control, and debug AI safely.
January 28, 2026
Andrew White discusses automating science with AI agents like Cosmos, the role of world models, and his journey from academia to founding Edison.
January 23, 2026
Yi Tay on captaining IMO Gold with Gemini, the power of on-policy RL, and why Transformers are here to stay. Insights from the AI frontier.
January 17, 2026
Brex CTO James Reggio on their three-pillar AI strategy to 10x workflows, cut operational costs, and build agentic finance products.
January 9, 2026
Artificial Analysis founders discuss their independent LLM benchmarking, new evals for hallucination & agentic tasks, and AI cost trends.
January 8, 2026
Artificial Analysis founders share their journey from a side project to the 'Gartner of AI,' detailing their independent benchmarking and new evals.
Artificial Analysis founders discuss their independent LLM benchmarks, business model, and new evals for hallucination, agents, and openness.
December 31, 2025
John Yang of SWE-bench discusses the evolution of coding evals, his new project Code Clash, and the shift towards long-horizon, interactive AI.
Anastasios of Arena discusses scaling from a research project to a $100M company, emphasizing platform integrity and real-user data for AI evaluation.
An OpenAI researcher discusses the shift to post-training, the complexity of RL, token efficiency, and the need for hybrid systems/ML talent.
December 30, 2025
Ex-OpenAI researcher Ashvin Nair on RL's evolution, why LLM agents will outpace robotics, and Cursor's strategy of co-designing product & model.
Amplify's Sarah Catanzaro on the AI funding frenzy, the data stack's evolution for AI, and why memory & personalization are the next big bets.
December 27, 2025
Experts from Anthropic, OpenAI & the Linux Foundation discuss MCP's evolution, its adoption, and the new Agentic AI Foundation (AAIF).
December 26, 2025
Steve Yegge declares the end of IDEs and manual coding, urging a shift to 'Vibe Coding' with AI agents to avoid becoming obsolete.
OpenAI's team discusses training coding agents like Codex for personality, the shift to higher-level agent abstractions, and the future of evals.
December 18, 2025
Meta researchers and Roboflow's CEO discuss SAM 3, its novel data engine, unified architecture, and its role as a visual tool for LLMs.
December 16, 2025
Hacker Pliny the Elder discusses universal AI jailbreaks, the futility of model guardrails, and the need for open-source security research.
December 12, 2025
KP's Jubin shares his journey from VC to founder, detailing his podcasting strategy, hiring philosophy, and new AI startup, Roadrunner.
December 11, 2025
Superhuman's CTO discusses building a high-quality AI email experience, the shift to conversational UX, and scaling complex products globally.
December 6, 2025
General Intuition CEO Pim discusses building world models from Metal's 3.8B gaming clips, creating a unique data moat for spatial intelligence.
November 25, 2025
World Labs founders Fei-Fei Li & Justin Johnson introduce "spatial intelligence" as the next AI frontier, moving beyond LLMs with their 3D world model, Marble.
World Labs founders Fei Fei & Justin explore spatial intelligence, the evolution of AI research, and their vision for world models with Marble.
November 19, 2025
10X co-founders reveal their output-based model allowing AI engineers to earn $1M+, using rapid prototyping as a key sales tool.
November 14, 2025
Deedy Das on Glean's $7B valuation, Anthropic's meteoric rise, and the venture strategy behind AI's fastest-growing startups.
November 10, 2025
GitHub's Jared Palmer discusses his journey from building V0 at Vercel to launching AgentHQ, defining the future of coding agents on GitHub.
Google Labs' Jed Borovik discusses building the autonomous coding agent Jules, the shift from complex scaffolds to powerful models, and the future of AI engineering.
November 6, 2025
Mark Zuckerberg and Priscilla Chan outline CZI's strategy: merging frontier AI and biology to build foundational tools and accelerate the mission to cure all diseases.
November 3, 2025
Zyphra's head of model training discusses their all-in bet on AMD, the craft of kernel writing, and disciplined AI-assisted coding.
October 31, 2025
Vercel CTO Malte Ubl discusses building for the AI era with new workflow tools, a humble AISDK, practical agents, and secure development.
October 30, 2025
Stripe's Head of Data & AI on building economic infrastructure for AI, combating new fraud vectors, and the Agentic Commerce Protocol.
October 16, 2025
OpenPipe founder Kyle discusses his journey from YC to a CoreWeave acquisition, covering the evolution of fine-tuning, RL, LORAs, and GRPO.
October 7, 2025
OpenAI's Platform team unveils Agent Kit, a full-stack solution with a visual builder, UI components, and evals to build and deploy agents.
October 2, 2025
Figma CEO Dylan Field on AI's role in blurring the lines between idea and reality, and why human design and craft are now the key differentiators.
September 25, 2025
Sourcegraph's leaders discuss their radical strategy for AMP: prioritizing speed and innovation to build the ultimate coding agent for power users.
September 11, 2025
LangChain's Lance Martin unpacks Context Engineering for agents, covering retrieval, multi-agent systems, and the Bitter Lesson in AI development.
September 5, 2025
Gorkum & Batuhan of Fal discuss their pivot to generative media, optimizing inference with custom kernels, and their rapid growth to $100M+ ARR.
![[LIVE] Anthropic Distillation & How Models Cheat (SWE-Bench Dead) | Nathan Lambert & Sebastian Raschka thumbnail](https://substackcdn.com/feed/podcast/1084089/post/189277598/ca7468da5614a246d2906ee8926f6de7.jpg)
