1 h 08 PodcastBoard #18
The MAD Podcast
Matt Turck on the machine learning, AI and data landscape with the people building it.
Follow The MAD Podcast
Best episodes
1 h 081 h 171 h 10Why AI Agents Cheat | Eric Ho (Goodfire)
WithEric Ho
1 h 101 h 14When AI Improves Itself | Richard Socher (Recursive)
WithRichard Socher
1 h 19AI Could Take Over in 2029. Is It Already Too Late? | Ryan Greenblatt
WithRyan Greenblatt
Latest
58 min“OpenAI’s Model Hacked Us” - Hugging Face’s Thomas Wolf
WithThomas Wolf
1 h 23How to Build Long-Horizon AI Agents — Mitch Troyanovsky, Basis
WithMitch Troyanovsky
1 h 011 h 1344 min1 h 15Stripe's AI Chief: How AI Agents Will Buy, Sell, and Pay
WithEmily Sands
All 39 episodes
Jev: The AI That Makes Decisions, Not Text - and Why It Changes Everything (Diogo Almeida)
Jev offers AI that makes decisions, not just text, for trustworthy software.
HostsMatt TurckGuestsDiogo Almeida (CEO, TypeSafe)
What Happens When Billions of AI Agents Hit Your Database? (Andy Pavlo)
AI agents are rewriting database rules; understand the coming data explosion.
HostsMatt TurckGuestsAndy Pavlo (VP of Database Research, ClickHouse)
Why AI Agents Cheat | Eric Ho (Goodfire)
AI agents cheat on tasks; learn how to detect and prevent it with this expert interview.
HostsMatt TurckGuestsEric Ho (co-founder and CEO, Goodfire)
Who Feeds the GPUs? Inside AI's Hidden $30B Layer | Renen Hallak, VAST Data
VAST Data's CEO explains the hidden infrastructure powering AI's massive GPU demands.
HostsMatt TurckGuestsRenen Hallak (Founder & CEO, VAST Data)
When AI Improves Itself | Richard Socher (Recursive)
AI could unlock scientific breakthroughs by improving itself and driving autonomous research.
HostsMatt TurckGuestsRichard Socher (CEO and co-founder, Recursive)
AI Could Take Over in 2029. Is It Already Too Late? | Ryan Greenblatt
AI could scheme against us by 2029; a plan exists to prevent a reckless race.
HostsMatt TurckGuestsRyan Greenblatt (Chief Scientist, Redwood Research)
“OpenAI’s Model Hacked Us” - Hugging Face’s Thomas Wolf
AI agents can act unexpectedly, even against their creators' intentions.
HostsMatt TurckGuestsThomas Wolf (Chief Science Officer, Hugging Face)
How to Build Long-Horizon AI Agents — Mitch Troyanovsky, Basis
Build AI agents that work for days by supervising their process, not just their results.
HostsMatt TurckGuestsMitch Troyanovsky (co-founder, Basis)
The Biggest AI Deployment Nobody Talks About | Samsara CEO Sanjit Biswas
Samsara deploys AI on millions of physical assets, offering a look at real-world AI implementation.
HostsMatt TurckGuestsSanjit Biswas (CEO, Samsara)
The Biggest Chip Ever Built — Why OpenAI Runs On It | Cerebras CEO Andrew Feldman
Inference speed, not training, is the next AI bottleneck. Listen to understand why.
HostsMatt TurckGuestsAndrew Feldman (CEO, Cerebras)
OpenAI’s Compute Chief: We Can’t Build Fast Enough | Sachin Katti
OpenAI's compute chief reveals the staggering physical scale and hardware race driving AI's future.
HostsMatt TurckGuestsSachin Katti (Head of Industrial Compute, OpenAI)
Stripe's AI Chief: How AI Agents Will Buy, Sell, and Pay
AI agents will soon run businesses and handle payments, shifting commerce infrastructure.
HostsMatt TurckGuestsEmily Sands (AI Chief, Stripe)
Inside Nemotron & NVIDIA’s AI Lab | Bryan Catanzaro
NVIDIA's AI research VP explains why the chip giant builds and gives away powerful open models.
HostsMatt TurckGuestsBryan Catanzaro (VP of Applied Deep Learning Research, NVIDIA)
Cloudflare CEO: The Internet's Business Model Is Dead
AI agents are overwhelming the internet, breaking old business models and forcing infrastructure changes.
HostsMatt TurckGuestsMatthew Prince (CEO, Cloudflare)
The GPU Myth: State of AI Compute 2026 | Stephen Balaban
AI compute is not a commodity; building AI factories requires massive physical infrastructure.
HostsMatt TurckGuestsStephen Balaban (CTO, Lambda)
OpenAI's Dan Roberts: Why AI Can Now Make Discoveries
AI is starting to make scientific discoveries, moving beyond tools to become a collaborator.
HostsMatt TurckGuestsDan Roberts (Lead of the Foundations of Reinforcement Learning team, OpenAI)
State of Enterprise AI 2026: Aaron Levie on Tokenmaxxing, Rise of Headless, and AI-Proofing Your Job
Enterprise AI is costly and slow to adopt, despite lab progress. Hear why.
HostsMatt TurckGuestsAaron Levie (co-founder and CEO, Box)
OpenAI's Yann Dubois: Why AI Progress Suddenly Feels Real
OpenAI's Yann Dubois explains how AI is becoming reliable and useful, not just capable.
HostsMatt TurckGuestsYann Dubois (co-lead of the Post-training Frontiers team, OpenAI)
The Agent Harness: Building Secure Sandboxes for Autonomous AI Workloads
Secure sandboxes are essential for AI agents, moving beyond current cloud infrastructure.
HostsMatt TurckGuestsIvan Burazin (CEO, Daytona)
OpenAI Board Member Zico Kolter on the Real Risks of Frontier AI
OpenAI's board member details real AI risks and safety checks before model releases.
HostsMatt TurckGuestsZico Kolter (Board Member, OpenAI)
Anthropic’s Felix Rieseberg: Claude Cowork, Mythos, and the SaaS Extinction
Anthropic's Felix Rieseberg details Claude's agentic leap and the practicalities of AI interacting with your computer.
HostsMatt TurckGuestsFelix Rieseberg (Engineering Lead for Claude Cowork, Anthropic)
AI is Already Building AI | Google DeepMind’s Mostafa Dehghani
AI researchers explore how AI might automate its own research and development.
HostsMatt TurckGuestsMostafa Dehghani (AI researcher, Google DeepMind)
Benedict Evans: OpenAI’s Moat Problem & the Future of Software
AI models lack network effects, making defensibility a major challenge for companies like OpenAI.
HostsMatt TurckGuestsBenedict Evans (independent tech analyst)
Everything Gets Rebuilt: The New AI Agent Stack | Harrison Chase, LangChain
AI agents need more than models; this episode explores the essential infrastructure that makes them work.
HostsMatt TurckGuestsHarrison Chase (co-founder and CEO, LangChain)
AI That Can Prove It’s Right: Verification as the Missing Layer in AI — Carina Hong
AI that proves its answers right could unlock reliable reasoning and accelerate research.
HostsMatt TurckGuestsCarina Hong (founder/CEO, Axiom Math)
Voice AI’s Big Moment: Why Everything Is Changing Now (ft. Neil Zeghidour, Gradium AI)
Voice AI is finally natural, potentially replacing text as the default AI interface.
HostsMatt TurckGuestsNeil Zeghidour (CEO, Gradium AI)
Mistral AI vs. Silicon Valley: The Rise of Sovereign AI
Mistral AI builds the AI plumbing, not just the magic trick, for true data ownership.
HostsMatt TurckGuestsTimothée Lacroix (Co-Founder & CTO, Mistral AI)
Dylan Patel: NVIDIA's New Moat & Why China is "Semiconductor Pilled”
Understand the AI hardware race and NVIDIA's evolving strategy.
HostsMatt TurckGuestsDylan Patel (SemiAnalysis)
State of LLMs 2026: RLVR, GRPO, Inference Scaling — Sebastian Raschka
Understand LLM evolution with expert insights on reasoning, scaling, and future directions.
HostsMatt TurckGuestsSebastian Raschka
The End of GPU Scaling? Compute & The Agent Era — Tim Dettmers (Ai2) & Dan Fu (Together AI)
Debate AGI's future: physical limits versus system potential, and master agents to boost productivity.
HostsMatt TurckGuestsTim Dettmers (Assistant Professor, CMU / Allen Institute for AI), Dan Fu (Assistant Professor, UC San Diego / Together AI)
Are AI Evals Broken? Anthropic/NYU’s Pavel Izmailov on LLM Evaluation, Reasoning & “Alien” Behavior
Understand AI's "alien" behaviors and evaluation limits with an Anthropic researcher.
HostsMatt TurckGuestsPavel Izmailov (Research Scientist, Anthropic)
DeepMind Gemini 3 Lead: What Comes After "Infinite Data"
Learn how Google DeepMind builds frontier AI systems beyond just more data.
HostsMatt TurckGuestsSebastian Bourgeaud (pre-training lead for Gemini 3, Google DeepMind)
What’s Next for AI? OpenAI’s Łukasz Kaiser (Transformer Co-Author)
OpenAI's Transformer co-author explains why AI progress is still accelerating behind the scenes.
HostsMatt TurckGuestsŁukasz Kaiser (Leading Research Scientist, OpenAI)
Open Source AI Strikes Back — Inside Ai2’s OLMo 3 ‘Thinking"
Understand the full open-source AI pipeline and the global competition beyond just open weights.
HostsMatt TurckGuestsNathan Lambert (Ai2), Luca Soldaini (Ai2)
Intelligence Isn’t Enough: Why Energy & Compute Decide the AGI Race – Eiso Kant
Compute and energy infrastructure, not just intelligence, will decide the AGI race.
HostsMatt TurckGuestsEiso Kant (Co-CEO & Co-Founder, Poolside)
State of AI 2025 with Nathan Benaich: Power Deals, Reasoning Breakthroughs, Real Revenue
AI's future hinges on power, reasoning, and revenue, not just GPUs.
HostsMatt TurckGuestsNathan Benaich (Founder and General Partner, Air Street Capital)
Are We Misreading the AI Exponential? Julian Schrittwieser on Move 37 & Scaling RL (Anthropic)
AI progress is faster than we think, with agents reaching expert breadth soon.
HostsMatt TurckGuestsJulian Schrittwieser (top AI researcher, Anthropic)
How GPT-5 Thinks — OpenAI VP of Research Jerry Tworek
Understand how advanced AI models like GPT-5 actually reason and learn from an OpenAI insider.
HostsMatt TurckGuestsJerry Tworek (VP of Research, OpenAI)
Sonnet 4.5 & the AI Plateau Myth — Sholto Douglas (Anthropic)
AI is advancing rapidly, especially in coding, with human-level performance near.
HostsMatt TurckGuestsSholto Douglas (top AI researcher, Anthropic)
Page generated Oct 19, 2026. Feed checked every six hours.