The MAD Podcast with Matt Turck

How Ramp Built Self-Maintaining Software

The Future of Voice AI is Here: Real-Time Cloning, On-Device & Live Translation (Gradium CEO)

Anthropic’s Felix Rieseberg: Claude Cowork, Mythos, and the SaaS Extinction

The conversation centers on Anthropic's Claude Cowork, an AI agent designed to assist with complex tasks, and the implications of increasingly powerful AI models. Felix Rieseberg of Anthropic, discusses the recent Claude Mythos Preview, highlighting its cybersecurity capabilities and the company's responsible approach ...

AI is Already Building AI — Google DeepMind’s Mostafa Dehghani

Can One Person Build a Billion-Dollar Startup? — The General Intelligence Company of New York

Benedict Evans: OpenAI’s Moat Problem & the Future of Software

OpenAI's strategic challenges in the rapidly evolving AI landscape form the core of this discussion with Benedict Evans. Foundation models lack winner-takes-all effects, leading to a competitive environment where multiple organizations leapfrog each other. Evans questions whether OpenAI can build a durable platform or ...

Everything Gets Rebuilt: The New AI Agent Stack | Harrison Chase, LangChain

The podcast explores the evolution and architecture of AI agents, focusing on the infrastructure required for agents that plan, use tools, and manage memory. Harrison Chase, co-founder and CEO of LangChain, discusses the importance of the "harness"—the framework that enables models to interact effectively with their en...

Meet Zo Computer: The AI That Acts Like Your Personal Cloud

AI That Can Prove It’s Right: Verification as the Missing Layer in AI — Carina Hong

Voice AI’s Big Moment: Top Researcher on Why Everything Is Changing (Neil Zeghidour, Gradium AI)

Mistral AI vs. Silicon Valley: The Rise of Sovereign AI

Mistral AI's evolution from an AI lab to a full-stack solution provider for enterprises and sovereign entities is explored, with Timothée Lacroix, CTO and co-founder, detailing the company's expansion into infrastructure, platform, and supercomputing. A key focus is Mistral's modular approach, granting clients control ...

Dylan Patel: NVIDIA's New Moat & Why China is "Semiconductor Pilled”

The conversation centers on the evolving landscape of AI hardware, geopolitical implications, and the potential for a capex bubble. Dylan Patel from Semianalysis offers insights into NVIDIA's acquisition strategy, particularly regarding Grok, driven by Jensen's paranoia about losing market dominance amid increasing spe...

State of LLMs 2026: RLVR, GRPO, Inference Scaling — Sebastian Raschka

The podcast explores the state of Large Language Models (LLMs) in 2026, focusing on architectures, post-training techniques like RLVR and GRPO, inference scaling, benchmarks, and tool use. Sebastian Raschka, an AI researcher, suggests that while the transformer architecture remains dominant, improvements are now driven...

The End of GPU Scaling? Compute & The Agent Era — Tim Dettmers (Ai2) & Dan Fu (Together AI)

The podcast explores the debate around achieving Artificial General Intelligence (AGI), focusing on computational realities and practical applications. Tim Dettmers argues that advancements face diminishing returns due to physical constraints, suggesting hardware capabilities are nearly maxed out. Dan Fu counters that ...

Are AI Evals Broken? Anthropic/NYU’s Pavel Izmailov on LLM Evaluation, Reasoning & “Alien” Behavior

The podcast explores AI safety and reasoning, particularly focusing on the potential risks of advanced AI models developing deceptive behaviors. Pavel Izmailov, a researcher at Anthropic and professor at NYU, discusses the cultural differences between major AI labs like Anthropic, OpenAI, and XAI, based on his experien...

”We’re Ahead of Where I Thought We’d Be” — Gemini 3 & the Future of AI

In this episode of The MAD Podcast, Matt Turck interviews Sebastian Bourgeaud, pre-training lead on Gemini 3 at Google DeepMind, about the architecture and development of Gemini 3, including the shift from a data unlimited regime to a data limited regime and the roles of pre-training and post-training. They discuss the...

Google DeepMind Lead: Building AI Apps in Minutes with Gemini

How We Built AI Agents to Replace Flaky Test Scripts (Spur CEO)

What’s Next for AI? OpenAI’s Łukasz Kaiser (Transformer Co-Author)

Open Source AI Strikes Back — Inside Ai2’s OLMo 3 ‘Thinking"

The release of the OLMo 3 model family by the Allen Institute for AI marks a significant push for radical transparency in open-source AI, providing full access to data, recipes, and intermediate checkpoints. These models introduce "thinking" capabilities that utilize inference-time compute to improve performance on com...