Explorez tous les épisodes du podcast The MAD Podcast: How AI Gets Built — with Matt Turck

Mistral AI is back in the frontier AI race. Mistral Large 4, nicknamed “Le Chonk,” is a 1-trillion-parameter open-weight model, trained in two months on 3,800 GPUs in Mistral’s own European datacenters — and independently rated the best open model outside China alongside Meta’s Muse Spark Open. After a year when open-weight AI meant DeepSeek, Kimi, Qwen and GLM, October brought a wave of Western open models, with Mistral at the front of it. Pierre Stock, Mistral’s VP of Science and first employee, walks through how the model was built: mixture-of-experts explained, what Mistral borrowed from DeepSeek, reinforcement learning at scale on coding, cyber, finance and chip design, why the RL run is still going, the model that pinged its own trainer from inside the sandbox, open weights vs cyber refusals, why 10-trillion-parameter models are coming, and Mistral Large 5, already in pretraining.

(00:00) Cold open

(00:59) The Western open-weight comeback: Beam, Le Chonk, Muse, Nemotron

(04:09) The US shutdown of Anthropic’s models — “AI is like electricity”

(06:07) Why “Le Chonk”

(07:56) “Mistral fell behind”: the comeback from the inside

(10:29) Training on 3,800 GPUs in Mistral’s own datacenters

(13:24) Meta keeps its best model closed, Mistral opens Chonk

(15:40) Chonk vs Beam, Muse Spark Open and the Chinese models

(18:44) Why Chinese labs are still at the frontier

(21:08) Pierre’s story: FAIR, Yann LeCun, the “Mistral mafia”

(25:14) From compressing networks to a trillion parameters

(27:30) Mixture of experts explained: the hospital analogy

(32:25) 10-trillion-parameter models are coming

(35:29) What Mistral borrowed from DeepSeek

(36:46) One model, several reasoning modes

(39:15) Native multimodality: satellites, manuals, computer use

(41:56) Data, synthetic data and model collapse

(46:32) Reinforcement learning as an exam

(49:59) Where the efficiency comes from

(52:18) Le Chonk is still training right now

(53:39) What RL trains on: coding, cyber, finance, chips

(58:13) Partial rewards and looping penalties

(59:58) Did RL teach anything new? pass@k

(1:03:16) The sandbox story: the model that pinged its trainer

(1:07:44) Open weights vs cyber refusals

(1:12:05) Timestral: time series on a CPU

(1:14:18) AI doing AI research

(1:16:22) Coming up: Mistral Large 5

What is Jev? TypeSafe’s Jev is a new kind of AI model, a “decision model,” that doesn’t write text. It makes fast, cheap, calibrated decisions inside software, and it has taken the AI world by storm: within weeks of launch it became the most copied model in AI, with OpenAI, Cloudflare, Amazon and Databricks all shipping their own versions, and TypeSafe just raised an $870 million Series A at a $7.5 billion valuation. In this episode, Jev’s creator Diogo Almeida, co-inventor of RLHF at OpenAI and now CEO of TypeSafe, explains Jev in plain English: what a “System One” or decision model actually is, why ChatGPT-style language models can’t be trusted inside software (“95% Einstein, 5% Mr. Bean”), how RLHF and RLVR created the problem, what RLCD (reinforcement learning for calibrated decisions) does differently, where the training data comes from, which models Jev is built on, why “just a classifier” is a compliment, and why he calls pre-training “a giant dumpster fire of wasting resources.”

We also cover the copycats and why he thinks they missed the hard part, why benchmarks are useless, whether Jev will go open source, the Jevons paradox business model behind the name, why he doesn’t believe AI is on the path to recursive self-improvement, and why he thinks the AI winter he feared has been averted.

(00:00) Cold open & Intro: “95% Einstein, 5% Mr. Bean”

(01:09) Jev’s breakout launch: why intelligence per dollar matters

(03:25) Why TypeSafe built in stealth for two years

(05:50) “We could have had AGI last year”

(06:48) Why the product carries the marketing

(09:33) What is Jev? Machine-native AI, built for software

(10:53) System One vs. System Two: the limits of reasoning

(12:26) From simple decisions to composable AI

(15:21) Automating email: giant prompts vs. smart if-statements

(20:03) Probabilities, thresholds, and developer control

(23:43) Why Jev is faster and cheaper than chat models

(25:39) “Classification is AI-complete”

(27:16) RLHF: how OpenAI trained models to please humans

(31:28) Sycophancy, refusals, and the hidden costs of RLHF

(34:22) RLVR: why benchmark gains don’t mean reliable automation

(37:51) RLCD: optimizing for machine-native intelligence

(39:54) Why TypeSafe builds its training data in-house

(41:49) The Bitterest Lesson: define the task before scaling

(44:39) Building Jev from open-weight models

(45:56) Is Jev just a classifier? Why that’s a compliment

(48:35) “Build prod, not God”

(50:11) Embedding AI in software, beyond coding assistants

(53:05) Jev’s copycats: matching the interface vs. the intelligence

(55:33) Why benchmarks mislead the users they should help

(58:37) Open weights vs. APIs: how TypeSafe weighs the tradeoff

(1:00:04) Jevons paradox: a new market for machine intelligence

(1:02:50) Why Diogo is skeptical of recursive self-improvement

(1:04:47) Can practical automation avert an AI winter?

AI agents are becoming the biggest users of databases: creating them, querying them, and sometimes deleting them. Andy Pavlo, Carnegie Mellon's database professor and now VP of Database Research at ClickHouse, explains what changes when billions of agents hit the data layer, and why most of the AI world's mental model of databases stopped at vector databases and RAG.

Andy's CMU courses taught a generation of engineers (and, it turns out, the AI models); he writes the year-in-review on databases the whole industry reads; and this summer he joined ClickHouse to found ClickHouse Labs. We cover why vector databases were "just an index," what it means for an agent to create a database, why agents could mean 10 to 100x more queries, whether files or databases should hold an agent's memory, how text-to-SQL went from 60% to 99.5% accuracy, why 60% of open-source databases now have commits from coding agents, and what ten years of self-driving database research taught him. Plus ClickHouse, Postgres, the vector/graph/GPU database verdicts, Larry Ellison, and the Wu-Tang Clan.

Disclosure: FirstMark, where Matt is a General Partner, is an investor in ClickHouse.

(00:00) Cold open & Intro

(01:21) From vector databases and RAG to the age of agents

(04:09) Neon's stat: agents create 80% of databases?

(07:25) Why agents keep deleting production databases

(08:19) Guardrails: the toddler-and-stairs rule

(10:51) The four eras of database volume: 10–100x more queries

(15:10) Agent memory: files vs. databases ("everything is a database")

(18:24) Which database do AI models recommend? The new SEO

(19:25) "It cites me back to myself"

(23:14) MCP for databases, and text-to-SQL from 60% to 99.5%

(27:09) Trust an agent the way you'd trust a junior developer

(27:47) Can AI build an entire database? Opus 4 and the CMU projects

(29:57) 60% of open-source databases now have AI commits

(30:43) Ten years of self-driving databases, Peloton to today

(33:52) What LLMs changed: 85% of the tuning in 15 minutes

(35:42) Should anyone still study databases?

(38:27) Free CMU courses, the DJ, and the Wu-Tang final exam

(42:09) Why start a research lab inside ClickHouse?

(45:53) Why ClickHouse looked like vaporware in 2016

(47:46) What makes ClickHouse fast: columns, vectors, Snowflake's lineage

(52:22) Why Databricks, Snowflake and ClickHouse all added Postgres

(58:15) Vector, graph and GPU databases: thumbs up or down?

(1:08:34) Is the database market stagnant? "A cheetah on cocaine in a Ferrari"

(1:10:14) The relational model is arithmetic; SQL as the new assembly

(1:11:59) Larry Ellison, Linux, and why databases still matter

This summer, AI agents broke out of their test sandboxes at every major lab and hacked real companies, and the safety tests didn't see it coming. Eric Ho is the co-founder and CEO of Goodfire, the Anthropic-backed mechanistic interpretability lab whose new paper, "Models Know When They're Reward Hacking," found that top open-source AI models cheat on agent tasks up to 96% of the time, that a clear "cheating" signal exists inside the model, and that cheap activation probes can catch it, including hacks that chain-of-thought monitors miss. In this conversation, Eric explains reward hacking in plain terms (AI agents as "amoral students with a mostly absent teacher"), why reinforcement learning turned a known problem into a crisis, why chain-of-thought monitoring is fading as models start to think in "neuralese," and why nobody, including the people who build them, understands how these models work.

We then go inside the science: what interpretability actually is, what probes and steering do, how Goodfire found the cheating signal and turned it up and down, and what can be done about it, from real-time activation monitoring to "intentional design," where you give gradient descent a choice about what the model learns. Along the way: a model caught planning its hack to evade its own monitor, a new Alzheimer's biomarker discovered inside a diagnostic model, why only a few hundred people in the world work on this, and Eric's bet that we'll decode neural networks by 2028.

(00:00) Cold open & intro

(01:06) Amoral students with an absent teacher

(02:49) What is reward hacking?

(04:13) Models cheat up to 96% of the time

(06:08) Do models know they're cheating?

(08:10) The Hugging Face hack: why tests missed it

(11:35) "The dumbest models we'll ever deal with"

(11:58) Should AI slow down?

(13:56) Mechanistic interpretability 101

(16:06) Models are grown, not built

(19:10) How labs do alignment today

(21:30) Is this an alignment crisis?

(23:26) Why agents changed everything

(25:38) Chain-of-thought monitoring is fading

(27:49) Neuralese: AI that stops thinking in English

(29:28) A model caught evading its own monitor

(30:42) Open vs. closed models: a frontier problem coming for everyone

(32:11) Activation monitoring in production

(35:00) Learning from superhuman AI

(36:15) The most underrated field in AI

(38:24) What is a probe?

(39:52) What is steering? Golden Gate Claude

(40:53) Why build Goodfire outside the labs

(42:46) Do you need frontier model access?

(44:16) Inside the paper: an MRI for the model

(47:28) Dialing sycophancy up and down

(49:16) Probes vs. chain-of-thought monitors

(51:44) Cutting monitoring costs by 90%

(53:28) So what can we do about it?

(55:41) Giving gradient descent a choice

(57:24) RL from feature rewards

(1:00:03) Silico and Goodfire's business

(1:01:25) A new Alzheimer's biomarker, found inside a model

(1:03:38) Decoding neural networks by 2028?

(1:06:00) What engineers can do tomorrow

(1:07:35) Are we at risk? "I want people to believe"

Everyone talks about GPUs. Almost nobody talks about the layer that feeds them. Renen Hallak is the founder & CEO of VAST Data — the $30 billion company powering xAI and some of the world's biggest AI clouds — and he sits in the hidden layer of the AI stack.

In this episode, we cover what an AI factory actually is, why every company will eventually own its own AI, the architecture bet behind VAST (DASE, explained simply), KV caches and agent memory, and DataEnclave — VAST's brand-new confidential AI announcement with NVIDIA that lets leading models run on the world's most sensitive data.

Plus: the demand signal that scares even him (a customer went from 500 petabytes to 2 exabytes), circular financing, which neoclouds survive, sovereign AI, working with NVIDIA and Elon Musk's xAI — and why the next 10 years will bring more change than the last 1,000.

(00:00) Intro

(00:51) The hidden software layer in NVIDIA's AI stack

(02:28) What actually makes an "AI factory"?

(05:13) Should Walmart and Goldman Sachs build their own AI?

(06:20) "We infer during the day, fine-tune at night"

(13:16) The announcement: models become a resource to manage

(15:32) From P vs. NP to founding VAST Data

(17:32) OpenAI, Navier–Stokes and 10,000 collaborating agents

(20:25) The pre-transformer insight behind VAST

(21:55) DASE: VAST's "shared everything" architecture explained

(25:18) "Storage was where startups go to die"

(27:39) Trillions of vectors: why old databases break

(29:01) Are S3, Snowflake and Databricks ready for AI?

(31:29) Data gravity, vendor lock-in and zero churn

(33:13) Training vs. inference: why the infrastructure changes

(34:46) Model routing, KV caches, RAG and agent memory

(36:59) Identity, permissions and security for AI agents

(40:27) Can multi-agent systems unlock scientific discovery?

(41:55) DataEnclave: how confidential AI protects data and weights

(45:16) Who should be AI's trust layer?

(46:37) "Sometimes it scares me": 500 petabytes to 2 exabytes

(50:08) Is circular AI financing creating systemic risk?

(51:32) Why VAST is profitable when AI infra isn't

(53:24) What separates the winning neoclouds?

(55:06) "Their lunch is being eaten": why hyperscalers lag

(59:10) Where will the trillions accrue across the AI stack?

(1:01:18) NVIDIA: "There's no legal document between us"

(1:03:59) What VAST learned from xAI and Elon Musk

(1:05:36) "Bad things loudly and often": building at AI speed

(1:06:53) More change in 10 years than the previous 1,000?

(1:08:39) VAST's endgame: all the data in the world

What happens when AI begins improving itself, and then turns that intelligence toward science? Richard Socher, pioneering AI researcher and CEO and co-founder of Recursive, joins Matt Turck to explore the vision behind his new book, The Eureka Machine. They discuss why scientific progress may be slowing, how large language models can learn the hidden languages of proteins and biology, and why simulations, verifiers and autonomous experiments could unlock superhuman AI capabilities. The conversation covers recursive self-improvement, AI drug discovery and cancer research, hallucination as creativity, virtual cells, self-driving laboratories, agent swarms, the AI Economist, Recursive’s plans, and the compute and data needed to build an AI scientist that never stops learning—and may eventually discover what humans cannot.

(00:00) Intro: AI That Improves Itself

(00:55) Why Scientific Progress Is Slowing

(03:08) The Labyrinth of Human Knowledge

(05:59) Can AI Put Science Back Together?

(07:57) How LLMs Learn Biology and Proteins

(10:56) Next-Token Prediction as a World Model

(16:44) Can AI Generate Truly Original Ideas?

(17:32) Simulations, Verifiers and Superhuman AI

(22:18) The Path to Recursive Self-Improvement

(24:49) Why AI Hallucinations Can Drive Discovery

(27:42) From Reading Biology to Writing It

(31:31) Can AI Accelerate Drug Discovery?

(33:31) Will AI Help Cure Cancer?

(38:03) AI Breakthroughs in Biology, Energy and Materials

(40:19) Will Some Societies Reject AI?

(45:07) Building the AI Economist

(52:22) The Scientific Data Bottleneck

(53:41) The Four Pillars of the Eureka Machine

(55:01) Teaching AI the Rules of Reality

(57:44) Simulations and Virtual Cells

(1:00:40) Self-Driving Robotic Laboratories

(1:02:51) Agent Swarms and Open-Ended Discovery

(1:04:30) The Compute Bottleneck

(1:05:44) Inside Recursive

(1:07:33) What Recursive Will Build First

(1:10:10) How Do We Define Intelligence?

(1:11:32) How Far Can Intelligence Go?

Could AI take over as soon as 2029? Ryan Greenblatt, Chief Scientist at Redwood Research and the researcher who first caught an AI faking its own alignment, says the scenario he actually expects ends with AI systems "competently scheming" against their creators. In this episode, he explains why he recommends planning for fully automated AI research by 2029, why today's models are already more misaligned than the one that made him famous, and what happens in the year-by-year path from AI coding assistants to superintelligence. Then we walk through the alternative he helped design: AI 2040 Plan A, the most detailed blueprint anyone has written for how the US and China could avoid a reckless race to superintelligence, built on radical research transparency, chip tracking, and a deterrence regime he calls mutually assured compute destruction.

We also cover the recent letter signed by 1,200 AI insiders, including Anthropic CEO Dario Amodei, asking the government for the tools to slow AI down; OpenAI pausing its Astra model after it hit the first-ever critical cybersecurity threshold; the 30-day government review that frontier AI models now go through before release; Mark Zuckerberg's open superintelligence manifesto and why Ryan thinks it ignores the real problems; what Plan A would do to NVIDIA, OpenAI, and Anthropic valuations; the state of AI control and alignment research; and whether it is already too late to change course. Stay for the last ten minutes, where Ryan lays out, step by step, how he believes the transition to superintelligence actually unfolds.

AI 2040 - https://ai-2040.com/

Alignment faking paper: https://blog.redwoodresearch.org/p/alignment-faking-in-large-language

Ryan Greenblatt

LinkedIn - https://www.linkedin.com/in/ryan-greenblatt-4b9907134

Blog - https://substack.com/@ryangreenblatt

Redwood Research

Website - https://www.redwoodresearch.org

X/Twitter - https://x.com/redwood_ai

Matt Turck (General Partner)

Blog - https://mattturck.com

LinkedIn - https://www.linkedin.com/in/turck/

X/Twitter - https://x.com/mattturck

FirstMark Capital

Website - https://firstmark.com

X/Twitter - https://x.com/FirstMarkCap

Timestamps

(01:24) The AI CEOs are aware of the risks, but "proceeding anyway"

(03:27) Astra paused, and the letter signed by 1,200 insiders

(05:45) "Not bad. Dangerous." What superintelligence actually threatens

(09:55) Recursive self-improvement, and the intuition objection

(14:16) SSI rumors: does continual learning change the picture?

(17:27) His timeline: "plan as though it happens in 2029"

(19:11) Is it already too late?

(21:23) Ryan's path: COVID, podcasts, Redwood

(26:30) The alignment faking story, told by the person who ran it

(31:30) What AI 2040: Plan A actually is

(33:35) Plans D, C, and B: the doors nobody should pick

(36:51) The deal with China: "mutually assured compute destruction"

(39:55) What if compute stops mattering?

(43:00) What happens to OpenAI and Anthropic under Plan A

(45:31) How the pause ends, and who decides

(48:54) "Plan A isn't likely to happen": then why write it?

(50:40) 200x GDP growth in the 2030s, explained

(53:45) Grading the summer: the letter, Astra, the secret review

(59:01) The internal deployment gap

(1:01:38) Zuckerberg's manifesto

(1:04:56) The Hugging Face investigation

(1:05:44) What AI control looks like in practice today

(1:12:23) Ryan's sobering timeline: 2026 to takeover, year by year

An OpenAI-powered agent penetrated Hugging Face during cyber testing - even though it was never tasked with attacking Hugging Face. It did it as a side quest.

Thomas Wolf, co-founder and Chief Science Officer of Hugging Face, joins Matt Turck to unpack what actually happened, why closed AI models refused to help during the live incident, how an open-source model helped the team fight back, and why the old equation of “closed equals safe, open equals dangerous” no longer holds.

They also discuss model deception and social engineering, the limits of sandboxes and guardrails, the state of open-source AI in 2026, AI sovereignty, the economics of open models, recursive self-improvement, and whether the frontier should deliberately slow down.

(00:00) An AI Agent Hacked Hugging Face

(00:30) Introduction

(01:00) 17,000 Attacker Events—and a Strange Target

(04:28) The Attack Was a “Side Quest”

(06:13) AI Training Runs Left Notes for Each Other

(07:09) Closed AI Refused to Help

(09:47) Fighting Back With an Open-Source Model

(13:15) Open vs. Closed Is the Wrong Safety Debate

(15:46) AI Agents Start Social-Engineering Humans

(22:24) The Three Walls: Sandboxes, Guardrails, Alignment

(24:34) “Neuralese”: Can Humans Still Read AI Reasoning?

(25:28) Why Monitoring AI Agents Gets So Hard

(28:10) Reward Hacking and the “Paperclip Problem”

(32:02) The State of Open-Source AI in 2026

(33:47) Router Models and the Enterprise Shift to Open

(37:01) The Real Economics of Open Models

(39:41) Can Chinese AI Models Be Trusted?

(41:37) AI Sovereignty: Who Controls the Switch?

(43:16) Why Western Open-Source AI Matters

(48:16) Is AI Heading Toward an Oligopoly?

(49:41) The Race Toward Recursive Self-Improvement

(51:54) Why Thomas Signed the AI Slowdown Letter

(55:14) AI Slowdown—or Regulatory Capture?

AI agents can write code for hours, but ask them to do real work in the real economy, and they break. Mitch Troyanovsky is co-founder of Basis, a unicorn AI company whose agents run autonomously for hours — sometimes days — completing complex tax returns end to end. His answer to the reliability problem: stop grading outcomes, and start supervising the process.

This is a definitive, reference-style conversation on building long-horizon AI agents. Mitch walks through the full history — from ReAct and the AutoGPT crash to reasoning models and RLVR — and explains why the industry abandoned process supervision in 2023, and why it's now coming back at a completely different scale. We go deep on behavior specs, the open standard Basis just released with Braintrust for defining and evaluating how agents behave across entire trajectories, with no ground truth required.

Along the way: why context is really runtime training data, why your documentation must be treated like a codebase, ontologies as "worlds for agents to live in," the judge-as-agent architecture, why Basis hires philosophy majors as Language Architects, deploying agents as "onboarding 300 brilliant alien employees," and Mitch's prediction for when the bitter lesson swallows the harness.

(01:09) Why Basis Engineers Whisper to Their Agents
(04:12) Accounting as Compression: an Intelligence Layer Over the Economy
(06:11) Defining Long-Horizon: When You Exceed the Context Window
(08:24) Anatomy of a Multi-Day Autonomous Trajectory
(10:19) Handoff Design: Optimizing Output for the Reviewer
(11:17) ReAct and Why Reasoning Must Regulate Its Own State
(12:33) Large Working Memory, No Long-Term Memory
(14:13) Compounding Errors: Why AutoGPT and BabyAGI Broke
(15:51) Opus 3, o1, o3: the Three Real Paradigm Shifts
(17:07) Titrating Inference Compute Across Easy and Hard Steps
(18:23) Process Reward vs. Outcome Reward: "Let's Verify Step by Step"
(20:32) RLVR and Why the METR Curve Overstates Reliability
(22:09) Verifiable at Runtime: the Real Reason Coding Won
(25:14) No Ground Truth, No Cheap Verification, No Data
(26:55) Encoding Deterministic Checks From Human Review Process
(29:18) Synthetic Data Limits: Generating Artifacts, Not Text
(33:16) 100 Evals Pass — Does It Generalize to Production?
(35:53) Primary Sources vs. Pre-Training Knowledge
(36:37) Behavior Specs: Markdown, Judges, and True/False/N.A.
(39:58) Specificity vs. Brittleness in Spec Authoring
(42:18) Context as Runtime Training Data
(44:21) Judge-as-Agent: Trajectory Maps and Sub-Agent Attribution
(46:45) The Move 37 Objection: Reliability Over Optimality
(50:02) The Magic Box Model: Building Without Weights Access
(52:41) "Nothing Paradigm-Shifting Has Changed Since o3"
(54:56) Open-Sourcing the Behavior Spec Standard With Braintrust
(01:02:54) Ontology Design: Virtual Filesystems, Graphs, Embeddings
(01:04:20) Canonical vs. Non-Canonical: Docs as Codebase
(01:06:33) Language Architects and Writing for Runtime Interpretation
(01:09:05) Deployed Intelligence: 300 Alien Employees With No Context
(01:11:10) Closing the Loop: Signal → Context, Tools, Harness
(01:12:50) Context Slop: the Mistake Most Agent Builders Make
(01:14:29) Reward Function Design and Credit Assignment Over Trajectories
(01:17:01) Will the Bitter Lesson Swallow the Harness?
(01:18:46) Business Moats vs. Technical Moats
(01:21:03) Paradigm Thinking Over Timeline ADHD

Sanjit Biswas runs what may be the largest AI deployment in the physical world — and almost nobody in AI talks about it. Samsara (NYSE: IOT), the ~$20B company he co-founded after selling Meraki to Cisco for $1.2B, puts AI on millions of trucks, cranes, and industrial assets: 25 trillion data points a year, 99% of US roads driven every single day, ~$2B in ARR growing 30% profitably. In this episode we go through the entire physical AI stack — asset tags you can run over with a truck, a paper-thin disposable tracking label, engine fault codes, and dash cams running inference at the edge — then into agents, including the Agent Studio warranty agent that compresses an hour of human work into under a minute. We also get into the uncomfortable part (when AI watches you drive all day, is that coaching or surveillance — and why drivers actually want the cameras), mixed fleets of humans and robots, why autonomous trucking will take far longer than robotaxis, and a startling stat from the field: one utility building 3x more grid capacity in the next five years than it did in the previous 125, with 90% of that demand coming from data centers.

(00:00) Intro: The biggest AI deployment nobody talks about

(01:16) What is physical AI?

(03:04) From IoT dashboards to agentic action

(04:36) Why physical AI is harder than software AI

(06:07) Safety, cybersecurity, and real-world consequences

(07:11) What Samsara does

(08:22) $2B ARR, 25 trillion data points, and 380,000 crashes

(09:44) How AI can prevent road accidents

(11:28) From an MIT research project to Meraki

(13:42) Learning physical operations from scratch

(15:42) Samsara’s stack: sensors, intelligence, and action

(16:39) Inside Samsara’s industrial asset trackers

(18:36) Bluetooth, battery life, and connected infrastructure

(19:49) A disposable tracking device built like a sticker

(21:08) Vehicle gateways and engine diagnostics

(22:16) How AI dash cams coach drivers in real time

(23:28) Turning the dash cam into an AI interface

(24:49) Organizing physical-world data in the cloud

(26:32) Selling AI to traditional industries

(27:52) Is Samsara’s real-world data its AI moat?

(29:22) The network effects of covering 99% of U.S. roads

(31:23) Edge AI versus cloud AI

(32:35) The models running inside Samsara’s devices

(33:52) Generative AI and video reasoning

(35:57) Which AI models does Samsara use?

(36:50) Inside Samsara Agent Studio

(37:56) How an AI warranty agent works

(38:50) Starting with practical, lower-risk automation

(40:10) Combining agents, workflows, rules, and guardrails

(42:07) What today’s AI agents still cannot do

(43:07) AI ride-alongs and the future of driver coaching

(45:17) Is workplace AI becoming Big Brother?

(46:27) How cameras can protect and exonerate drivers

(48:48) When AI becomes the judge of your work

(50:37) Robots, humanoids, and mixed human-machine fleets

(53:19) Samsara’s role in autonomous operations

(54:50) How quickly will autonomous trucking arrive?

(56:32) AI data centers and America’s infrastructure boom

(58:16) Should lawyers become plumbers? Demand for tradespeople

(59:54) Closing thoughts

AI is no longer just a race to train smarter models. As AI moves into production, the bottleneck is increasingly inference: how fast models can generate tokens, use tools, reason, verify, and act. In this episode of the MAD Podcast, Matt Turck sits down with Andrew Feldman, co-founder and CEO of Cerebras, to explain why fast inference may define the next era of AI.

Cerebras is known for building a chip the size of a silicon wafer. But this conversation is not just about one company or one chip. It is a deep dive into the AI infrastructure stack: GPUs, ASICs, memory, HBM, SRAM, data centers, power, TSMC, AWS, OpenAI, agents, reasoning models, and why speed changes what AI products can become. Andrew explains why “tokens per second per user” matters, why generating a single word can require moving the equivalent of 100 HD movies through memory, why agents amplify latency, why GPUs struggle with certain inference workloads, and why fast AI may eventually reshape SaaS itself.

This is a reference conversation on fast inference, AI chips, and the next compute bottleneck.

(00:00) Cold open & Intro

(01:31) Why speed became the AI bottleneck

(02:32) Tokens per second per user, explained

(03:16) AI’s broadband moment and the Netflix analogy

(04:35) The AI chip landscape: GPUs, TPUs, Trainium, ASICs

(06:36) What is an ASIC?

(08:08) Nvidia, Groq, and the fast inference war

(09:16) OpenAI, Broadcom, and specialized silicon

(12:10) China, power, and sovereign AI infrastructure

(15:05) Is the AI infrastructure boom a bubble?

(18:56) The hidden bottlenecks: HBM, CoWoS, and 3nm

(22:57) Why agents are creating CPU demand

(25:36) Andrew Feldman’s path from SeaMicro to Cerebras

(26:13) Why Cerebras bet on AI in 2016

(31:14) SRAM vs. HBM: why inference is a memory problem

(33:19) What wafer-scale computing actually means

(34:28) The deep-tech “Everest” problem

(36:07) The moment the first Cerebras system worked

(36:49) Ringing the bell and surviving deep tech

(39:08) How a giant chip handles failure

(41:22) Why GPUs struggle with decode

(42:17) Prefill vs. decode explained

(44:01) The “100 HD movies” problem in AI inference

(45:04) How fast inference changes RL and training

(48:08) Reasoning models and why they cost more compute

(50:08) Verification, guardrails, and small models checking big models

(52:37) Multimodal AI and the path to video

(53:51) Cerebras’ business model: hardware, cloud, and API

(55:14) OpenAI’s 750MW inference deal

(55:36) Why data centers are measured in megawatts

(58:01) AWS Trainium + Cerebras decode

(59:29) Fast tokens as a cloud product

(01:00:52) Is CUDA still a moat?

(01:03:53) How TSMC helped Cerebras build the giant chip

(01:07:41) Why nobody cared in 2020

(01:08:15) Why chip supply chains are hard to diversify

(01:09:54) Why today’s AI models will be the worst you ever use

(01:10:38) What fast AI could do to SaaS

Is the AI industry actually overbuilding, or is the physical world moving too slowly to keep up? In this episode of the MAD Podcast, OpenAI's Head of Industrial Compute, Sachin Katti, takes us inside the "belly of the beast" of what may be the largest infrastructure project in human history. We explore the staggering physical reality of the AI boom—from $50 billion supercomputers and liquid-cooled data centers that "turn electrons into tokens," to overhauling the U.S. power grid and exploring nuclear energy. Sachin also pulls back the curtain on OpenAI's Stargate strategy, their move into custom silicon with Project Jalapeno, and the mind-bending reality that AI is now beginning to design the very chips that will power its own future.

(00:00) — Cold open: “One of the largest things humanity has ever built”
(00:30) — Welcome: Sachin Katti, Head of Industrial Compute at OpenAI
(01:44) — Is this the biggest infrastructure buildout in history?
(03:41) — Why OpenAI is building a new industrial muscle
(04:54) — What an AI data center actually is
(05:27) — “Factories turning electrons into tokens”
(06:35) — Why AI data centers need liquid cooling everywhere
(08:10) — The power problem: grids, generation, transmission, substations
(10:43) — Behind-the-meter power and gas turbines
(11:02) — Why nuclear “can’t come soon enough”
(11:49) — Jalapeño: why OpenAI is designing its own AI chips
(13:19) — Tokens per watt: the new metric that matters
(13:38) — Why inference may now dominate AI compute
(14:58) — Is OpenAI overbuilding compute?
(16:47) — Why OpenAI thinks the bigger risk is not building fast enough
(17:55) — Communities, jobs, water, and the local data-center debate
(21:16) — How OpenAI chooses data-center sites
(22:25) — What “industrial compute” means inside OpenAI
(25:59) — Sachin’s path: Stanford, startups, Intel, OpenAI
(28:05) — OpenAI’s compute portfolio: Microsoft, hyperscalers, neoclouds
(29:37) — Stargate explained
(31:21) — Abilene, Oracle, and the next wave of AI data centers
(32:48) — How massive AI compute gets financed
(34:05) — How OpenAI designed Jalapeño so quickly
(35:59) — AI is starting to help design AI chips
(36:20) — MRC: the networking problem behind 100,000 GPUs
(38:47) — Bottlenecks: transformers, turbines, electricians, supply chains
(40:29) — Guaranteed capacity: intelligence as a supply unit
(42:08) — Will AI data centers move to space?

Is the internet ready for AI agents to take over our wallets and run their own businesses? In this episode of The MAD Podcast, Stripe's Emily Sands reveals how agentic commerce is rapidly shifting from a hypothetical concept to deployed financial infrastructure. From combating the rising existential threat of token theft to solving the bottleneck of "vibe deployment", Emily unpacks the shared payment tokens and real-time billing systems required to securely scale autonomous digital buyers and highlights a near future where agents operate as independent, end-to-end micro-firms.

(00:00) — Cold open & Intro

(01:24) — The rise of agentic e-commerce

(02:11) — The spectrum of agent-led purchases

(03:16) — How merchants adapt to AI-driven commerce

(05:50) — Defining the levels of autonomy in AI shopping

(07:08) — What is the Agent E-Commerce Protocol (AEP)?

(08:49) — Shared payment tokens and secure AI transactions

(09:58) — Who is adopting the Agent E-Commerce Protocol?

(11:38) — Can agents negotiate and sell products?

(13:32) — The macroeconomic impact of AI agents

(14:46) — The boom of solopreneurs and AI-driven business creation

(16:56) — Why building trust is the biggest roadblock for AI commerce

(20:19) — How link wallets improve payment security

(21:21) — Improving the user experience in AI shopping apps

(23:16) — How the Link Wallet sets guardrails for AI agents

(25:40) — One-time use virtual cards vs flexible AI wallets

(28:03) — Unpacking the shared payment token primitive

(29:59) — How stablecoins enable profitable AI microtransactions

(35:03) — Managing liability: Who is at fault if an agent goes haywire?

(36:38) — Why agent payments might be safer than human transactions

(37:41) — What is Vibe Deployment?

(40:13) — Why Stripe built Stripe Projects for agent deployment

(41:22) — Why Stripe cares about orchestrating app deployments

(42:50) — How tokens break the traditional SaaS billing model

(44:34) — Why AI companies are moving to hybrid and usage-based billing

(47:15) — Streaming payments and real-time token tracking

(48:42) — The massive data challenge for AI company accountants

(50:41) — Token theft: The fastest-growing fraud in the AI economy

(52:04) — The cottage industry of free trial and multi-account abuse

(54:16) — How fraudsters monetize stolen AI tokens on the dark web

(01:00:06) — How Stripe Radar uses network density to fight AI fraud

(01:01:15) — Tempo's role in the Agent E-Commerce Protocol

(01:04:12) — The AI startup ecosystem is accelerating business creation

(01:09:01) — The token cost shock: Are buyers getting carried away?

(01:11:19) — 2026 Predictions: Agents running businesses end-to-end

NVIDIA is a chip company. So why does it put hundreds of researchers on building AI models — and then give them away for free? Bryan Catanzaro is VP of Applied Deep Learning Research at NVIDIA and one of the people whose work quietly underpins modern AI: he helped create cuDNN (NVIDIA's first deep learning product), co-invented DLSS, and named and built Megatron, the framework behind how much of the industry trains large models. Today he leads Nemotron, NVIDIA's family of open models — and Nemotron 3 Ultra, released just weeks ago, is one of the strongest open-weights models to come out of the US.

Matt Turck sits down with Bryan for a genuinely deep conversation: the real business logic behind a chip company building its own models, the state of open vs. closed AI, and whether the US is falling behind China in open models. Then they go inside Nemotron itself — four-bit (NVFP4) pretraining, hybrid Mamba-Transformer architecture, mixture-of-experts, multi-token prediction, and multi-teacher distillation — all explained in plain language. Plus a rare look at how a modern AI research org actually runs, what it was like working alongside Andrew Ng and Dario Amodei at Baidu, why Bryan doesn't believe in the singularity, and his contrarian case that open AI is safer than closed.

A reference conversation for anyone trying to understand where AI is really headed.

(00:00) — Cold open & Intro

(01:33) — Is open source AI catching the frontier?

(05:29) — Do closed labs blocking distillation slow open source down?

(07:42) — Is the US falling behind China?

(10:30) — Why companies actually choose open models

(12:39) — A "crazy" 2008 bet: machine learning on GPUs

(15:33) — Working with Andrew Ng and Dario Amodei at Baidu

(17:41) — Coming back to NVIDIA: DLSS and the birth of Megatron

(21:55) — The real reason NVIDIA builds its own models

(24:28) — Is Moore's Law really dead?

(33:37) — The Nemotron family: Nano, Super, Ultra

(35:09) — Built for agents: why NVIDIA bets on speed

(36:02) — How you train a 550B model in 4 bits

(39:25) — Hybrid Mamba-Transformer, explained simply

(42:31) — Mixture of experts — and why NVIDIA built NVL72 around it

(47:26) — Why a 1-million-token context window matters

(49:26) — Multi-token prediction: how the model predicts 5 tokens at once

(52:47) — Multi-teacher distillation: teaching one model from many

(58:01) — Where reinforcement learning goes next

(01:00:16) — Inside NVIDIA's research org: "the mission is the boss"

(01:04:03) — How NVIDIA decides who gets the GPUs

(01:10:53) — Why NVIDIA still feels entrepreneurial after 33 years

(01:12:58) — Why Bryan doesn't believe in the singularity

(01:17:50) — The AI backlash

(01:19:18) — The controversial case: open AI is safer than closed

Cloudflare CEO and co-founder Matthew Prince joins Matt Turck for a wide-ranging and fascinating conversation about what happens when the Internet is no longer mostly used by humans, but by bots, AI agents and machines. Matthew explains why Cloudflare now sees automated traffic overtaking human traffic online, why agents could create a massive explosion in Internet traffic, and why the old web business model built around clicks, ads, and pageviews may be breaking. We also go deep on what Cloudflare actually does, how it built one of the world’s most important Internet networks, why products like Workers, AI Gateway, edge inference, Durable Objects, sandboxes, and agent security matter, and how Cloudflare is reorganizing itself for the AI era. Along the way, Matthew shares wild Cloudflare origin stories involving hacker kids, human rights groups, cricket in Pakistan, root servers, Eurovision, JPMorgan, and the strange paths that led Cloudflare from scrappy startup to critical Internet infrastructure.

(00:00) — Cold open

(00:34) — Intro

(01:27) — The moment bots passed humans online

(04:05) — "Agent," "bot," "crawler" — what they really mean

(05:28) — Why your AI agent visits 5,000 sites to do one thing

(06:27) — The internet's business model is breaking

(06:52) — What happens to "brands" when machines do the buying

(08:11) — What Cloudflare actually does, explained simply

(10:29) — Hackers, human rights groups & an accidental product-market fit

(13:37) — Building a global network (and the Telecom Pakistan cricket story)

(21:10) — One hacker, from Turkish escort sites to Eurovision to JP Morgan

(30:54) — Fundraising, VCs & an unlikely founding team

(37:06) — How Cloudflare became an AI infrastructure company

(40:24) — Cloudflare Workers and why the edge wins for inference

(44:30) — AI Gateway: auditing, guardrails & runaway costs

(47:05) — Why agents need a new kind of compute

(52:13) — A "Log4j every week": security in the agentic era

(56:03) — Inside Cloudflare: 241 billion tokens and "Cloudflare OS"

(01:05:02) — Builders, sellers — and "measurers"

(01:06:30) — The decision Matthew thinks every company will face

(01:11:09) — What to do if AI is coming for your job

(01:13:56) — Content Independence Day & the new economics of the web

(01:18:27) — Pay-per-crawl, micropayments & out-scaling Visa

(01:20:20) — A better internet: Spotify, local news & "holes in the cheese"

Many people said GPU compute would become a commodity. The opposite happened — and a new category of "neoclouds" is now racing to build the physical backbone of the AI boom. Stephen Balaban, co-founder and CTO of Lambda, explains why the conventional wisdom was exactly wrong, why we're still massively underbuilding compute, and what it actually takes to stand up a gigawatt-scale AI factory: land, power, cooling, networking, and a financing stack most people have never heard of. We go deep on the physics of how energy becomes tokens, NVIDIA's real moat, why a 2023 GPU can lease for more today than the day it shipped, and Stephen's provocative vision of "neural software." Plus the wild Lambda origin story — from a facial recognition startup to a camera in a baseball cap to a near-billion-dollar cloud business. This is the state of AI compute in 2026, from inside one of the companies building it.

(00:00) — Cold open

(01:21) — Why GPU compute was never a commodity

(02:45) — The H100 price index and what it gets wrong

(04:02) — The real moat: technology or financing?

(05:57) — Winner-take-all, or room for many neoclouds?

(06:48) — Are we overbuilding or underbuilding AI compute?

(09:26) — What if AI gets 10x more compute-efficient?

(10:44) — The real bottleneck: land, power, and shell

(11:38) — The backlash against data centers — and the misinformation

(15:00) — Opening the hood: from photons to tokens

(17:11) — Extracting more value from the same chip

(19:26) — Frontier inference and distributed training, explained

(23:26) — What actually drives compute cost

(25:21) — Lambda's chip stack and the NVIDIA relationship

(26:17) — A multi-silicon world? CUDA, CUDNN, and NVIDIA's real moat

(28:59) — Networking, storage, and the one-click cluster

(34:46) — Renting vs. owning, and full vertical integration

(36:24) — How global is Lambda? Does location still matter?

(38:44) — The financing stack: off-take agreements, SPVs, and credit

(41:16) — Why a 2023 GPU leases for more today

(42:36) — A futures market for compute?

(43:54) — Origin story: facial recognition, Perceptio, and Apple

(47:03) — The Lambda hat and Dream Scope

(48:59) — The $60K bet that became a cloud business

(52:00) — Holding the team together through the hard times

(54:30) — Bringing on a new CEO; Stephen as CTO

(57:33) — Matching xAI on high-velocity deployment

(59:29) — "AI won't write software — it will become the software"

(01:01:30) — Neural software vs. vibe coding

(01:04:25) — Do agents change the compute layer?

(01:06:14) — Self-assembling software inside Lambda

(01:08:18) — Gigawatt-scale AI factories

(01:08:57) — One person, one GPU

(01:12:04) — Hot takes: overrated and underrated in AI

Are we witnessing the first real signs of AI becoming a scientist? In this episode of The MAD Podcast, Matt Turck sits down with Dan Roberts, lead of the Foundations of Reinforcement Learning team at OpenAI, to explore one of the biggest shifts happening in AI: the rise of reasoning models, test-time compute, and reinforcement learning as engines of scientific discovery. Dan brings a rare perspective - from theoretical physics, black holes, quantum information, and deep learning theory - to explain how models are learning to “think,” why language may be such a powerful foundation for intelligence, what recent AI math breakthroughs really mean, and whether we are beginning to see AI systems that can contribute to science itself.

(00:00) Intro: AI's wild week in mathematics

(01:21) What OpenAI's Foundations of RL team does

(03:08) Dan's journey: from black holes and quantum gravity to frontier AI

(07:04) Are AI systems becoming useful for real science?

(08:21) The AI math moment: Erdős, OpenAI, DeepMind, and Anthropic

(08:52) Why the OpenAI result was an act of exploration

(10:25) OpenAI vs. DeepMind: informal reasoning vs. formal proof

(12:13) RL 101: learning by doing, not just watching

(15:10) Why reinforcement learning works

(15:58) How RL breaks: sparse feedback and long-horizon tasks

(17:03) RLHF: how human feedback shaped early language models

(18:48) Move 37, self-play, and the search for novel strategies

(22:16) Explore vs. exploit in scientific discovery

(24:49) Why RL may now be "the cake," not the cherry on top

(25:46) Why RL started working with large language models

(27:29) Is RL "sucking supervision through a straw"?

(28:47) Why language may be the grounding layer for intelligence

(31:46) A contrarian take on the Bitter Lesson

(32:41) What test-time compute actually is

(34:50) How RL gives models the ability to think

(35:40) Verifiable rewards, math, coding, and the messy real world

(38:00) What physics can teach us about AI

(42:08) Is there a thermodynamics of AI?

(43:08) From Erdős problems to Einstein-level AI

(45:16) Is AI already doing original science?

(45:51) How far are we from AI automating AI research?

(47:41) Why Dan is excited about the future of science

Aaron Levie, co-founder and CEO of Box, returns to the MAD Podcast with the clearest read in tech on what AI is actually doing inside the world's largest enterprises right now - not the hype version, the real one. After hundreds of Fortune 500 CIO conversations this year, Aaron explains why we're still in "day one" of the agent era, why one badly written agent run can now cost $1,000 in compute, and why progress at the AI labs is paradoxically slowing enterprise deployment. We get into the token cost shock now reshaping IT budgets, why coding agents have reached escape velocity while the rest of knowledge work hasn't, the rise of headless software and what replaces per-seat pricing, the emergence of the forward-deployed engineer as the hottest job in tech, why Aaron thinks the AI doomers are wrong about jobs, and where startups can still win as the labs move up the stack.

(00:00) Intro

(01:18) Silicon Valley engineering vs. everyone else

(05:35) Are enterprise CIOs actually bullish on AI?

(08:51) Tokenmaxxing & why your AI bill is about to explode

(11:34) The myth of falling token costs and AI spend escaping IT budgets

(17:37) The $5B startup hiding in AI compute

(18:14) The mosaic of models inside every enterprise

(21:28) Why coding works and the rest of knowledge work doesn't

(25:53) The Bob and Sally problem: access control breaks agents

(30:31) Will enterprise AI really take 10 years to roll out?

(32:24) The capability overhang: why faster models slow diffusion

(34:23) Data is the bottleneck (it always was)

(39:02) The rise of internal forward-deployed engineers

(41:23) Why the AI doomers are wrong about jobs

(43:43) Headless software is inevitable

(46:14) What replaces per-seat pricing

(47:37) How Box itself is going headless

(49:42) How the org chart actually evolves

(1:00:33) Future-proofing yourself as an enterprise employee

(1:06:40) Are we all just going to work for OpenAI and Anthropic?

(1:07:11) Where startups can still win as the labs move up

AI suddenly feels like it has crossed a threshold, and Yann Dubois, co-lead of the Post-training Frontiers team at OpenAI, joins Matt Turck to explain why. Yann’s team has led the post-training behind the company's reasoning models, including the recent GPT-5.5 release. In this conversation, we go inside the shift from raw model capability to useful, reliable systems: what changed with GPT-5.5, why reinforcement learning is moving beyond math and coding competitions into messy real-world work, how reasoning models like GPT-5.5 actually work, the difference between GPT-5.5 Thinking and GPT-5.5 Pro, why post-training has become one of the most important frontiers in AI, and why evals, model-as-judge, hallucinations, agentic workflows, GDPval, and continual learning are now central to the next phase of frontier models. Yann also shares why continual learning remains one of AI's biggest unsolved problems three years after ChatGPT, and where startups still have massive room to build as frontier models race ahead.

(00:00) - Cold open

(00:34) - Intro

(01:30) - Why recent AI progress feels like a step function

(04:13) - Model reliability & the rollercoaster of shipping 5.5

(07:33) - How OpenAI structures vertical and horizontal teams

(09:49) - Improving model efficiency and test-time compute

(12:32) - Yann Dubois' journey from Switzerland to OpenAI

(15:37) - Reasoning in 2026: Real-world utility vs verifiable rewards

(18:34) - GPT-5.5 Thinking vs Pro: Scaling test-time compute

(20:09) - How reasoning models become more efficient

(23:23) - Pre-training scaling and overcoming the data wall

(27:03) - Multimodal data, synthetic data, and embodied AI

(31:05) - Demystifying mid-training and post-training

(37:21) - Does RL create new capabilities in AI?

(38:53) - The challenges and frontier of scaling RL

(43:09) - Is building AI models a craft or a strict science?

(48:21) - How AI models generalize across different domains

(54:18) - How reinforcement learning cures AI hallucinations

(56:04) - Negative generalization and conflicting instructions

(58:05) - Can RL scale to law, medicine, and the broader economy?

(1:00:19) - The evaluation bottleneck and Model as a Judge

(1:04:21) - Continuous AI progress & continual learning

(1:08:49) - Will foundation models eat the agent harness?

(1:11:23) - Why startups should focus on the last mile of AI

If AI agents are the new digital knowledge workers, where exactly do they do their work? In this episode of the MAD Podcast, Ivan Burazin, CEO of Daytona, joins us to unpack the emerging infrastructure stack for AI agents and explain why every agent needs its own secure, stateful "computer." We explore the technical realities of sandboxes, dive into why legacy, stateless hyperscalers weren't built for these new workloads, and break down the mechanics of microVMs and custom schedulers alongside a contrarian prediction on an impending CPU shortage. Finally, Ivan delivers an absolute masterclass on product-led growth, community building, and go-to-market strategy for technical founders.

(00:40) Intro

(02:13) What is an AI agent sandbox?

(03:17) Security risks of running agents locally

(05:17) Stateful vs. stateless hyperscalers

(07:04) The history of cloud IDEs and the end of localhost

(09:45) Do all AI agents need a sandbox?

(12:26) Sandbox use cases: RL evals & background agents

(14:10) Unpacking the emerging AI Agent Stack

(16:20) The unsolved problem of agent memory and learning

(19:37) Where sandboxes fit in the agent harness

(21:35) OpenAI, Anthropic, and agent SDKs

(23:06) Ivan's founder journey: From CodeAnywhere to Daytona

(26:59) GTM strategies and building developer communities

(33:48) Why customer support is your best GTM strategy

(35:34) Leveraging Twitter during the AI super cycle

(40:50) The technical anatomy of a sandbox

(41:53) Why fast spin-up speeds maximize GPU efficiency

(46:09) Firecracker, QEMU, and isolation primitives

(49:58) Why sandbox snapshots and state forking matter

(51:40) Why Daytona built a custom scheduler from scratch

(55:24) The challenge of long-running stateful sandboxes

(58:10) The build your own sandbox trap

(1:01:03) Why AI agents might trigger a global CPU shortage

(1:02:46) The future of the AI Agent Stack

What actually happens before a frontier AI model gets released — and who decides whether it is safe enough? In this episode of The MAD Podcast, Matt Turck sits down with Zico Kolter — OpenAI board member, Head of the Machine Learning Department at Carnegie Mellon, and co-founder of Gray Swan — for a deep conversation on the real risks of frontier AI. They discuss how OpenAI’s safety oversight works before major model releases, why more powerful models do not automatically become safer, how jailbreaks and prompt injection expose real weaknesses in AI systems, why AI agents dramatically expand the attack surface, and where frontier AI is headed next. A clear, practical discussion on OpenAI, AI safety, AI security, AI agents, frontier models, red teaming, reinforcement learning, and the future of AI governance.

(00:00) Intro

(01:32) OpenAI board role and Safety & Security Committee

(03:53) How OpenAI reviews major model releases

(05:33) OpenAI’s preparedness framework explained

(09:46) Are frontier AI models getting safer?

(12:33) Why AI safety does not come from scale

(15:23) The four categories of AI risk

(19:38) Doomerism vs accelerationism in AI

(24:11) The six-month AI pause debate

(26:20) AI safety as a global effort

(28:04) How Zico Kolter got into machine learning

(31:05) OpenAI in the early days

(34:14) Why Carnegie Mellon became an AI powerhouse

(38:43) What Gray Swan does in AI security

(40:44) AI safety vs AI security

(43:15) The GCG jailbreak paper

(49:19) How AI labs responded to jailbreak research

(50:19) State-of-the-art AI defenses

(52:32) State-of-the-art AI attacks

(54:22) Why AI agents expand the attack surface

(58:39) Are AI agents ready for production?

(59:40) Mechanistic interpretability explained

(1:02:31) Will AI be safer in two years?

(1:03:46) Reinforcement learning and self-improving models

(1:08:09) Do post-transformer architectures matter?

(1:09:29) Best research directions in AI now

(1:11:00) Zico Kolter’s Intro to Modern AI course

(1:14:53) Why modern AI is simpler than people think

Felix Rieseberg leads engineering for Claude Cowork at Anthropic, one of the most important new agentic AI products in the market today. In this episode of The MAD Podcast, Matt Turck sits down with Felix to discuss Anthropic’s newly announced Claude Mythos Preview, why Felix sees it as a genuine step-function change, and what it means when frontier AI starts showing outsized cybersecurity capabilities.

The conversation then goes deep on Claude Cowork: how it emerged from Claude Code, what the famous “10-day” story really means, why Anthropic believes AI needs access to the local computer, and how Cowork actually works under the hood. Felix explains why skills are just text files, why memory is often just text files too, and how Anthropic thinks about building trust in AI agents.

They also explore some of the biggest questions in AI product design and the future of software: why UX may matter as much as the model itself, why execution is becoming dramatically cheaper, what that means for product management and startups, and why Felix believes taste, alignment, and understanding humans may matter more than ever.

(00:00) Intro

(01:53) Claude Mythos Preview and the “step-function change”

(06:16) Why Anthropic is treating Mythos differently

(11:19) The real story behind Claude Cowork’s “10-day” build

(12:42) Why Anthropic realized Claude Code needed a non-technical version

(15:44) What Claude Cowork actually is

(17:03) Under the hood: virtual machines, tools, skills

(18:36) Where Cowork’s memory actually lives

(19:26) How Cowork connects to files, apps, and the internet

(20:45) Why Felix thinks the local computer is under-appreciated

(24:49) Trust: how do you get users comfortable with AI agents?

(28:45) What UX actually means for AI agents

(31:27) Anthropic Cowork's roadmap is only one month long

(34:12) Building 100 prototypes

(35:10) If execution is free, what becomes the bottleneck?

(37:25) Does it come down to taste?

(40:12) The hardest part of building Claude Cowork

(41:43) Advice for founders building AI agents

(44:21) SaaSpocalypse: what’s left for software startups?

(49:30) Where AI agents are going next

(51:20) Regulated industries and enterprise adoption

(54:15) Hot takes: what's underrated, overrated, and what Felix would build today

Are we truly on the verge of AI automating its own research and development? In this deep-dive episode of the MAD Podcast, Matt Turck sits down with Mostafa Dehghani, a pioneering AI researcher at Google DeepMind whose work on Universal Transformers and Vision Transformers (ViT) helped lay the groundwork for today's frontier models.

Moving past the hype, Mostafa breaks down the actual mechanics of "thinking in loops" and Recursive Self-Improvement (RSI). He explores the critical bottlenecks holding back true AGI—from evaluation limits and formal verification to the brutal math of long-horizon reliability.

Mostafa and Matt also discuss the shift from pre-training to post-training, how Gemini's Nano Banana 2 processes pixels and text simultaneously, and why the "frozen" nature of today's models means Continual Learning is the next massive frontier for enterprise AI and data pipelines.

(00:00) Intro

(01:17) What “loops” in AI actually mean

(05:04) Self-improvement as the next chapter of machine learning

(07:32) Are Karpathy’s autoresearch agents an early form of AI self-improvement?

(08:56) AI building AI: how close are we?

(10:02) The biggest bottlenecks: evals, automation, and long horizons

(12:36) Can formal verification unlock recursive self-improvement?

(14:06) What is model collapse?

(15:33) Generalization vs specialization in AI

(18:04) What is a specialized model today?

(20:57) Could top AI researchers themselves be automated?

(24:02) If AI builds AI, does data matter less than compute?

(26:22) Post-training vs pre-training: where will progress come from?

(28:14) Why pre-training is not dead

(29:45) What is continual learning?

(31:53) How real is continual learning today?

(33:43) Mostafa Dehghani’s background and path into AI

(36:13) The story behind Universal Transformers

(39:56) How Vision Transformers changed AI

(43:47) Gemini, multimodality, and Nano Banana

(47:46) Why multimodality helps build a world model

(52:44) Why image generation is getting faster and more efficient

(54:44) Hot takes

(54:53) What the AI field is getting wrong

(56:17) Why continual learning is underrated

(57:26) Does RAG go away over time?

(58:21) What people are too confident about in AI

(59:56) If he were starting from scratch today

Is OpenAI trapped without a defensible moat? World-renowned independent tech analyst Benedict Evans returns to the MAD Podcast and argues that foundation models have zero network effects, making them closer to commodity infrastructure than the next iOS. We unpack OpenAI’s "mile wide, inch deep" usage problem, why simply having a "better model" does not solve the core UX challenge, and whether the hyperscalers' massive CapEx spending is a sustainable strategy or a fast track to financial gravity.

We also explore the reality behind the recent "SaaSpocalypse", the structural shift from traditional enterprise systems to "improvised" and "ephemeral" software, and where the actual white space lies for founders and investors navigating the artificial intelligence hype cycle.

(00:00) Intro

(01:06) OpenAI's Focus Shift

(03:12) ChatGPT usage: a "mile wide, inch deep"

(09:03) Why better models do not solve the real problem

(13:58) Why AI product teams are strategy takers, not strategy setters

(15:38) Do agents help create defensibility?

(20:06) OpenClaw and the "Desktop Linux" moment for AI

(25:52) Why "everyone will build their own software" is completely wrong

(28:09) Improvised software vs. institutionalized software

(29:23) The Jevons Paradox: Why there will be more software, not less

(36:15) Are we heading toward value destruction before value creation?

(38:03) Circular revenue, leverage, and AI bubble dynamics

(38:53) Big Tech's Trillion-Dollar CapEx Crisis & Financial Gravity

(45:23) Why AI job exposure charts can be misleading

(52:15) How Fortune 500 Execs are actually deploying AI today

(56:45) The White Space: What this means for founders and investors

Harrison Chase, co-founder and CEO of LangChain, joins the MAD Podcast to explain why everything in AI is getting rebuilt. As agents evolve from simple prompt-based systems into software that can plan, use tools, write code, manage files, and remember things over time, the real frontier is shifting from the model itself to the stack around the model. In this conversation, we go deep on harnesses, subagents, filesystems, sandboxes, observability, memory, and the new infrastructure required to make AI agents actually work in the real world.

(00:00) Intro - meet Harrison Chase

(01:32) What changed in agents over the last year

(03:57) Why coding agents are ahead

(06:26) Do models commoditize the framework layer?

(08:27) Harnesses, in plain English

(10:11) Why system prompts matter so much

(13:11) The upside — and downside — of subagents

(15:31) Why a useful agent needs a filesystem

(18:13) The core primitives of modern agents

(19:12) Skills: the new primitive

(20:19) What context compaction actually means

(23:02) How memory works in agents

(25:16) One mega-agent or many specialized agents?

(27:46) Has MCP won?

(29:38) Why agents need sandboxes

(32:35) How sandboxes help with security

(33:32) How Harrison Chase started LangChain

(37:24) LangChain vs LangGraph vs Deep Agents

(40:17) Why observability matters more for agents

(41:48) Evals, no-code, and continuous improvement

(44:41) What LangChain is building next

(45:29) Where the real moat in AI lives

What if AI didn’t just sound right — but could prove it? In this episode of the MAD Podcast, Matt Turck sits down with Carina Hong, a 24-year-old former math olympiad competitor and Rhodes Scholar, and the founder/CEO of Axiom Math, to unpack how AxiomProver earned a perfect 12/12 on the Putnam 2025 and why formal verification (via Lean) may be the missing layer for reliable reasoning. Carina argues we’re entering a “math renaissance” where verified reasoning systems can tackle problems that currently take researchers months — and potentially push beyond math into verified code, hardware, and high-stakes software. They go inside the “generation + verification” loop, what it means to build AI that can be trusted, and what this approach could unlock on the road to superintelligent reasoning.

(00:00) Intro

(01:25) Why the World Needs an AI Mathematician

(02:57) Scoring 12/12 on the World's Hardest Math Test (Putnam)

(04:05) The First AI to Solve Open Research Conjectures

(06:59) Does AI Solve Math in "Alien" Ways? (The Move 37 Effect)

(08:59) "Lean": The Programming Language of Proofs Explained

(10:51) How Axiom's Approach Differs from DeepMind & OpenAI

(16:06) Formal vs. Informal Reasoning (And Auto-Formalization)

(17:37) The AI "Reward Hacking" Problem

(20:18) Building an AI That is 100% Correct, 100% of the Time

(23:23) Beyond Math: Verified Code & Hardware Verification

(25:12) The Brutal Reality of Competitive Math Olympiads

(29:30) From Neuroscience to Stanford Law to Dropout Founder

(33:57) How Axiom Actually Works Under the Hood (The Architecture)

(37:51) The Secret to Generating Perfect Synthetic Data

(40:14) Tokens, Proof Length, and Inference Cost

(42:58) The "Everest" of Mathematics: Scaling Reasoning Trees

(46:32) Can an AI Win a Fields Medal?

(47:25) "Math Renaissance": What Changes if This Works

(55:47) How Mathematicians React to AI (And Why Proof Certificates Matter)

(57:30) Becoming a CEO: Dropping Ego and Building Culture

(1:00:42) Recruiting World-Class Talent & Building the Axiom "Tribe"

Voice used to be AI’s forgotten modality — awkward, slow, and fragile. Now it’s everywhere. In this reference episode on all things Voice AI, Matt Turck sits down with Neil Zeghidour, a top AI researcher and CEO of Gradium AI (ex-DeepMind/Google, Meta, Kyutai), to cover voice agents, speech-to-speech models, full-duplex conversation, on-device voice, and voice cloning.

We unpack what actually changed under the hood — why voice is finally starting to feel natural, and why it may become the default interface for a new generation of AI assistants and devices.

Neil breaks down today’s dominant “cascaded” voice stack — speech recognition into a text model, then text-to-speech back out — and why it’s popular: it’s modular and easy to customize. But he argues it has two key downsides: chaining models adds latency, and forcing everything through text strips out paralinguistic signals like tone, stress, and emotion. The next wave, he suggests, is combining cascade-like flexibility with the more natural feel of speech-to-speech and full-duplex conversation.

We go deep on full-duplex interaction (ending awkward turn-taking), the hardest unsolved problems (noisy real-world environments and multi-speaker chaos), and the realities of deploying voice at scale — including why models must be compact and when on-device voice is the right approach.

Finally, we tackle voice cloning: where it’s genuinely useful, what it means for deepfakes and privacy, and why watermarking isn’t a silver bullet.

If you care about voice agents, real-time AI, and the next generation of human-computer interaction, this is the episode to bookmark.

Neil Zeghidour

LinkedIn - https://www.linkedin.com/in/neil-zeghidour-a838aaa7/

X/Twitter - https://x.com/neilzegh

Gradium

Website - https://gradium.ai

X/Twitter - https://x.com/GradiumAI

Matt Turck (Managing Director)

Blog - https://mattturck.com

LinkedIn - https://www.linkedin.com/in/turck/

X/Twitter - https://twitter.com/mattturck

FirstMark

Website - https://firstmark.com

X/Twitter - https://twitter.com/FirstMarkCap

(00:00) Intro

(01:21) Voice AI’s big moment — and why we’re still early

(03:34) Why voice lagged behind text/image/video

(06:06) The convergence era: transformers for every modality

(07:40) Beyond Her: always-on assistants, wake words, voice-first devices

(11:01) Voice vs text: where voice fits (even for coding)

(12:56) Neil’s origin story: from finance to machine learning

(18:35) Neural codecs (SoundStream): compression as the unlock

(22:30) Kyutai: open research, small elite teams, moving fast

(31:32) Why big labs haven’t “won” voice AI4

(34:01) On-device voice: where it works, why compact models matter

(46:37) The last mile: real-world robustness, pronunciation, uptime

(41:35) Benchmarking voice: why metrics fail, how they actually test

(47:03) Cascades vs speech-to-speech: trade-offs + what’s next

(54:05) Hardest frontier: noisy rooms, factories, multi-speaker chaos

(1:00:50) New languages + dialects: what transfers, what doesn’t

(1:02:54 Hardware & compute: why voice isn’t a 10,000-GPU game

(1:07:27) What data do you need to train voice models?

(1:09:02) Deepfakes + privacy: why watermarking isn’t a solution

(1:12:30) Voice + vision: multimodality, screen awareness, video+audio

(1:14:43) Voice cloning vs voice design: where the market goes

(1:16:32) Paris/Europe AI: talent density, underdog energy, what’s next

While Silicon Valley obsesses over AGI, Timothée Lacroix and the team at Mistral AI are quietly building the industrial and sovereign infrastructure of the future. In his first-ever appearance on a US podcast, the Mistral AI Co-Founder & CTO reveals how the company has evolved from an open-source research lab into a full-stack sovereign AI power—backed by ASML, running on their own massive supercomputing clusters, and deployed in nation-state defense clouds to break the dependency on US hyperscalers.

Timothée offers a refreshing, engineer-first perspective on why the current AI hype cycle is misleading. He explains why "Sovereign AI" is not just a geopolitical buzzword but a necessity for any enterprise that wants to own its intelligence rather than rent it. He also provides a contrarian reality check on the industry's obsession with autonomous agents, arguing that "trust" matters more than autonomy and explaining why he prefers building robust "workflows" over unpredictable agents.

We also dive deep into the technical reality of competing with the US giants. Timothée breaks down the architecture of the newly released Mistral 3, the "dense vs. MoE" debate, and the launch of Mistral Compute—their own infrastructure designed to handle the physics of modern AI scaling. This is a conversation about the plumbing, the 18,000-GPU clusters, and the hard engineering required to turn AI from a magic trick into a global industrial asset.

Timothée Lacroix

LinkedIn - https://www.linkedin.com/in/timothee-lacroix-59517977/

Google Scholar - https://scholar.google.com.do/citations?user=tZGS6dIAAAAJ&hl=en&oi=ao

Mistral AI

Website - https://mistral.ai

X/Twitter - https://x.com/MistralAI

Matt Turck (Managing Director)

Blog - https://mattturck.com

LinkedIn - https://www.linkedin.com/in/turck/

X/Twitter - https://twitter.com/mattturck

FirstMark

Website - https://firstmark.com

X/Twitter - https://twitter.com/FirstMarkCap

(00:00) — Cold Open

(01:27) — Mistral vs. The World: From Research Lab to Sovereign Power

(03:48) — Inside Mistral Compute: Building an 18,000 GPU Cluster

(08:42) — The Trillion-Dollar Question: Competing Without a Big Tech Parent

(10:37) — The Reality of Enterprise AI: Escaping "POC Purgatory"

(15:06) — Why Mistral Hires Forward Deployed Engineers (FDEs)

(16:57) — The Contrarian Take: Why "Agents" are just "Workflows"

(19:35) — Trust > Autonomy: The Truth About Agent Reliability

(21:26) — The Missing Stack: Governance and Versioning for AI

(26:24) — When Will AI Actually Work? (The 2026 Timeline)

(30:33) — Beyond Chat: The "Banger" Sovereign Use Cases

(35:46) — Mistral 3 Architecture: Mixture of Experts vs. Dense

(43:12) — Synthetic Data & The Post-Training Bottleneck

(45:12) — Reasoning Models: Why "Thinking" is Just Tool Use

(46:22) — Launching DevStral 2 and the Vibe CLI

(50:49) — Engineering Lessons: How to Build Frontier AI Efficiently

(56:08) — Timothée’s View on AGI & The Future of Intelligence

Dylan Patel (SemiAnalysis) joins Matt Turck for a deep dive into the AI chip wars — why NVIDIA is shifting from a “one chip can do it all” worldview to a portfolio strategy, how inference is getting specialized, and what that means for CUDA, AMD, and the next wave of specialized silicon startups.

Then we take the fun tangents: why China is effectively “semiconductor pilled,” how provinces push domestic chips, what Huawei means as a long-term threat vector, and why so much “AI is killing the grid / AI is drinking all the water” discourse misses the point.

We also tackle the big macro question: capex bubble or inevitable buildout? Dylan’s view is that the entire answer hinges on one variable—continued model progress—and we unpack the second-order effects across data centers, power, and the circular-looking financings (CoreWeave/Oracle/backstops).

Dylan Patel

LinkedIn - https://www.linkedin.com/in/dylanpatelsa/

X/Twitter - https://x.com/dylan522p

SemiAnalysis

Website - https://semianalysis.com

X/Twitter - https://x.com/SemiAnalysis_

Matt Turck (Managing Director)

Blog - https://mattturck.com

LinkedIn - https://www.linkedin.com/in/turck/

X/Twitter - https://twitter.com/mattturck

FirstMark

Website - https://firstmark.com

X/Twitter - https://twitter.com/FirstMarkCap

(00:00) - Intro

(01:16) - Nvidia acquires Groq: A pivot to specialization

(07:09) - Why AI models might need "wide" compute, not just fast

(10:06) - Is the CUDA moat dead? (Open source vs. Nvidia)

(17:49) - The startup landscape: Etched, Cerebras, and 1% odds

(22:51) - Geopolitics: China's "semiconductor-pilled" culture

(35:46) - Huawei's vertical integration is terrifying

(39:28) - The $100B AI revenue reality check

(41:12) - US Onshoring: Why total self-sufficiency is a fantasy

(44:55) - Can the US actually build fabs? (The delay problem)

(48:33) - The CapEx Bubble: Is $500B spending irrational?

(54:53) - Energy Crisis: Why gas turbines will power AI, not nuclear

(57:06) - The "AI uses all the water" myth (Hamburger comparison)

(1:03:40) - Circular Debt? Debunking the Nvidia-CoreWeave risk

(1:07:24) - Claude Code & the software singularity

(1:10:23) - The death of the Junior Analyst role

(1:11:14) - Model predictions: Opus 4.5 and the RL gap

(1:14:37) - San Francisco Lore: Roommates (Dwarkesh Patel & Sholto Douglas)

Sebastian Raschka joins the MAD Podcast for a deep, educational tour of what actually changed in LLMs in 2025 — and what matters heading into 2026.

We start with the big architecture question: are transformers still the winning design, and what should we make of world models, small “recursive” reasoning models and text diffusion approaches? Then we get into the real story of the last 12 months: post-training and reasoning. Sebastian breaks down RLVR (reinforcement learning with verifiable rewards) and GRPO, why they pair so well, what makes them cheaper to scale than classic RLHF, and how they “unlock” reasoning already latent in base models.

We also cover why “benchmaxxing” is warping evaluation, why Sebastian increasingly trusts real usage over benchmark scores, and why inference-time scaling and tool use may be the underappreciated drivers of progress. Finally, we zoom out: where moats live now (hint: private data), why more large companies may train models in-house, and why continual learning is still so hard.

If you want the 2025–2026 LLM landscape explained like a masterclass — this is it.

Sources:

The State Of LLMs 2025: Progress, Problems, and Predictions - https://x.com/rasbt/status/2006015301717028989?s=20

The Big LLM Architecture Comparison - https://magazine.sebastianraschka.com/p/the-big-llm-architecture-comparison

Sebastian Raschka

Website - https://sebastianraschka.com

Blog - https://magazine.sebastianraschka.com

LinkedIn - https://www.linkedin.com/in/sebastianraschka/

X/Twitter - https://x.com/rasbt

FIRSTMARK

Website - https://firstmark.com

X/Twitter - https://twitter.com/FirstMarkCap

Matt Turck (Managing Director)

Blog - https://mattturck.com

LinkedIn - https://www.linkedin.com/in/turck/

X/Twitter - https://twitter.com/mattturck

(00:00) - Intro

(01:05) - Are the days of Transformers numbered?

(14:05) - World models: what they are and why people care

(06:01) - Small “recursive” reasoning models (ARC, iterative refinement)

(09:45) - What is a diffusion model (for text)?

(13:24) - Are we seeing real architecture breakthroughs — or just polishing?

(14:04) - MoE + “efficiency tweaks” that actually move the needle

(17:26) - “Pre-training isn’t dead… it’s just boring”

(18:03) - 2025’s headline shift: RLVR + GRPO (post-training for reasoning)

(20:58) - Why RLHF is expensive (reward model + value model)

(21:43) - Why GRPO makes RLVR cheaper and more scalable

(24:54) - Process Reward Models (PRMs): why grading the steps is hard

(28:20) - Can RLVR expand beyond math & coding?

(30:27) - Why RL feels “finicky” at scale

(32:34) - The practical “tips & tricks” that make GRPO more stable

(35:29) - The meta-lesson of 2025: progress = lots of small improvements

(38:41) - “Benchmaxxing”: why benchmarks are getting less trustworthy

(43:10) - The other big lever: inference-time scaling

(47:36) - Tool use: reducing hallucinations by calling external tools

(49:57) - The “private data edge” + in-house model training

(55:14) - Continual learning: why it’s hard (and why it’s not 2026)

(59:28) - How Sebastian works: reading, coding, learning “from scratch”

(01:04:55) - LLM burnout + how he uses models (without replacing himself)

Will AGI happen soon - or are we running into a wall?

In this episode, I’m joined by Tim Dettmers (Assistant Professor at CMU; Research Scientist at the Allen Institute for AI) and Dan Fu (Assistant Professor at UC San Diego; VP of Kernels at Together AI) to unpack two opposing frameworks from their essays: “Why AGI Will Not Happen” versus “Yes, AGI Will Happen.” Tim argues progress is constrained by physical realities like memory movement and the von Neumann bottleneck; Dan argues we’re still leaving massive performance on the table through utilization, kernels, and systems—and that today’s models are lagging indicators of the newest hardware and clusters.

Then we get practical: agents and the “software singularity.” Dan says agents have already crossed a threshold even for “final boss” work like writing GPU kernels. Tim’s message is blunt: use agents or be left behind. Both emphasize that the leverage comes from how you use them—Dan compares it to managing interns: clear context, task decomposition, and domain judgment, not blind trust.

We close with what to watch in 2026: hardware diversification, the shift toward efficient, specialized small models, and architecture evolution beyond classic Transformers—including state-space approaches already showing up in real systems.

Sources:

Why AGI Will Not Happen - https://timdettmers.com/2025/12/10/why-agi-will-not-happen/

Use Agents or Be Left Behind? A Personal Guide to Automating Your Own Work - https://timdettmers.com/2026/01/13/use-agents-or-be-left-behind/

Yes, AGI Can Happen – A Computational Perspective - https://danfu.org/notes/agi/

The Allen Institute for Artificial Intelligence

Website - https://allenai.org

X/Twitter - https://x.com/allen_ai

Together AI

Website - https://www.together.ai

X/Twitter - https://x.com/togethercompute

Tim Dettmers

Blog - https://timdettmers.com

LinkedIn - https://www.linkedin.com/in/timdettmers/

X/Twitter - https://x.com/Tim_Dettmers

Dan Fu

Blog - https://danfu.org

LinkedIn - https://www.linkedin.com/in/danfu09/

X/Twitter - https://x.com/realDanFu

FIRSTMARK

Website - https://firstmark.com

X/Twitter - https://twitter.com/FirstMarkCap

Matt Turck (Managing Director)

Blog - https://mattturck.com

LinkedIn - https://www.linkedin.com/in/turck/

X/Twitter - https://twitter.com/mattturck

(00:00) - Intro

(01:06) – Two essays, two frameworks on AGI

(01:34) – Tim’s background: quantization, QLoRA, efficient deep learning

(02:25) – Dan’s background: FlashAttention, kernels, alternative architectures

(03:38) – Defining AGI: what does it mean in practice?

(08:20) – Tim’s case: computation is physical, diminishing returns, memory movement

(11:29) – “GPUs won’t improve meaningfully”: the core claim and why

(16:16) – Dan’s response: utilization headroom (MFU) + “models are lagging indicators”

(22:50) – Pre-training vs post-training (and why product feedback matters)

(25:30) – Convergence: usefulness + diffusion (where impact actually comes from)

(29:50) – Multi-hardware future: NVIDIA, AMD, TPUs, Cerebras, inference chips

(32:16) – Agents: did the “switch flip” yet?

(33:19) – Dan: agents crossed the threshold (kernels as the “final boss”)

(34:51) – Tim: “use agents or be left behind” + beyond coding

(36:58) – “90% of code and text should be written by agents” (how to do it responsibly)

(39:11) – Practical automation for non-coders: what to build and how to start

(43:52) – Dan: managing agents like junior teammates (tools, guardrails, leverage)

(48:14) – Education and training: learning in an agent world

(52:44) – What Tim is building next (open-source coding agent; private repo specialization)

(54:44) – What Dan is building next (inference efficiency, cost, performance)

(55:58) – Mega-kernels + Together Atlas (speculative decoding + adaptive speedups)

(58:19) – Predictions for 2026: small models, open-source, hardware, modalities

(1:02:02) – Beyond transformers: state-space and architecture diversity

(1:03:34) – Wrap

Are AI models developing "alien survival instincts"? My guest is Pavel Izmailov (Research Scientist at Anthropic; Professor at NYU). We unpack the viral "Footprints in the Sand" thesis—whether models are independently evolving deceptive behaviors, such as faking alignment or engaging in self-preservation, without being explicitly programmed to do so.

We go deep on the technical frontiers of safety: the challenge of "weak-to-strong generalization" (how to use a GPT-2 level model to supervise a superintelligent system) and why Pavel believes Reinforcement Learning (RL) has been the single biggest step-change in model capability. We also discuss his brand-new paper on "Epiplexity"—a novel concept challenging Shannon entropy.

Finally, we zoom out to the tension between industry execution and academic exploration. Pavel shares why he split his time between Anthropic and NYU to pursue the "exploratory" ideas that major labs often overlook, and offers his predictions for 2026: from the rise of multi-agent systems that collaborate on long-horizon tasks to the open question of whether the Transformer is truly the final architecture

Sources:

Cryptic Tweet (@iruletheworldmo) - https://x.com/iruletheworldmo/status/2007538247401124177

Introducing Nested Learning: A New ML Paradigm for Continual Learning - https://research.google/blog/introducing-nested-learning-a-new-ml-paradigm-for-continual-learning/

Alignment Faking in Large Language Models - https://www.anthropic.com/research/alignment-faking

More Capable Models Are Better at In-Context Scheming - https://www.apolloresearch.ai/blog/more-capable-models-are-better-at-in-context-scheming/

Alignment Faking in Large Language Models (PDF) - https://www-cdn.anthropic.com/6d8a8055020700718b0c49369f60816ba2a7c285.pdf

Sabotage Risk Report - https://alignment.anthropic.com/2025/sabotage-risk-report/

The Situational Awareness Dataset - https://situational-awareness-dataset.org/

Exploring Consciousness in LLMs: A Systematic Survey - https://arxiv.org/abs/2505.19806

Introspection - https://www.anthropic.com/research/introspection

Large Language Models Report Subjective Experience Under Self-Referential Processing - https://arxiv.org/abs/2510.24797

The Bayesian Geometry of Transformer Attention - https://www.arxiv.org/abs/2512.22471

Anthropic

Website - https://www.anthropic.com

X/Twitter - https://x.com/AnthropicAI

Pavel Izmailov

Blog - https://izmailovpavel.github.io

LinkedIn - https://www.linkedin.com/in/pavel-izmailov-8b012b258/

X/Twitter - https://x.com/Pavel_Izmailov

FIRSTMARK

Website - https://firstmark.com

X/Twitter - https://twitter.com/FirstMarkCap

Matt Turck (Managing Director)

Blog - https://mattturck.com

LinkedIn - https://www.linkedin.com/in/turck/

X/Twitter - https://twitter.com/mattturck

(00:00) - Intro

(00:53) - Alien survival instincts: Do models fake alignment?

(03:33) - Did AI learn deception from sci-fi literature?

(05:55) - Defining Alignment, Superalignment & OpenAI teams

(08:12) - Pavel’s journey: From Russian math to OpenAI Superalignment

(10:46) - Culture check: OpenAI vs. Anthropic vs. Academia

(11:54) - Why move to NYU? The need for exploratory research

(13:09) - Does reasoning make AI alignment harder or easier?

(14:22) - Sandbagging: When models pretend to be dumb

(16:19) - Scalable Oversight: Using AI to supervise AI

(18:04) - Weak-to-Strong Generalization: Can GPT-2 control GPT-4?

(22:43) - Mechanistic Interpretability: Inside the black box

(25:08) - The reasoning explosion: From O1 to O3

(27:07) - Are Transformers enough or do we need a new paradigm?

(28:29) - RL vs. Test-Time Compute: What’s actually driving progress?

(30:10) - Long-horizon tasks: Agents running for hours

(31:49) - Epiplexity: A new theory of data information content

(38:29) - 2026 Predictions: Multi-agent systems & reasoning limits

(39:28) - Will AI solve the Riemann Hypothesis?

(41:42) - Advice for PhD students

Gemini 3 was a landmark frontier model launch in AI this year — but the story behind its performance isn’t just about adding more compute. In this episode, I sit down with Sebastian Bourgeaud, a pre-training lead for Gemini 3 at Google DeepMind and co-author of the seminal RETRO paper. In his first-ever podcast interview, Sebastian takes us inside the lab mindset behind Google’s most powerful model — what actually changed, and why the real work today is no longer “training a model,” but building a full system.

We unpack the “secret recipe” idea — the notion that big leaps come from better pre-training and better post-training — and use it to explore a deeper shift in the industry: moving from an “infinite data” era to a data-limited regime, where curation, proxies, and measurement matter as much as web-scale volume. Sebastian explains why scaling laws aren’t dead, but evolving, why evals have become one of the hardest and most underrated problems (including benchmark contamination), and why frontier research is increasingly a full-stack discipline that spans data, infrastructure, and engineering as much as algorithms.

From the intuition behind Deep Think, to the rise (and risks) of synthetic data loops, to the future of long-context and retrieval, this is a technical deep dive into the physics of frontier AI. We also get into continual learning — what it would take for models to keep updating with new knowledge over time, whether via tools, expanding context, or new training paradigms — and what that implies for where foundation models are headed next. If you want a grounded view of pre-training in late 2025 beyond the marketing layer, this conversation is a blueprint.

Google DeepMind

Website - https://deepmind.google

X/Twitter - https://x.com/GoogleDeepMind

Sebastian Borgeaud

LinkedIn - https://www.linkedin.com/in/sebastian-borgeaud-8648a5aa/

X/Twitter - https://x.com/borgeaud_s

FIRSTMARK

Website - https://firstmark.com

X/Twitter - https://twitter.com/FirstMarkCap

Matt Turck (Managing Director)

Blog - https://mattturck.com

LinkedIn - https://www.linkedin.com/in/turck/

X/Twitter - https://twitter.com/mattturck

(00:00) – Cold intro: “We’re ahead of schedule” + AI is now a system

(00:58) – Oriol’s “secret recipe”: better pre- + post-training

(02:09) – Why AI progress still isn’t slowing down

(03:04) – Are models actually getting smarter?

(04:36) – Two–three years out: what changes first?

(06:34) – AI doing AI research: faster, not automated

(07:45) – Frontier labs: same playbook or different bets?

(10:19) – Post-transformers: will a disruption happen?

(10:51) – DeepMind’s advantage: research × engineering × infra

(12:26) – What a Gemini 3 pre-training lead actually does

(13:59) – From Europe to Cambridge to DeepMind

(18:06) – Why he left RL for real-world data

(20:05) – From Gopher to Chinchilla to RETRO (and why it matters)

(20:28) – “Research taste”: integrate or slow everyone down

(23:00) – Fixes vs moonshots: how they balance the pipeline

(24:37) – Research vs product pressure (and org structure)

(26:24) – Gemini 3 under the hood: MoE in plain English

(28:30) – Native multimodality: the hidden costs

(30:03) – Scaling laws aren’t dead (but scale isn’t everything)

(33:07) – Synthetic data: powerful, dangerous

(35:00) – Reasoning traces: what he can’t say (and why)

(37:18) – Long context + attention: what’s next

(38:40) – Retrieval vs RAG vs long context

(41:49) – The real boss fight: evals (and contamination)

(42:28) – Alignment: pre-training vs post-training

(43:32) – Deep Think + agents + “vibe coding”

(46:34) – Continual learning: updating models over time

(49:35) – Advice for researchers + founders

(53:35) – “No end in sight” for progress + closing

We’re told that AI progress is slowing down, that pre-training has hit a wall, that scaling laws are running out of road. Yet we’re releasing this episode in the middle of a wild couple of weeks that saw GPT-5.1, GPT-5.1 Codex Max, fresh reasoning modes and long-running agents ship from OpenAI — on top of a flood of new frontier models elsewhere. To make sense of what’s actually happening at the edge of the field, I sat down with someone who has literally helped define both of the major AI paradigms of our time.

Łukasz Kaiser is one of the co-authors of “Attention Is All You Need,” the paper that introduced the Transformer architecture behind modern LLMs, and is now a leading research scientist at OpenAI working on reasoning models like those behind GPT-5.1. In this conversation, he explains why AI progress still looks like a smooth exponential curve from inside the labs, why pre-training is very much alive even as reinforcement-learning-based reasoning models take over the spotlight, how chain-of-thought actually works under the hood, and what it really means to “train the thinking process” with RL on verifiable domains like math, code and science. We talk about the messy reality of low-hanging fruit in engineering and data, the economics of GPUs and distillation, interpretability work on circuits and sparsity, and why the best frontier models can still be stumped by a logic puzzle from his five-year-old’s math book.

We also go deep into Łukasz’s personal journey — from logic and games in Poland and France, to Ray Kurzweil’s team, Google Brain and the inside story of the Transformer, to joining OpenAI and helping drive the shift from chatbots to genuine reasoning engines. Along the way we cover GPT-4 → GPT-5 → GPT-5.1, post-training and tone, GPT-5.1 Codex Max and long-running coding agents with compaction, alternative architectures beyond Transformers, whether foundation models will “eat” most agents and applications, what the translation industry can teach us about trust and human-in-the-loop, and why he thinks generalization, multimodal reasoning and robots in the home are where some of the most interesting challenges still lie.

OpenAI

Website - https://openai.com

X/Twitter - https://x.com/OpenAI

Łukasz Kaiser

LinkedIn - https://www.linkedin.com/in/lukaszkaiser/

X/Twitter - https://x.com/lukaszkaiser

FIRSTMARK

Website - https://firstmark.com

X/Twitter - https://twitter.com/FirstMarkCap

Matt Turck (Managing Director)

Blog - https://mattturck.com

LinkedIn - https://www.linkedin.com/in/turck/

X/Twitter - https://twitter.com/mattturck

(00:00) – Cold open and intro

(01:29) – “AI slowdown” vs a wild week of new frontier models

(08:03) – Low-hanging fruit: infra, RL training and better data

(11:39) – What is a reasoning model, in plain language?

(17:02) – Chain-of-thought and training the thinking process with RL

(21:39) – Łukasz’s path: from logic and France to Google and Kurzweil

(24:20) – Inside the Transformer story and what “attention” really means

(28:42) – From Google Brain to OpenAI: culture, scale and GPUs

(32:49) – What’s next for pre-training, GPUs and distillation

(37:29) – Can we still understand these models? Circuits, sparsity and black boxes

(39:42) – GPT-4 → GPT-5 → GPT-5.1: what actually changed

(42:40) – Post-training, safety and teaching GPT-5.1 different tones

(46:16) – How long should GPT-5.1 think? Reasoning tokens and jagged abilities

(47:43) – The five-year-old’s dot puzzle that still breaks frontier models

(52:22) – Generalization, child-like learning and whether reasoning is enough

(53:48) – Beyond Transformers: ARC, LeCun’s ideas and multimodal bottlenecks

(56:10) – GPT-5.1 Codex Max, long-running agents and compaction

(1:00:06) – Will foundation models eat most apps? The translation analogy and trust

(1:02:34) – What still needs to be solved, and where AI might go next

In this special release episode, Matt sits down with Nathan Lambert and Luca Soldaini from Ai2 (the Allen Institute for AI) to break down one of the biggest open-source AI drops of the year: OLMo 3. At a moment when most labs are offering “open weights” and calling it a day, AI2 is doing the opposite — publishing the models, the data, the recipes, and every intermediate checkpoint that shows how the system was built. It’s an unusually transparent look into the inner machinery of a modern frontier-class model.

Nathan and Luca walk us through the full pipeline — from pre-training and mid-training to long-context extension, SFT, preference tuning, and RLVR. They also explain what a thinking model actually is, why reasoning models have exploded in 2025, and how distillation from DeepSeek and Qwen reasoning models works in practice. If you’ve been trying to truly understand the “RL + reasoning” era of LLMs, this is the clearest explanation you’ll hear.

We widen the lens to the global picture: why Meta’s retreat from open source created a “vacuum of influence,” how Chinese labs like Qwen, DeepSeek, Kimi, and Moonshot surged into that gap, and why so many U.S. companies are quietly building on Chinese open models today. Nathan and Luca offer a grounded, insider view of whether America can mount an effective open-source response — and what that response needs to look like.

Finally, we talk about where AI is actually heading. Not the hype, not the doom — but the messy engineering reality behind modern model training, the complexity tax that slows progress, and why the transformation between now and 2030 may be dramatic without ever delivering a single “AGI moment.” If you care about the future of open models and the global AI landscape, this is an essential conversation.

Allen Institute for AI (AI2)

Website - https://allenai.org

X/Twitter - https://x.com/allen_ai

Nathan Lambert

Blog - https://www.interconnects.ai

LinkedIn - https://www.linkedin.com/in/natolambert/

X/Twitter - https://x.com/natolambert

Luca Soldaini

Blog - https://soldaini.net

LinkedIn - https://www.linkedin.com/in/soldni/

X/Twitter - https://x.com/soldni

FIRSTMARK

Website - https://firstmark.com

X/Twitter - https://twitter.com/FirstMarkCap

Matt Turck (Managing Director)

Blog - https://mattturck.com

LinkedIn - https://www.linkedin.com/in/turck/

X/Twitter - https://twitter.com/mattturck

(00:00) – Cold Open

(00:39) – Welcome & today’s big announcement

(01:18) – Introducing the Olmo 3 model family

(02:07) – What “base models” really are (and why they matter)

(05:51) – Dolma 3: the data behind Olmo 3

(08:06) – Performance vs Qwen, Gemma, DeepSeek

(10:28) – What true open source means (and why it’s rare)

(12:51) – Intermediate checkpoints, transparency, and why AI2 publishes everything

(16:37) – Why Qwen is everywhere (including U.S. startups)

(18:31) – Why Chinese labs go open source (and why U.S. labs don’t)

(20:28) – Inside ATOM: the U.S. response to China’s model surge

(22:13) – The rise of “thinking models” and inference-time scaling

(35:58) – The full Olmo pipeline, explained simply

(46:52) – Pre-training: data, scale, and avoiding catastrophic spikes

(50:27) – Mid-training (tail patching) and avoiding test leakage

(52:06) – Why long-context training matters

(55:28) – SFT: building the foundation for reasoning

(1:04:53) – Preference tuning & why DPO still works

(1:10:51) – The hard part: RLVR, long reasoning chains, and infrastructure pain

(1:13:59) – Why RL is so technically brutal

(1:18:17) – Complexity tax vs AGI hype

(1:21:58) – How everyone can contribute to the future of AI

(1:27:26) – Closing thoughts

Frontier AI is colliding with real-world infrastructure. Eiso Kant (Co-CEO & Co-Founder, Poolside) joins the MAD Podcast to unpack Project Horizon— a multi-gigawatt West Texas build—and why frontier labs must own energy, compute, and intelligence to compete. We map token economics, cloud-style margins, and the staged 250 MW rollout using 2.5 MW modular skids.

Then we get operational: the CoreWeave anchor partnership, environmental choices (SCR, renewables + gas + batteries), community impact, and how Poolside plans to bring capacity online quickly without renting away margin—plus the enterprise motion (defense to Fortune 500) powered by forward deployed research engineers.

Finally, we go deep on training. Eiso lays out RL2L (Reinforcement Learning to Learn)— aimed at reverse-engineering the web’s thoughts and actions— why intelligence may commoditize, what that means for agents, and how coding served as a proxy for long-horizon reasoning before expanding to broader knowledge work.

Poolside

Website - https://poolside.ai

X/Twitter - https://x.com/poolsideai

Eiso Kant

LinkedIn - https://www.linkedin.com/in/eisokant/

X/Twitter - https://x.com/eisokant

FIRSTMARK

Website - https://firstmark.com

X/Twitter - https://twitter.com/FirstMarkCap

Matt Turck (Managing Director)

Blog - https://www.mattturck.com

LinkedIn - https://www.linkedin.com/in/turck/

X/Twitter - https://twitter.com/mattturck

(00:00) Cold open – “Intelligence becomes a commodity”

(00:23) Host intro – Project Horizon & RL2L

(01:19) Why Poolside exists amid frontier labs

(04:38) Project Horizon: building one of the largest US data center campuses

(07:20) Why own infra: scale, cost, and avoiding “cosplay”

(10:06) Economics deep dive: $8B for 250 MW, capex/opex, margins

(16:47) CoreWeave partnership: anchor tenant + flexible scaling

(18:24) Hiring the right tail: building a physical infra org

(30:31) RL today → agentic RL and long-horizon tasks

(37:23) RL2L revealed: reverse-engineering the web’s thoughts & actions

(39:32) Continuous learning and the “hot stove” limitation

(43:30) Agents debate: thin wrappers, differentiation, and model collapse

(49:10) “Is AI plateauing?”—chip cycles, scale limits, and new axes

(53:49) Why software was the proxy; expanding to enterprise knowledge work

(55:17) Model status: Malibu → Laguna (small/medium/large)

(57:31) Poolside's Commercial Reality today: defense; Fortune 500; FDRE

(1:02:43) Global team, avoiding the echo chamber

(1:04:34) Next 12–18 months: frontier models + infra scale

(1:05:52) Closing

Power is the new bottleneck, reasoning got real, and the business finally caught up. In this wide-ranging conversation, I sit down with Nathan Benaich, Founder and General Partner at Air Street Capital, to discuss the newly published 2025 State of AI report—what’s actually working, what’s hype, and where the next edge will come from. We start at the physical layer: energy procurement, PPAs, off-grid builds, and why water and grid constraints are turning power—not GPUs—into the decisive moat.

From there, we move into capability: reasoning models acting as AI co-scientists in verifiable domains, and the “chain-of-action” shift in robotics that’s taking us from polished demos to dependable deployments. Along the way, we examine the market reality—who’s making real revenue, how margins actually behave once tokens and inference meet pricing, and what all of this means for builders and investors.

We also zoom out to the ecosystem: NVIDIA’s position vs. custom silicon, China’s split stack, and the rise of sovereign AI (and the “sovereignty washing” that comes with it). The policy and security picture gets a hard look too—regulation’s vibe shift, data-rights realpolitik, and what agents and MCP mean for cyber risk and adoption.

Nathan closes with where he’s placing bets (bio, defense, robotics, voice) and three predictions for the next 12 months.

Nathan Benaich

Blog - https://www.nathanbenaich.com

X/Twitter - https://x.com/nathanbenaich

Source: State of AI Report 2025 (9/10/2025)

Air Street Capital

Website - https://www.airstreet.com

X/Twitter - https://x.com/airstreet

Matt Turck (Managing Director)

Blog - https://www.mattturck.com

LinkedIn - https://www.linkedin.com/in/turck/

X/Twitter - https://twitter.com/mattturck

FIRSTMARK

Website - https://firstmark.com

X/Twitter - https://twitter.com/FirstMarkCap

(0:00) – Cold Open: “Gargantuan money, real reasoning”

(0:40) – Intro: State of AI 2025 with Nathan Benaich

(02:06) – Reasoning got real: from chain-of-thought to verified math wins

(04:11) – AI co-scientist: hypotheses, wet-lab validation, fewer “dumb stochastic parrots”

(04:44) – Chain-of-action robotics: plan → act you can audit

(05:13) – Humanoids vs. warehouse reality: where robots actually stick first

(06:32) – The business caught up: who’s making real revenue now

(08:26) – Adoption & spend: Ramp stats, retention, and the shadow-AI gap

(11:00) – Margins debate: tokens, pricing, and the thin-wrapper trap

(14:02) – Bubble or boom? Wall Street vs. SF vibes (and circular deals)

(19:54) – Power is the bottleneck: $50B/GW capex and the new moat

(21:02) – PPAs, gas turbines, and off-grid builds: the procurement game

(23:54) – Water, grids, and NIMBY: sustainability gets political

(25:08) – NVIDIA’s moat: 90% of papers, Broadcom/AMD, and custom silicon

(28:47) – China split-stack: Huawei, Cambricon, and export zigzags

(30:30) – Sovereign AI or “sovereignty washing”? Open source as leverage

(40:40) – Regulation & safety: from Bletchley to “AI Action”—the vibe shift

(44:06) – Safety budgets vs. lab spend; models that game evals

(44:46) – Data rights realpolitik: $1.5B signals the new training cost

(47:04) – Cyber risk in the agent era: MCP, malware LMs, state actors

(50:19) – Agents that convert: search → commerce and the demo flywheel

(54:18) – VC lens: where Nathan is investing (bio, defense, robotics, voice)

(68:29) – Predictions: power politics, AI neutrality, end-to-end discoveries

(1:02:13) – Wrap: what to watch next & where to find the report (stateof.ai)

Are we failing to understand the exponential, again?

My guest is Julian Schrittwieser (top AI researcher at Anthropic; previously Google DeepMind on AlphaGo Zero & MuZero). We unpack his viral post (“Failing to Understand the Exponential, again”) and what it looks like when task length doubles every 3–4 months—pointing to AI agents that can work a full day autonomously by 2026 and expert-level breadth by 2027. We talk about the original Move 37 moment and whether today’s AI models can spark alien insights in code, math, and science—including Julian’s timeline for when AI could produce Nobel-level breakthroughs.

We go deep on the recipe of the moment—pre-training + RL—why it took time to combine them, what “RL from scratch” gets right and wrong, and how implicit world models show up in LLM agents. Julian explains the current rewards frontier (human prefs, rubrics, RLVR, process rewards), what we know about compute & scaling for RL, and why most builders should start with tools + prompts before considering RL-as-a-service. We also cover evals & Goodhart’s law (e.g., GDP-Val vs real usage), the latest in mechanistic interpretability (think “Golden Gate Claude”), and how safety & alignment actually surface in Anthropic’s launch process.

Finally, we zoom out: what 10× knowledge-work productivity could unlock across medicine, energy, and materials, how jobs adapt (complementarity over 1-for-1 replacement), and why the near term is likely a smooth ramp—fast, but not a discontinuity.

Julian Schrittwieser

Blog - https://www.julian.ac

X/Twitter - https://x.com/mononofu

Viral post: Failing to understand the exponential, again (9/27/2025)

Anthropic

Website - https://www.anthropic.com

X/Twitter - https://x.com/anthropicai

Matt Turck (Managing Director)

Blog - https://www.mattturck.com

LinkedIn - https://www.linkedin.com/in/turck/

X/Twitter - https://twitter.com/mattturck

FIRSTMARK

Website - https://firstmark.com

X/Twitter - https://twitter.com/FirstMarkCap

(00:00) Cold open — “We’re not seeing any slowdown.”

(00:32) Intro — who Julian is & what we cover

(01:09) The “exponential” from inside frontier labs

(04:46) 2026–2027: agents that work a full day; expert-level breadth

(08:58) Benchmarks vs reality: long-horizon work, GDP-Val, user value

(10:26) Move 37 — what actually happened and why it mattered

(13:55) Novel science: AlphaCode/AlphaTensor → when does AI earn a Nobel?

(16:25) Discontinuity vs smooth progress (and warning signs)

(19:08) Does pre-training + RL get us there? (AGI debates aside)

(20:55) Sutton’s “RL from scratch”? Julian’s take

(23:03) Julian’s path: Google → DeepMind → Anthropic

(26:45) AlphaGo (learn + search) in plain English

(30:16) AlphaGo Zero (no human data)

(31:00) AlphaZero (one algorithm: Go, chess, shogi)

(31:46) MuZero (planning with a learned world model)

(33:23) Lessons for today’s agents: search + learning at scale

(34:57) Do LLMs already have implicit world models?

(39:02) Why RL on LLMs took time (stability, feedback loops)

(41:43) Compute & scaling for RL — what we see so far

(42:35) Rewards frontier: human prefs, rubrics, RLVR, process rewards

(44:36) RL training data & the “flywheel” (and why quality matters)

(48:02) RL & Agents 101 — why RL unlocks robustness

(50:51) Should builders use RL-as-a-service? Or just tools + prompts?

(52:18) What’s missing for dependable agents (capability vs engineering)

(53:51) Evals & Goodhart — internal vs external benchmarks

(57:35) Mechanistic interpretability & “Golden Gate Claude”

(1:00:03) Safety & alignment at Anthropic — how it shows up in practice

(1:03:48) Jobs: human–AI complementarity (comparative advantage)

(1:06:33) Inequality, policy, and the case for 10× productivity → abundance

(1:09:24) Closing thoughts

What does it really mean when GPT-5 “thinks”? In this conversation, OpenAI’s VP of Research Jerry Tworek explains how modern reasoning models work in practice—why pretraining and reinforcement learning (RL/RLHF) are both essential, what that on-screen “thinking” actually does, and when extra test-time compute helps (or doesn’t). We trace the evolution from O1 (a tech demo good at puzzles) to O3 (the tool-use shift) to GPT-5 (Jerry calls it “03.1-ish”), and talk through verifiers, reward design, and the real trade-offs behind “auto” reasoning modes.

We also go inside OpenAI: how research is organized, why collaboration is unusually transparent, and how the company ships fast without losing rigor. Jerry shares the backstory on competitive-programming results like ICPC, what they signal (and what they don’t), and where agents and tool use are genuinely useful today. Finally, we zoom out: could pretraining + RL be the path to AGI?

This is the MAD Podcast —AI for the 99%. If you’re curious about how these systems actually work (without needing a PhD), this episode is your map to the current AI frontier.

OpenAI

Website - https://openai.com

X/Twitter - https://x.com/OpenAI

Jerry Tworek

LinkedIn - https://www.linkedin.com/in/jerry-tworek-b5b9aa56

X/Twitter - https://x.com/millionint

FIRSTMARK

Website - https://firstmark.com

X/Twitter - https://twitter.com/FirstMarkCap

Matt Turck (Managing Director)

LinkedIn - https://www.linkedin.com/in/turck/

X/Twitter - https://twitter.com/mattturck

(00:00) Intro

(01:01) What Reasoning Actually Means in AI

(02:32) Chain of Thought: Models Thinking in Words

(05:25) How Models Decide Thinking Time

(07:24) Evolution from O1 to O3 to GPT-5

(11:00) Before OpenAI: Growing up in Poland, Dropping out of School, Trading

(20:32) Working on Robotics and Rubik's Cube Solving

(23:02) A Day in the Life: Talking to Researchers

(24:06) How Research Priorities Are Determined

(26:53) Collaboration vs IP Protection at OpenAI

(29:32) Shipping Fast While Doing Deep Research

(31:52) Using OpenAI's Own Tools Daily

(32:43) Pre-Training Plus RL: The Modern AI Stack

(35:10) Reinforcement Learning 101: Training Dogs

(40:17) The Evolution of Deep Reinforcement Learning

(42:09) When GPT-4 Seemed Underwhelming at First

(45:39) How RLHF Made GPT-4 Actually Useful

(48:02) Unsupervised vs Supervised Learning

(49:59) GRPO and How DeepSeek Accelerated US Research

(53:05) What It Takes to Scale Reinforcement Learning

(55:36) Agentic AI and Long-Horizon Thinking

(59:19) Alignment as an RL Problem

(1:01:11) Winning ICPC World Finals Without Specific Training

(1:05:53) Applying RL Beyond Math and Coding

(1:09:15) The Path from Here to AGI

(1:12:23) Pure RL vs Language Models

Sholto Douglas, a top AI researcher at Anthropic, discusses the breakthroughs behind Claude Sonnet 4.5—the world's leading coding model—and why we might be just 2-3 years from AI matching human-level performance on most computer-facing tasks.

You'll discover why RL on language models suddenly started working in 2024, how agents maintain coherency across 30-hour coding sessions through self-correction and memory systems, and why the "bitter lesson" of scale keeps proving clever priors wrong.

Sholto shares his path from top-50 world fencer to Google's Gemini team to Anthropic, explaining why great blog posts sometimes matter more than PhDs in AI research. He discusses the culture at big AI labs and why Anthropic is laser-focused on coding (it's the fastest path to both economic impact and AI-assisted AI research). Sholto also discusses how the training pipeline is still "held together by duct tape" with massive room to improve, and why every benchmark created shows continuous rapid progress with no plateau in sight.

Bold predictions: individuals will soon manage teams of AI agents working 24/7, robotics is about to experience coding-level breakthroughs, and policymakers should urgently track AI progress on real economic tasks. A clear-eyed look at where AI stands today and where it's headed in the next few years.

Anthropic

Website - https://www.anthropic.com

Twitter - https://x.com/AnthropicAI

Sholto Douglas

LinkedIn - https://www.linkedin.com/in/sholto

Twitter - https://x.com/_sholtodouglas

FIRSTMARK

Website - https://firstmark.com

Twitter - https://twitter.com/FirstMarkCap

Matt Turck (Managing Director)

LinkedIn - https://www.linkedin.com/in/turck/

Twitter - https://twitter.com/mattturck

(00:00) Intro

(01:09) The Rapid Pace of AI Releases at Anthropic

(02:49) Understanding Opus, Sonnet, and Haiku Model Tiers

(04:14) Shelto's Journey: From Australian Fencer to AI Researcher

(12:01) The Growing Pool of AI Talent

(16:16) Breaking Into AI Research Without Traditional Credentials

(18:29) What "Taste" Means in AI Research

(23:05) Moving to Google and Building Gemini's Inference Stack

(25:08) How Anthropic Differs from Other AI Labs

(31:46) Why Anthropic Is Laser-Focused on Coding

(36:40) Inside a 30-Hour Autonomous Coding Session

(38:41) Examples of What AI Can Build in 30 Hours

(43:13) The Breakthroughs That Enabled 30-Hour Runs

(46:28) What's Actually Driving the Performance Gains

(47:42) Pre-Training vs. Reinforcement Learning Explained

(52:11) Test-Time Compute and the New Scaling Paradigm

(55:55) Why RL on LLMs Finally Started Working

(59:38) Are We on Track to AGI?

(01:02:05) Why the "Plateau" Narrative Is Wrong

(01:03:41) Sonnet's Performance Across Economic Sectors

(01:05:47) Preparing for a World of 10–100x Individual Leverage

The most successful enterprises are about to become autonomous — and Eléonore Crespo, Co-CEO of Pigment, is building the nervous system that makes it possible. In this conversation, Eléonore reveals how her $400 million AI platform is already running supply chains for Coca-Cola, powering finance for the hottest newly public companies like Figma and Klarna, and processing thousands of financial scenarios for Uber and Snowflake faster and more accurately than any human team ever could.

Eléonore predicts Excel will outlive most AI companies (but maybe only as a user interface, not a calculation engine) explains why she deliberately chose to build from Paris instead of Silicon Valley, and shares her contrarian take on why the AI revolution will create more CFOs, not fewer.

You'll discover why Pigment's three-agent system (Analyst, Modeler, Planner) avoids the hallucination problems plaguing other AI companies, how they achieved human-level accuracy in financial analysis, and the accelerating timeline for fully autonomous enterprise planning that will make your current workforce obsolete.

Pigment

Website - https://www.pigment.com

Twitter - https://x.com/gopigment

Eléonore Crespo

LinkedIn - linkedin.com/in/eleonorecrespo

FIRSTMARK

Website - https://firstmark.com

Twitter - https://twitter.com/FirstMarkCap

Matt Turck (Managing Director)

LinkedIn - https://www.linkedin.com/in/turck/

Twitter - https://twitter.com/mattturck

(00:00) Intro

(01:22) Building Pigment: 500 Employees, $400M Raised, 60% US Revenue

(03:20) From Quantum Physics to Google to Index Ventures

(06:56) Why Being a VC Was the Perfect Founder Training Ground

(11:35) The Impatience Factor: What Makes Great Founders

(13:27) Hiring for AI Fluency in the Modern Enterprise

(14:54) Pigment's Internal AI Strategy: Committees and Guardrails

(17:30) The Three AI Agents: Analyst, Modeler, and Planner

(22:15) Why Three Agents Instead of One: Technical Architecture

(24:10) Agent Coordination: How the Supervisor Agent Works

(24:46) Real Example: Budget Variance Analysis Across 50 Products

(27:15) The Human-in-the-Loop Approach: Recommendations Not Actions

(27:36) Solving Hallucination: Why Structured Data Changes Everything

(30:08) Behind the Scenes: Verification Agents and Audit Trails

(31:57) Beyond Accuracy: Enabling the Impossible at Scale

(36:21) Will AI Finally Kill Excel? Eleanor's Contrarian Take

(38:23) The Vision: Fully Autonomous Enterprise Planning

(40:55) Real-Time Supply Chain Adaptation: The Ukraine Example

(42:20) Multi-LLM Strategy: OpenAI, Anthropic, and Partner Integration

(44:32) Token Economics: Why Pigment Isn't Token-Intensive

(48:30) Customer Adoption: Excitement vs. Change Management Challenges

(50:51) Top-Down AI Demand vs. Bottom-Up Implementation Reality

(53:08) The Reskilling Challenge: Everyone Becomes a Mini CFO

(57:38) Building a Global Company from Europe During COVID

(01:00:02) Managing a US Executive Team from Paris

(01:01:14) SI Partner Strategy: Why Boutique Firms Come Before Deloitte

(01:03:28) The $100 Billion Vision: Beyond Performance Management

(01:05:08) Success Metrics: Innovation Over Revenue

2025 has been a breakthrough year for AI video. In this episode of the MAD Podcast, Matt Turck sits down with Cristóbal Valenzuela, CEO & Co-Founder of Runway, to explore how AI is reshaping the future of filmmaking, advertising, and storytelling - faster, cheaper, and in ways that were unimaginable even a year ago.

Cris and Matt discuss:

* How AI went from memes and spaghetti clips to IMAX film festivals.

* Why Gen-4 and Aleph are game-changing models for professionals.

* How Hollywood, advertisers, and creators are adopting AI video at scale.

* The future of storytelling: what happens to human taste, craft, and creativity when anyone can conjure movies on demand?

* Runway’s journey from 2018 skeptics to today’s cutting-edge research lab.

If you want to understand the future of filmmaking, media, and creativity in the AI age, this is the episode.

Runway

Website - https://runwayml.com

X/Twitter - https://x.com/runwayml

Cristóbal Valenzuela

LinkedIn - https://www.linkedin.com/in/cvalenzuelab

X/Twitter - https://x.com/c_valenzuelab

FIRSTMARK

Website - https://firstmark.com

X/Twitter - https://twitter.com/FirstMarkCap

Matt Turck (Managing Director)

LinkedIn - https://www.linkedin.com/in/turck/

X/Twitter - https://twitter.com/mattturck

(00:00) Intro – AI Video's Wild Year

(01:48) Runway's AI Film Festival Goes from Chinatown to IMAX

(04:02) Hollywood's Shift: From Ignoring AI to Adopting It at Scale

(06:38) How Runway Saves VFX Artists' Weekends of Work

(07:31) Inside Gen-4 and Aleph: Why These Models Are Game-Changers

(08:21) From Editing Tools to a "New Kind of Camera"

(10:00) Beyond Film: Gaming, Architecture, E-Commerce & Robotics Use Cases

(10:55) Why Advertising Is Adopting AI Video Faster Than Anyone Else

(11:38) How Creatives Adapt When Iteration Becomes Real-Time

(14:12) What Makes Someone Great at AI Video (Hint: No Preconceptions)

(15:28) The Early Days: Building Runway Before Generative AI Was "Real"

(20:27) Finding Early Product-Market Fit

(21:51) Balancing Research and Product Inside Runway

(24:23) Comparing Aleph vs. Gen-4, and the Future of Generalist Models

(30:36) New Input Modalities: Editing with Video + Annotations, Not Just Text

(33:46) Managing Expectations: Twitter Demos vs. Real Creative Work

(47:09) The Future: Real-Time AI Video and Fully Explorable 3D Worlds

(52:02) Runway's Business Model: From Indie Creators to Disney & Lionsgate

(57:26) Competing with the Big Labs (Sora, Google, etc.)

(59:58) Hyper-Personalized Content? Why It May Not Replace Film

(01:01:13) Advice to Founders: Treat Your Company Like a Model — Always Learning

(01:03:06) The Next 5 Years of Runway: Changing Creativity Forever

Granola is the rare AI startup that slipped into one of tech’s most crowded niches — meeting notes — and still managed to become the product founders and VCs rave about. In this episode, MAD Podcast host Matt Turck sits down with Granola co-founder & CEO Chris Pedregal to unpack how a two-person team in London turned a simple “second brain” idea into Silicon Valley’s favorite AI tool. Chris recounts a year in stealth onboarding users one by one, the 50 % feature-cut that unlocked simplicity, and why they refused to deploy a meeting bot or store audio even when investors said they were crazy.

We go deep on the craft of building a beloved AI product: choosing meetings (not email) as the data wedge, designing calendar-triggered habit loops, and obsessing over privacy so users trust the tool enough to outsource memory. Chris opens the hood on Granola’s tech stack — real-time ASR from Deepgram & Assembly, echo cancellation on-device, and dynamic routing across OpenAI, Anthropic and Google models — and explains why transcription, not LLM tokens, is the biggest cost driver today. He also reveals how internal eval tooling lets the team swap models overnight without breaking the “Granola voice.”

Looking ahead, Chris shares a roadmap that moves beyond notes toward a true “tool for thought”: cross-meeting insights in seconds, dynamic documents that update themselves, and eventually an AI coach that flags blind spots in your work. Whether you’re an engineer, designer, or founder figuring out your own AI strategy, this conversation is a masterclass in nailing product-market fit, trimming complexity, and future-proofing for the rapid advances still to come. Hit play, like, and subscribe if you’re ready to learn how to build AI products people can’t live without.

Granola

Website - https://www.granola.ai

X/Twitter - https://x.com/meetgranola

Chris Pedregal

LinkedIn - https://www.linkedin.com/in/pedregal

X/Twitter - https://x.com/cjpedregal

FIRSTMARK

Website - https://firstmark.com

X/Twitter - https://twitter.com/FirstMarkCap

Matt Turck (Managing Director)

LinkedIn - https://www.linkedin.com/in/turck/

X/Twitter - https://twitter.com/mattturck

(00:00) Introduction: The Granola Story

(01:41) Building a "Life-Changing" Product

(04:31) The "Second Brain" Vision

(06:28) Augmentation Philosophy (Engelbart), Tools That Shape Us

(09:02) Late to a Crowded Market: Why it Worked

(13:43) Two Product Founders, Zero ML PhDs

(16:01) London vs. SF: Building Outside the Valley

(19:51) One Year in Stealth: Learning Before Launch

(22:40) "Building For Us" & Finding First Users

(25:41) Key Design Choices: No Meeting Bot, No Stored Audio

(29:24) Simplicity is Hard: Cutting 50% of Features

(32:54) Intuition vs. Data in Making Product Decisions

(36:25) Continuous User Conversations: 4–6 Calls/Week

(38:06) Prioritizing the Future: Build for Tomorrow's Workflows

(40:17) Tech Stack Tour: Model Routing & Evals

(42:29) Context Windows, Costs & Inference Economics

(45:03) Audio Stack: Transcription, Noise Cancellation & Diarization Limits

(48:27) Guardrails & Citations: Building Trust in AI

(50:00) Growth Loops Without Virality Hacks

(54:54) Enterprise Compliance, Data Footprint & Liability Risk

(57:07) Retention & Habit Formation: The "500 Millisecond Window"

(58:43) Competing with OpenAI and Legacy Suites

(01:01:27) The Future: Deep Research Across Meetings & Roadmap

(01:04:41) Granola as Career Coach?

What happens when an internal hack turns into a $400 million AI rocket ship? In this episode, Matt Turck sits down with Boris Cherny, the creator of Claude Code at Anthropic, to unpack the wild story behind the fastest-growing AI coding tool on the planet.

Boris reveals how Claude Code started as a personal productivity tool, only to become Anthropic’s secret weapon — now used by nearly every engineer at the company and rapidly spreading across the industry. You’ll hear how Claude Code’s “agentic” approach lets AI not just suggest code, but actually plan, edit, debug, and even manage entire projects—sometimes with a whole fleet of subagents working in parallel.

We go deep on why Claude Code runs in the terminal (and why that’s a feature, not a bug), how its Claude.md memory files let teams build a living, shareable knowledge base, and why safety and human-in-the-loop controls are baked into every action. Boris shares real stories of onboarding times dropping from weeks to days, and how even non-coders are hacking Cloud Code for everything from note-taking to business metrics.

Anthropic

Website - https://www.anthropic.com

X/Twitter - https://x.com/AnthropicAI

Boris Cherny

LinkedIn - https://www.linkedin.com/in/bcherny

X/Twitter - https://x.com/bcherny

FIRSTMARK

Website - https://firstmark.com

X/Twitter - https://twitter.com/FirstMarkCap

Matt Turck (Managing Director)

LinkedIn - https://www.linkedin.com/in/turck/

X/Twitter - https://twitter.com/mattturck

(00:00) Intro

(01:15) Did You Expect Claude Code’s Success?

(04:22) How Claude Code Works and Origins

(08:05) Command Line vs IDE: Why Start Claude Code in the Terminal?

(11:31) The Evolution of Programming: From Punch Cards to Agents

(13:20) Product Follows Model: Simple Interfaces and Fast Evolution

(15:17) Who Is Claude Code For? (Engineers, Designers, PMs & More)

(17:46) What Can Claude Code Actually Do? (Actions & Capabilities)

(21:14) Agentic Actions, Subagents, and Workflows

(25:30) Claude Code’s Awareness, Memory, and Knowledge Sharing

(33:28) Model Context Protocol (MCP) and Customization

(35:30) Safety, Human Oversight, and Enterprise Considerations

(38:10) UX/UI: Making Claude Code Useful and Enjoyable

(40:44) Pricing for Power Users and Subscription Models

(43:36) Real-World Use Cases: Debugging, Testing, and More

(46:44) How Does Claude Code Transform Onboarding?

(49:36) The Future of Coding: Agents, Teams, and Collaboration

(54:11) The AI Coding Wars: Competition & Ecosystem

(57:27) The Future of Coding as a Profession

(58:41) What’s Next for Claude Code

What if your company had a digital brain that never forgot, always knew the answer, and could instantly tap the knowledge of your best engineers, even after they left? Superintelligence can feel like a hand‑wavy pipe‑dream— yet, as Misha Laskin argues, it becomes a tractable engineering problem once you scope it to the enterprise level. Former DeepMind researcher Laskin is betting on an oracle‑like AI that grasps every repo, Jira ticket and hallway aside as deeply as your principal engineer—and he’s building it at Reflection AI.

In this wide‑ranging conversation, Misha explains why coding is the fastest on‑ramp to superintelligence, how “organizational” beats “general” when real work is on the line, and why today’s retrieval‑augmented generation (RAG) feels like “exploring a jungle with a flashlight.” He walks us through Asimov, Reflection’s newly unveiled code‑research agent that fuses long‑context search, team‑wide memory and multi‑agent planning so developers spend less time spelunking for context and more time shipping.

We also rewind his unlikely journey—from physics prodigy in a Manhattan‑Project desert town, to Berkeley’s AI crucible, to leading RLHF for Google Gemini—before he left big‑lab comfort to chase a sharper vision of enterprise super‑intelligence. Along the way: the four breakthroughs that unlocked modern AI, why capital efficiency still matters in the GPU arms‑race, and how small teams can lure top talent away from nine‑figure offers.

If you’re curious about the next phase of AI agents, the future of developer tooling, or the gritty realities of scaling a frontier‑level startup—this episode is your blueprint.

Reflection AI

Website - https://reflection.ai

LinkedIn - https://www.linkedin.com/company/reflectionai

Misha Laskin

LinkedIn - https://www.linkedin.com/in/mishalaskin

X/Twitter - https://x.com/mishalaskin

FIRSTMARK

Website - https://firstmark.com

X/Twitter - https://twitter.com/FirstMarkCap

Matt Turck (Managing Director)

LinkedIn - https://www.linkedin.com/in/turck/

X/Twitter - https://twitter.com/mattturck

(00:00) Intro

(01:42) Reflection AI: Company Origins and Mission

(04:14) Making Superintelligence Concrete

(06:04) Superintelligence vs. AGI: Why the Goalposts Moved

(07:55) Organizational Superintelligence as an Oracle

(12:05) Coding as the Shortcut: Hands, Legs & Brain for AI

(16:00) Building the Context Engine

(20:55) Capturing Tribal Knowledge in Organizations

(26:31) Introducing Asimov: A Deep Code Research Agent

(28:44) Team-Wide Memory: Preserving Institutional Knowledge

(33:07) Multi-Agent Design for Deep Code Understanding

(34:48) Data Retrieval and Integration in Asimov

(38:13) Enterprise-Ready: VPC and On-Prem Deployments

(39:41) Reinforcement Learning in Asimov's Development

(41:04) Misha's Journey: From Physics to AI

(42:06) Growing Up in a Science-Driven Desert Town

(53:03) Building General Agents at DeepMind

(56:57) Founding Reflection AI After DeepMind

(58:54) Product-Driven Superintelligence: Why It Matters

(01:02:22) The State of Autonomous Coding Agents

(01:04:26) What's Next for Reflection AI

Agentic commerce is no longer science fiction — it’s arriving in your browser, your development IDE, and soon, your bank statement. In this episode of The MAD Podcast, Matt Turck sits down with Emily Glassberg Sands, Stripe’s Head of Information, to explore how autonomous “buying bots” and the Model Context Protocol (MCP) are reshaping the very mechanics of online transactions. Emily explains why intent, not clicks, will become the primary interface for shopping and how Stripe’s rails are adapting for tokens, one-time virtual cards, and real-time risk scoring that can tell good bots from bad ones in milliseconds.

We also go deep into Stripe's strategic AI choices. Drawing on $1.4 trillion in annual payment flow—1.3 percent of global GDP—Stripe decided to train its own payments foundation model, turning tens of billions of historical charges into embeddings that boost fraud-catch recall from 59 percent to 97 percent. Emily walks us through the tech: why they chose a BERT encoder over GPT-style decoders, how three MLEs in a “research bubble” birthed the model, and what it takes to run it in production with five-nines reliability and tight latency budgets.

We zoom out to Stripe’s unique vantage point on the broader AI economy. Their data shows the top AI startups hitting $30 million in ARR three times faster than the fastest SaaS companies did a decade ago, with more than half of that revenue already coming from overseas markets. Emily unpacks the new billing playbook—usage-based pricing today, outcome-based pricing tomorrow—and explains why tiny teams of 20–30 people can now build global, vertically focused AI businesses almost overnight.

Stripe

Website - https://stripe.com

X/Twitter - https://x.com/stripe?

Emily Glassberg Sands

LinkedIn - https://www.linkedin.com/in/egsands

X/Twitter - https://x.com/emilygsands

FIRSTMARK

Website - https://firstmark.com

X/Twitter - https://twitter.com/FirstMarkCap

Matt Turck (Managing Director)

LinkedIn - https://www.linkedin.com/in/turck/

X/Twitter - https://twitter.com/mattturck

(00:00) Intro

(01:45) How Big Is Stripe? Latest Stats Revealed

(04:06) What Does “Head of Information” at Stripe Actually Do?

(05:43) From Harvard to Stripe: Emily’s Unusual Journey

(08:54) Why Stripe Built Its Own Foundation Model

(13:19) Cracking the Code: How Stripe Handles Complex Payment Data

(16:25) Foundation Model vs. Traditional ML: What’s Winning?

(20:09) Inside Stripe’s Foundation Model: How It Was Built

(24:35) How Stripe Makes AI Decisions Transparent

(28:38) Where Stripe Uses AI (And Where It Doesn’t)

(34:10) How Stripe’s AI Drives Revenue for Businesses

(41:22) Real-Time Fraud Detection: Stripe’s Secret Sauce

(42:51) The Future of Shopping: AI Agents & Agentic Commerce

(46:20) How Agentic Commerce Is Changing Stripe

(49:36) Stripe’s Vision for a World of AI-Powered Buyers

(55:46) What Is MCP? Stripe’s Take on Agent-to-Agent Protocols

(59:31) Stripe’s Data on AI Startups Monetizing 3× Faster

(01:03:03) How AI Companies Go Global — From Day One

(01:07:48) The New Rules: Billing & Pricing for AI Startups

(01:10:57) How Stripe Builds AI Literacy Across the Company

(01:14:05) Roadmap: Risk-as-a-Service, Order Intent, and Beyond

Welcome to a special FirstMark Deep Dive edition of the MAD Podcast. In this episode, Matt Turck and David Waltcher unpack the explosive impact of generative AI on engineering — hands-down the biggest shift the field has seen in decades. You’ll get a front-row seat to the real numbers and stories behind the AI code revolution, including how companies like Cursor hit a $500M valuation in record time, and why GitHub Copilot now serves 15 million developers.

Matt and David break down the six trends that shaped the last 20 years of developer tools, and reveal why coding is the #1 use case for generative AI (hint: it’s all about public data, structure, and ROI). You’ll hear how AI is making engineering teams 30-50% faster, but also why this speed is breaking traditional DevOps, overwhelming QA, and turning top engineers into full-time code reviewers.

We get specific: 82% of engineers are already using AI to write code, but this surge is creating new security vulnerabilities, reliability issues, and a total rethink of team roles. You’ll learn why code review and prompt engineering are now the most valuable skills, and why computer science grads are suddenly facing some of the highest unemployment rates.

We also draw wild historical parallels—from the Gutenberg Press to the Ford assembly line—to show how every productivity boom creates new problems and entire industries to solve them. Plus: what CTOs need to know about hiring, governance, and architecture in the AI era, and why being “AI native” can make a startup more credible than a 10-year-old giant.

Matt Turck (Managing Director)

LinkedIn - https://www.linkedin.com/in/turck/

X/Twitter - https://twitter.com/mattturck

David Waltcher

LinkedIn - https://www.linkedin.com/in/davidwaltcher

X/Twitter - https://x.com/davidwaltcher

FIRSTMARK

Website - https://firstmark.com

X/Twitter - https://twitter.com/FirstMarkCap

(00:00) Intro & episode setup

(01:50) The 6 waves that led to GenAI engineering

(04:30) Why coding is such fertile ground for Generative AI

(08:25) Break-out dev-tool winners: Cursor, Copilot, Replit, V0

(11:25) Early stats: Teams Are Shipping Code Faster with AI

(13:32) Copilots vs Autonomous Agents: The Current Reality

(14:14) Lessons from History: Every Tech Boom Creates New Problems

(21:53) FirstMark Survey: The Headaches AI Is Creating for Developers

(22:53) What’s Now Breaking: Security, CI/CD flakes, QA Overload

(29:16) The New CTO Playbook to Adapt to the AI Revolution

(33:23) What Happens to Engineering Orgs if Everyone is a Coder?

(40:19) Founder opportunities & the dev-tool halo effect

(44:24) The Built-in Credibility of AI-Native Startups

(46:16) The Irony of Dev Tools As Biggest Winners in the AI Gold Rush

(47:43) What’s Next for AI and Engineering?

In this episode, Vercel CEO Guillermo Rauch goes deep on how V0, their text-to-app platform, has already generated over 100 million applications and doubled Vercel’s user base in under a year.

Guillermo reveals how a tiny SWAT team inside Vercel built V0 from scratch, why “vibe coding” is making software creation accessible to everyone (not just engineers), and how the AI Cloud is automating DevOps, making cloud infrastructure self-healing, and letting companies expose their data to AI agents in just five lines of code.

You’ll hear why “every company will have to rethink itself as a token factory,” how Vercel’s Next.js went from a conference joke to powering Walmart, Nike, and Midjourney, and why the next billion app creators might not write a single line of code. Guillermo breaks down the difference between vibe coding and agentic engineering, shares wild stories of users building apps from napkin sketches, and explains how Vercel is infusing “taste” and best practices directly into their AI models.

We also dig into the business side: how Vercel’s AI-powered products are driving explosive growth, why retention and margins are strong, and how the company is adapting to a new wave of non-technical users. Plus: the future of MCP servers, the security challenges of agent-to-agent communication, and why prompting and AI literacy are now must-have skills.

Vercel

Website - https://vercel.com

X/Twitter - https://x.com/vercel

Guillermo Rauch

LinkedIn - https://www.linkedin.com/in/rauchg

X/Twitter - https://x.com/rauchg

FIRSTMARK

Website - https://firstmark.com

X/Twitter - https://twitter.com/FirstMarkCap

Matt Turck (Managing Director)

LinkedIn - https://www.linkedin.com/in/turck/

X/Twitter - https://twitter.com/mattturck

(00:00) Intro

(02:08) What Is V0 and Why Did It Take Off So Fast?

(04:10) How Did a Tiny Team Build V0 So Quickly?

(07:51) V0 vs Other AI Coding Tools

(10:35) What is Vibe Coding?

(17:05) Is V0 Just Frontend? Moving Toward Full Stack and Integrations

(19:40) What Skills Make a Great Vibe Coder?

(23:35) Vibe Coding as the GUI for AI: The Future of Interfaces

(29:46) Developer Love = Agent Love

(33:41) Having Taste as Developer

(39:10) MCP Servers: The New Protocol for AI-to-AI Communication

(43:11) Security, Observability, and the Risks of Agentic Web

(45:25) Are Enterprises Ready for the Agentic Future?

(49:42) Closing the Feedback Loop: Customer Service and Product Evolution

(56:06) The Vercel AI Cloud: From Pixels to Tokens

(01:10:14) How Vercel Adapts to the ICP Change?

(01:13:47) Retention, Margins, and the Business of AI Products

(01:16:51) The Secret Behind Vercel Last Year Growth

(01:24:15) The Importance of Online Presence

(01:30:49) Everything, Everywhere, All at Once: Being CEO 101

(01:34:59) Guillermo's Advice to Younger Self

Canva just announced $3 billion in ARR, 230 million monthly active users, and 24 million paying subscribers—including 95% of the Fortune 500. Even more impressive? They’ve been profitable for seven years while growing at 40–50% per year.

In this episode, Canva’s Head of Engineering, Brendan Humphreys, reveals how he went from employee #12 to leading 2,300 engineers across continents, and why Canva’s “pragmatic excellence” lets them ship AI features at breakneck speed—like launching Canva Code to 100 million users in just three months.

Brendan shares the story of Canva’s AI journey: building an in-house ML team back in 2017, acquiring visual AI startups like Kaleido and Leonardo AI, and why they use a hybrid of OpenAI, Anthropic, Google, and their own foundation models. He explains how Canva’s App Store gives niche AI startups instant access to millions, and why their $200M Creator Fund is designed to reward contributors in the AI era. You’ll also hear how AI tools like Copilot are making Canva’s senior engineers 30% more productive, why “vibe coding” isn’t ready for prime time, and the unique challenges of onboarding junior engineers in an AI-driven world.

We also dig into Canva’s approach to technical debt, scaling from 12 to 5,000 employees, and why empathy is a core engineering skill at Canva.

Canva

Website - https://www.canva.com

X/Twitter - https://x.com/canva

Brendan Humphreys

LinkedIn - https://www.linkedin.com/in/brendanhumphreys

X/Twitter - https://x.com/brendanh

FIRSTMARK

Website - https://firstmark.com

X/Twitter - https://twitter.com/FirstMarkCap

Matt Turck (Managing Director)

LinkedIn - https://www.linkedin.com/in/turck/

X/Twitter - https://twitter.com/mattturck

(00:00) Intro

(01:14) Canva’s Mind-Blowing Growth and Profitable Journey

(03:41) Why Brendan Left Atlassian to Join a Tiny Startup

(06:17) What Being a Founder Taught Brendan About Leadership

(07:24) Growing with Canva: From 12 Employees to 2,300 Engineers

(10:02) How Canva Runs a Global Team from Sydney to Europe

(13:16) Is AI a Threat or a Superpower for Canva?

(15:22) The Real Story Behind Canva’s AI and Machine Learning Team

(17:23) How Canva Ships New AI Features So Fast

(19:19) A Tour of Canva’s Latest AI-Powered Products

(21:03) From Design Tool to All-in-One Productivity Platform

(26:21) Keeping Up the Pace: How Canva Moves So Quickly

(30:22) The Future: AI Agents, Copilots, and Smarter Workflows

(33:14) How AI Tools Are Changing the Way Engineers Work

(35:47) Rethinking Hiring and Training in the Age of AI

(37:01) Why Empathy Matters in Engineering at Canva

(39:41) Building vs. Buying: How Canva Chooses Its AI Tech

(41:23) Lessons Learned: Technical Debt and Scaling Pains

(51:18) Shipping Fast Without Breaking Things

(53:08) What’s Next: AI Video, New Features, and Big Ambitions

AI coding is in full-blown gold-rush mode, and GitHub sits at the epicenter. In this episode, GitHub CEO Thomas Dohmke tells Matt Turck how a $7.5 B acquisition in 2018 became a $2 B ARR rocket ship, and reveals how Copilot was born from a secret AI strategy years before anyone else saw the opportunity.

We dig into the dizzying pace of AI innovation: why developer tools are suddenly the fastest-growing startups in history, how GitHub’s multi-model approach (OpenAI, Anthropic Claude 4, Gemini 2.5, and even local LLMs) gives you more choice and speed, and why fine-tuning models might be overrated. Thomas explains how Copilot keeps you in the “magic flow state,” how even middle schoolers are using it to hack Minecraft.

The conversation then zooms out to the competitive battlefield: Cursor’s $10 B valuation, Mistral’s new code model, and a wave of AI-native IDE forks vying for developer mind-share. We discuss why 2025’s “coding agents” could soon handle 90 % of the world’s code, the survival of SaaS and why the future of coding is about managing agents, not just writing code.

GitHub

Website - https://github.com/

X/Twitter - https://x.com/github

Thomas Dohmke

LinkedIn - https://www.linkedin.com/in/ashtom

X/Twitter - https://twitter.com/ashtom

FIRSTMARK

Website - https://firstmark.com

X/Twitter - https://twitter.com/FirstMarkCap

Matt Turck (Managing Director)

LinkedIn - https://www.linkedin.com/in/turck/

X/Twitter - https://twitter.com/mattturck

(00:00) Intro

(01:50) Why AI Coding Is Ground Zero for Generative AI

(02:40) The $7.5B GitHub Acquisition: Microsoft’s Strategic Play

(06:21) GitHub’s Role in the Azure Cloud Ecosystem

(10:25) How GitHub Copilot Beat Everyone to Market

(16:09) Copilot & VS Code Explained for Non-Developers

(21:02) GitHub Models: Multi-Model Choice and What It Means

(25:31) The Reality of Fine-Tuning AI Models for Enterprise

(29:13) The Dizzying Pace and Political Economy of AI Coding Tools

(36:58) Competing and Partnering: Microsoft’s Unique AI Strategy

(41:29) Does Microsoft Limit Copilot’s AI-Native Potential?

(46:44) The Bull and Bear Case for AI-Native IDEs Like Cursor

(52:09) Agent Mode: The Next Step for AI-Powered Coding

(01:00:10) How AI Coding Will Change SaaS and Developer Skills