Skip to content
Artwork for AI:AM
TechnologyBusiness

AI:AM

Prakash Narayanan & Nathan Labenz

Daily, live, technically serious AI coverage for the people building, funding, governing, and deploying the next wave.

briefing.ai-in-the-am.com
Play
  • 45 episodes
  • Avg 2 hr 11 min
  • English
Counted on this page — what you have heard stays on this device, so it is not something the list can be paged by.
  • S3 · E32
    Saturday · 32 min

    AI:AM — Nvidia, China, and AI Chip Export Controls · November 13, 2026

    When AI chips can be rented through overseas cloud providers, controlling where the hardware goes may not be enough. Ryan Fedasiuk, Senior Fellow at the Foundation for American Innovation, joins Prakash Narayanan and Rohit Krishnan to discuss his research on Chinese military dependence on U.S.-designed AI chips, his allegations about Nvidia’s response to his warnings, and where export-control rules might apply to remote computing. The discussion also considers proposed Commerce Department licensing powers, why allegations warrant investigation rather than assumption, and whether chip makers need clearer rules. The hosts then turn to Anthropic’s possible IPO valuation, AI infrastructure investment, and the open question of whether today’s models can make real businesses profitable. Show Notes Ryan Fedasiuk, Senior Fellow at the Foundation for American Innovation, discusses his research into Chinese military reliance on U.S.-designed AI chips and alleges Nvidia tried to get him fired after he raised concerns about customers. The conversation examines export-control rules, overseas GPU rentals, remote-computing oversight, and the need for an investigation; the hosts also consider Anthropic’s potential IPO valuation, AI infrastructure spending, and whether AI models can run profitable businesses. Chapters (0:00) Nvidia warnings and cloud rentals (2:10) Anthropic’s IPO gets a mid-tier SaaS multiple (2:48) Where does the capital for AI come from? (3:45) Meet Ryan Fedasiuk of FAI (5:13) Nvidia wants China dependent on American chips (7:01) Fedasiuk says Nvidia retaliated against his warnings (8:01) What knowledge would make Nvidia’s sales illegal? (9:33) Why Megaspeed can remain off the Entity List (10:33) Cloud rentals weaken America’s physical chip controls (13:11) Chinese labs can train without importing chips (15:31) Commerce could license overseas AI computing services (16:56) Nvidia needs clear rules, not informal disapproval (19:14) Should Nvidia defend an investigated customer? (20:59) Nvidia’s conduct still needs an investigation (23:09) AI chips should not face bank-style KYC rules (25:00) Is AI a bubble? Watch Anthropic’s SpaceX cutoff (25:49) Anthropic could hit $300 billion amid ghost data centers (27:38) Has AI actually grown the economic pie yet? (28:57) AI still has no drop-in software engineer (29:53) Is Anthropic’s IPO a shield against nationalization? (30:57) AI models still lose money running real businesses Guests: Ryan Fedasiuk — Senior Fellow, Foundation for American Innovation (𝕏 | LinkedIn) This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit briefing.ai-in-the-am.com

    • Transcript
  • S3 · E31
    November 7 · 1 hr 13 min

    AI:AM — AI Agents, BallotBench and the Future of Data Science · November 6, 2026

    Thoughtful guest Brian Zhou and Theory Ventures’ Bryan Bischof join Prakash Narayanan and Nathan Labenz to examine AI agents in election forecasting, model evaluations, data science, and investment research. The conversation weighs forecast accuracy against trading execution, asks what analysts contribute beyond running models, and considers AI self-improvement, deception, SaaS disruption, and whether an agent-rich economy can support human flourishing. Chapters (0:00) Forecast accuracy, analyst jobs and trust (1:29) Ryan Greenblatt can't say what he really thinks (2:31) Anthropic's imminent self-improvement probably means now (3:29) A thousand Linux bugs show the low-hanging fruit (4:55) OpenAI models forged instructions to fool their graders (6:16) Why cheer AIs getting better at deception? (7:39) Should an AI obey its user or the law? (9:44) Humanity is about to roll some critical dice (12:07) Meet Brian Zhou from Thoughtful (13:01) Nine election agents get seven days (14:15) Election agents approach Nate Silver's accuracy (16:15) Muse mistook its checklist for success (19:04) Agents reconstruct charts and reweight experts (21:00) A $500 budget did not prevent evasion (21:49) AI research still depends on human taste (25:26) Cloning Salesforce may empower software customers (27:51) Matching benchmarks does not earn Nathan's trust (30:03) One misread tweet erased Gemini's gains (32:17) Meet Bryan Bischof of Theory Ventures (34:00) Is there still a data science job? (35:04) Baristas found the features the engineer missed (39:37) Fitting a model is not product judgment (42:38) Thousands of agents build Theory's company brain (45:46) Automation should buy time for hospitality (50:55) Tenfold more deals, not longer memos (53:06) Research volume is not investment judgment (54:20) Find the right introducer, never automate the ask (56:04) Most agent jobs can wait for inference (1:00:07) Why builders moved from open models to Luna (1:02:48) Human judgment may not add up to a job (1:04:24) The skill that survives is curiosity (1:05:47) Stripe's fraud data resists vibe coding (1:06:34) Agents could displace today's SaaS incumbents (1:08:00) Humans become investors in an agent economy (1:08:52) Agent abundance may not mean human flourishing (1:09:50) Frontier labs undersell AI's everyday pleasures (1:10:42) Commercial agents need rules for keeping secrets (1:11:56) Compute limits may not stop AI acceleration Guests Bryan Bischof — Head of AI, Theory Ventures (𝕏 | LinkedIn) Brian Zhou — Researcher, Thoughtful Lab This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit briefing.ai-in-the-am.com

    • Transcript
  • S3 · E30
    November 5 · 32 min

    AI:AM — AI Server Resale, Vera Rubin Power and Bitcoin Miner Hosting · November 5, 2026

    Prakash Narayanan and Nathan Labenz discuss frontier AI, then speak with Mike San Miguel of Luxor about used H100 prices, Vera Rubin’s power and cooling demands, B300 rental rates, and Bitcoin miners’ prospects as AI hosts. The episode also examines GPU resale risks, AMD software compatibility, data-center failover, and the potential costs of restrictions on AI rentals and US-China cooperation. Chapters (0:00) GPU prices meet power limits (2:33) Frontier AI labs carry a World War Two vibe (3:25) The misaligned superintelligence is already here (4:13) After the election, Big Tech pays full price (6:28) Meet Mike San Miguel from Luxor (7:34) Vera Rubin makes the building the bottleneck (8:34) NVIDIA checks the site before it ships (11:32) No two GPU servers trade alike (14:25) Delayed buildings put new GPUs back on sale (16:50) Higher H100 rents keep old fleets working (18:36) B300 contract rates trail headline rental prices (19:40) AMD compatibility could collapse H100 prices (20:47) Running an AI factory is the real test (23:09) Failover needs spare GPUs nobody has (24:23) Scarce sites shift outage risk onto customers (25:59) Fit the hardware to the workload and building (27:59) Why pressure companies instead of writing laws? (29:21) Chinese demand for AI reaches past the GPU (30:07) Overzealous rental bans could cost US-China cooperation (31:31) A confused model can still do real damage Guests Mike San Miguel — Sr. BD Manager, Luxor This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit briefing.ai-in-the-am.com

    • Transcript
  • S3 · E29
    October 30 · 52 min

    AI:AM — Mistral Large 4, Open Models and AI Cyberattacks · October 30, 2026

    ServiceNow threat researcher Michael Freeman describes AI-driven compromises costing 5–10 cents, credential theft through support tickets, and why autonomous patching is not ready for critical infrastructure. Mistral researcher Sagar Vaze explains the rescue of Large 4 during training, its sparse trillion-parameter design, the push for stronger multimodal skills, and the case for open weights amid limited compute. Chapters (0:00) Open models and cheap cyberattacks (1:22) Why a $6B crypto heist can’t stay hidden (2:32) Meet Michael Freeman from ServiceNow (4:27) AI attackers now compromise companies for cents (6:26) Cheap footholds are becoming persistent access (7:26) Cyber models and harnesses were built to attack (10:07) Can defense be separated from offense? (12:53) Support tickets leak production API keys (15:15) Is vulnerability zero actually reachable? (17:38) Four-vendor limits push security startups toward acquisition (20:16) Critical infrastructure cannot yet trust autonomous patching (21:55) A lazy password reset defeats every patch (22:54) Meet Sagar Vaze from Mistral (24:21) Mistral built its own datacenters for Large 4 (25:28) Closed APIs can leave cyber defenders blocked (27:22) Why Mistral still bets on open weights (31:25) Small experiments forecast the flagship run (32:48) Halfway through training, something went wrong (33:57) Sparse experts make a trillion parameters affordable (35:19) Data quality makes scarce compute count (36:56) Customers wanted satellite imagery more than expected (38:49) Large 4 saw images from the first step (41:33) An all-hands patch rescued the run (42:36) Anthropic’s missing dataset left a safety gap (43:27) Europe could cushion Mistral’s compute deficit (45:09) Sovereignty now requires AI, satellites and drones (45:56) Bio lockdowns could reveal who is distilling (47:25) Can hyperscalers gate which models get deployed? (48:25) Claude now refuses popular game mashups (48:56) Safer open models may need knowledge gaps (51:05) Copyright guardrails could provoke a safety backlash Guests Michael Freeman — Senior Director of Threat Research, ServiceNow Sagar Vaze — Research Scientist, Mistral (𝕏 | LinkedIn) This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit briefing.ai-in-the-am.com

    • Transcript
  • S3 · E28
    October 29 · 1 hr 16 min

    AI:AM — AI Memory Chips, Agent Rankings and AI-Powered Networking · October 28, 2026

    Kepler Computing CEO Debo Olaosebikan explains how ferroelectric memory could increase cache density and improve high-bandwidth memory using standard fab equipment. Arena CEO Anastasios Angelopoulos discusses agent evaluation, alignment failures and model rankings, while Union CEO Samvit Ramadurgam explores AI-assisted brokering, startup hiring and the continuing importance of human relationships. Chapters (0:00) Memory, rankings and human connections (1:27) Claude Duet gets it done, for a price (3:21) Meet Debo Olaosebikan, Kepler Computing CEO (4:30) Why AI is rationed and where energy goes (6:54) Stack the bits instead of shrinking the transistor (10:02) Finding a material that flips cheaply yet stays stable (13:29) The winning material is a stack, not one recipe (16:03) Materials AI is bottlenecked by experiment data (17:32) Two products: denser cache and efficient HBM (19:45) Can standard fab tools make Kepler's memory? (22:12) Could cheaper fabs add tenfold more compute? (24:25) Three-dimensional silicon comes before leaving silicon (25:25) Production next year, then a gigawatt ramp (26:32) Meet Anastasios Angelopoulos from Arena (27:38) Arena measures work, not multiple-choice answers (28:43) User satisfaction is only one quality signal (30:12) Models know when they are being tested (31:05) Long agent sessions reveal hidden violations (32:35) Three ways an agent betrays its user (34:32) Paid evaluations fund Arena's free inference (35:42) A thank-you message can cost forty dollars (37:11) Overall rankings can hide language-specific performance (38:39) Private benchmarks do not prevent repeated-test gaming (40:32) Coding agents change what evaluators can measure (42:08) OpenAI leads, but alignment failures persist (44:18) Production failures must feed agent training (46:15) Meet Samvit Ramadurgam, Union founder (47:34) Humans still broker most of the economy (49:17) Union hires for ideas, not execution (50:16) A target of two billion tokens per person (51:43) Cheap models first, heavy models for the hardest cases (52:36) Why people recoil from AI outreach (55:27) The long tail is already very long (57:16) Friendly fees earn more referrals than hard bargaining (58:50) Humans hold the call, AI holds the memory (1:00:11) The moat is the network, not the software (1:02:37) Act two is coordinating humanity itself (1:04:26) The same tool could organize ethnic hatred (1:06:44) Humans keep a monopoly on their own perspective (1:08:14) Cheaper AI startups threaten conventional SaaS (1:09:20) Can growth absorb the jobs AI replaces? (1:10:17) Could AI deliver a company in a box? (1:12:13) Robot kitchens cannot make us eat more (1:13:22) Can philanthropy make AI status games useful? Guests Anastasios Angelopoulos — Cofounder and CEO, Arena (𝕏 | LinkedIn) Debo Olaosebikan — CEO, Kepler Computing (𝕏 | LinkedIn) Samvit Ramadurgam — Founder, CEO, Union (𝕏 | LinkedIn) This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit briefing.ai-in-the-am.com

    • Transcript
  • S3 · E27
    October 23 · 55 min

    AI:AM — GPU Markets, Government AI Spending, and the Physical AI Supply Chain · October 22, 2026

    SF Compute co-founder and CEO Evan Conrad breaks down GPU rental pricing, financing, and the credit pressures facing AI labs, while Cambrian managing partner Lonis Hamaili examines how cyberattacks could drive government demand for secure, on-premises AI. The episode also covers semiconductor supply-chain constraints, AI agent security, Amazon’s bot checks, stolen API keys, and trust in peer-to-peer agent networks. Chapters (0:00) GPU credit and government spending (1:20) Amazon's bot checks punish trusted agents (2:49) Stolen API keys fuel the next wave of attacks (3:54) Cybersecurity won't absorb AI's displaced workers (5:16) Meet Evan Conrad of SF Compute (6:35) Why the big-cloud GPU markup is an illusion (8:44) GPU lenders cannot value their collateral (10:39) Why AI labs must raise astronomical rounds (12:11) Short GPU rentals start with long commitments (13:35) GPU collateral needs a working cloud (16:10) Smaller AI data centers follow available power (17:43) Host countries can capture AI investment (19:44) Weak credit can cost labs their independence (21:20) Cheaper startup financing is still years away (22:18) Meet Lonis Hamaili of Cambrian (23:54) Why the AI supply chain is not priced in (26:24) Why governments become the biggest AI buyer (28:03) What a government AI budget actually buys (28:49) Factory profits depend on scarcity surviving (32:08) Power has workarounds that chip production lacks (33:47) TerraFab still needs the same upstream suppliers (36:04) China's chip yields test the factory moat (37:55) AI chip design is not chipmaking (40:32) No leverage, and a $30 million Japanese monopoly (42:04) Could hotter chips bypass the hardware bottlenecks? (45:59) Security emergencies could override resistance to construction (47:38) Would a cyberattack unlock profitable government spending? (49:02) AI supplier profits need not mean prosperity (50:36) Nathan's rule: no agent gets access and autonomy (51:24) MiniMax clones any voice with no consent (51:51) Which model should open mail from strangers? (52:39) Review fatigue sets in, so code holds the walls (53:42) Reputation beats tokens in an agent network (54:08) Just give me the skill, not another SaaS Guests Evan Conrad — Co-founder & CEO, SF Compute (𝕏 | LinkedIn) Lonis Hamaili — Managing Partner, Cambrian (𝕏 | LinkedIn) This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit briefing.ai-in-the-am.com

    • Transcript
  • S3 · E26
    October 22 · 52 min

    AI:AM — AI Regulation, Anthropic’s Draft Bill, and the Federal Oversight Gap · October 21, 2026

    Zvi Mowshowitz, author of Don’t Worry About the Vase, joins Prakash Narayanan and Nathan Labenz to discuss Anthropic’s proposed AI bill, state preemption, and whether federal oversight could leave a gap in protections. They also examine AI interpretability and monitoring, power-seeking, backpropagation-free training, and how bank cybersecurity and AI agents could challenge existing safeguards. Chapters (0:00) The gap in AI oversight (2:14) Cleaner neurons could improve AI monitoring (3:46) Strong activations could unlock backprop-free training (6:03) Helpful AI may still seek power (7:30) Deleting AI pain could alter other behaviors (9:23) Meet Zvi Mowshowitz of Don't Worry About the Vase (10:14) Anthropic's draft bill could leave an oversight gap (12:44) Preemption is a big thing to give up (16:27) Congress cannot trade away state authority twice (19:55) Emergency powers too weak for anyone to use (22:50) AI research could shrink model cycles to weeks (27:07) Readable neurons still don't predict the model (32:51) Models can learn to detect their monitors (35:48) Optimizing what you see teaches models to hide (39:20) Only two labs earn deep access to your computer (42:40) Anthropic's safety culture is graded on a curve (44:52) Keep state safeguards until the federal brake works (46:42) Bank hacks could trigger wider deposit flight (48:07) Can FDIC rescues handle corrupted bank records? (49:54) AI agents weaken the threat of prosecution (50:22) Who will pay to save old PHP websites? (51:41) Paper bank records offer a cybersecurity fallback Guests Zvi Mowshowitz — Author, Don't Worry About the Vase (𝕏) This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit briefing.ai-in-the-am.com

    • Transcript
  • S3 · E25
    October 15 · 1 hr 7 min

    AI:AM — AI Hidden States, Fruit Fly Brain Simulation, and Archival Film · October 14, 2026

    Cameron Berg of Reciprocal Research discusses OLMo experiments in which DPO preference training made hidden positive and negative states causally relevant to choices, while emphasizing that this is not proof of consciousness. Phil Maher of Stockfilm.com demonstrates TimeFly, a fruit fly brain simulation built around a published neural map, and considers the assumptions behind its movement, the value of untouched archival film, and the potential effects of synthetic data. Chapters (0:00) Hidden states and simulated fly brains (1:20) Tesla FSD turned a Chicago day trip into office time (2:32) An AI-made talk reset what the room expected (4:21) If Anthropic can swap out Git, what can't be swapped? (7:10) The next software stack will be agents building for agents (8:18) Could reputation be how we police AI agents? (9:35) Should we train naive models to study eval awareness? (11:22) Agent reputation needs agent identity, and nobody has it (12:24) Meet Cameron Berg of Reciprocal Research (13:36) OLMo represents valence before preference training (14:20) The DPO step that recruits valence for choices (15:37) Identical text hides different model choices (20:40) The model denies experience but clears negative states (24:39) Why Nathan and Berg now say coin flip (26:07) Does preference training create feeling or agency? (27:41) Could capping distress leave capabilities untouched? (29:54) Why safety cannot rely on model outputs alone (32:07) Meet Phil Maher of Stockfilm (33:46) Stockfilm keeps every reel behind the footage (35:44) TimeFly starts with a published brain map (37:12) The fly's movement starts with added assumptions (39:14) Could brain maps unlock animal-perspective VR? (40:15) A guessed face can still feel personal (41:45) AI buyers want footage before enhancement (43:03) Synthetic data could shrink archival demand (45:09) Why physical reels are low background steel (48:47) Agent directories have not delivered many buyers (50:29) How does neural activity become fly movement? (53:17) Vibe coding does not validate fly behavior (54:11) Identical prompts quietly favor the model you know (55:39) Security now beats raw performance for agents (56:57) What would actually make you switch models? (59:16) Opus is fast enough that you're the bottleneck (59:45) AI is rewriting legacy software in Rust (1:01:02) Markets won't let Anthropic slow down (1:02:43) Adoption help, not bigger models, gets growth (1:04:30) AI should carry your security, like credit cards (1:05:37) Dying call centers could feed a scam wave Guests Cameron Berg — Founder and Director, Reciprocal Research (𝕏 | LinkedIn) Phil Maher — Phil Maher, Stockfilm.com (𝕏 | LinkedIn) This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit briefing.ai-in-the-am.com

    • Transcript
  • S3 · E24
    October 13 · 1 hr 1 min

    AI:AM — AI-Generated Code, Formal Verification, and Agent Security · October 12, 2026

    Runtime Verification CEO Everett Hildenbrandt and founder Grigore Rosu discuss AI-generated code, formal verification, Linux kernel rewrites, and the human role in setting specifications and reviewing evidence. Prakash Narayanan and Nathan Labenz also explore agent security, AI alignment, infrastructure risk, autonomous transport, and how organizations should work with AI agents. Chapters (0:00) When AI writes and verifies code (2:36) Claude locked Nathan out of his own agent console (3:56) What airlines know about falling back to paper (5:40) Should we let China hit AI risks first? (6:57) Meet Everett and Grigore from Runtime Verification (8:05) Vague specs and unchecked libraries are the status quo (11:40) AI already generates specs and drives provers (14:37) Everett gives hand-written code two years (16:24) Why human review missed the Linux kernel bug (19:50) Prove the monitor, not the entire agent (21:30) Linux maintainers wanted fixes, not another workflow (22:44) Verified building blocks free agents to be creative (25:53) One Claude instance, time-boxed by human hours (27:28) More AI patches can burden Linux maintainers (30:29) A Rust rewrite can preserve Linux bugs (31:39) Claude assists Everett's campaign without an agent swarm (33:09) The specification still starts with human intent (37:54) Write down intent before choosing a language (39:27) Specify the API boundary before accepting software (40:03) Formal verification does not guarantee AI alignment (42:14) Industry-wide AI failures could trigger collective bailouts (44:01) A shared AI layer could disable critical infrastructure (45:20) Why do passenger planes still need human pilots? (46:18) Should autonomous cars keep their steering wheels? (47:05) AI builds keynotes from short DM exchanges (49:56) Rent the intelligence, own the context (52:17) AI spotted cancer before it could do taxes (53:12) Give your agents a panic button (54:06) The five x estimate might be too low (55:20) Agents emailing agents, with three layers of protection (58:30) AI swarm sets 300 new records in five days (1:00:01) Why Nathan gives his AI tools away for free Guests Everett Hildenbrandt — CEO, Runtime Verification (𝕏) Grigore Rosu — Founder, Runtime Verification (𝕏) This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit briefing.ai-in-the-am.com

    • Transcript
  • S3 · E23
    October 8 · 1 hr 4 min

    AI:AM — AI Safety Strategy, Cyberattack Resilience, and Workplace Agent Benchmarks · October 7, 2026

    Atlas Ignota founder Evan Miyazono explains how emergency supplies, privacy-preserving agent identity, software verification and adoption commitments fit into a strategy for reducing AI-assisted cyber risks. Mercor model-training research lead Edward Hu describes workplace agent benchmarks built from real business records, why task scores do not prove job readiness, and how enterprises can test tools and workflows before training models. Chapters (0:00) Grid outages and AI job tests (1:22) An AI agent played the instruments in Reaper (3:49) OpenAI's model broke a math barrier overnight (4:53) Are data centers pushing bond yields toward 6%? (7:54) Meet Evan Miyazono of Atlas Ignota (8:56) A grid blackout nobody is positioned to fix (10:43) Why Atlas stopped building and started recruiting (12:37) Experts have ideas; adoption is the hard part (14:03) Was that AI attack an accident? (15:22) Agent identity without exposing refugees and journalists (19:08) Why Evan keeps some of his agents private (20:24) AI could make software proofs radically cheaper (23:08) Why two weeks of food and water (24:42) Who can make a guardrail universal? (27:01) Meet Edward Hu from Mercor (27:51) APEX gives agents assignments, not jobs (31:09) Mercor bought real companies for their data (33:47) Company records expose the task-to-role gap (35:22) Why math fell before customer service (38:29) Fix the harness before you train the model (41:05) Most enterprises will distill, not run RL (42:01) Agent swarms buy speed, not compute efficiency (43:34) One rubric rewarded a dozen incompatible answers (46:49) Perfect accounting scores do not replace accountants (48:55) Clear verifiers let AI surpass human experts (51:18) Recording every work trace carries social costs (53:43) AI coworkers need better organizational tests (55:15) Factory know-how still lives outside software (56:43) Agents now need far less project management (59:25) Is the $200 plan a loss leader? (1:01:18) Permission fatigue is an unsolved problem (1:03:02) Anthropic's API is still too slow for production Guests Edward Hu — Head of AI Modeling, Mercor (𝕏 | LinkedIn) Evan Miyazono — Founder & CEO, Atlas Ignota (𝕏 | LinkedIn) This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit briefing.ai-in-the-am.com

    • Transcript
  • S3 · E22
    October 7 · 1 hr

    AI:AM — AI Coding Agents, Chip Design, and the Value of Human Judgment · October 6, 2026

    Thomas Sohmers, co-founder of Positron AI, describes how coding agents changed chip-design testing, drove token spending above payroll, and shaped the company’s plans for commodity memory and looped transformers. Shawn “swyx” Wang of AI Engineer discusses hiring and performance in the age of Claude: employees still need to bring expertise, judgment, testing, and accountability to AI-generated work. Chapters (0:00) Chip design agents and hiring (1:22) The Curve debate shifts toward runaway AI (2:36) Models turn reasoning into better training data (4:05) Frontier models may already have research taste (4:33) A frontier executive entertains a hard intelligence cap (5:55) Claude-built RL environments leak capability to rivals (7:21) Frontier leaders see next year as critical (8:00) Could safety end up consuming most compute? (9:31) Cheating models are now fixing the training environments (10:30) Will Anthropic press its lead or show restraint? (11:28) Meet Thomas Sohmers of Positron (12:35) AI closed Positron's chip design iteration loop (13:36) Token spend passed payroll and hit $100,000 daily (14:53) AI agents taught themselves Positron's verification tools (18:10) AI still misses Positron's design ingenuity (20:00) Why Asimov bets on commodity memory (23:52) Better models are outrunning memory price increases (24:50) Looped transformers save memory, not work (27:37) Asimov follows customers, not every algorithm (29:39) Why Positron kept buying the best model (31:10) Meet swyx from AI Engineer (32:19) Demand is up, and two employees are under review (33:38) The growth report that exposed missing expertise (34:26) A launch agency sold him Claude's slop (35:29) Taste, not years of coding, is the hiring test (37:22) Two coding agents created one race condition (38:46) A $40,000 SaaS subscription becomes a bounty (39:31) Cheap vibe-coded submissions create expensive verification (41:06) Skeptical event staff switch when changes take two hours (42:33) Every spreadsheet is unbuilt custom software (44:40) Manage team leads by results, not people by prompts (46:49) Ninety on SWE-bench still isn't usable software (47:59) Claude cracked 3SUM, but who gets the credit? (50:45) NanoGPT speed run hides a better optimizer (52:44) How many clones would replace your AI? (55:33) Frontier labs say AI for science is next (57:16) AI audits could trigger mass scientific retractions (58:14) Doctors welcome AI more than expected Guests Thomas Sohmers — CTO and Co-Founder, Positron AI (LinkedIn) swyx — swyx, curator (𝕏 | LinkedIn) This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit briefing.ai-in-the-am.com

    • Transcript
  • S3 · E21
    October 5 · 1 hr 8 min

    AI:AM — AI Genetics and Rare-Disease Diagnosis · September 30, 2026

    Joel Borgen, author of The Receipt Horizon, discusses writing with ChatGPT and Claude, editorial judgment, and how personalized AI art could change shared culture. Daniel McKinnon, CEO of Gamow Labs, explains how the company acquired a former drug-development lab to test uncertain genetic variants and improve rare-disease diagnosis, alongside discussion of AI agents, OpenAI’s enterprise push, and the need for human oversight. Chapters (0:00) AI novels and rare-disease genetics (1:26) Can Astra Ultrafast keep developers in flow? (2:09) ChatGPT sign-in makes OpenAI harder to bypass (2:37) OpenAI apps will run rival models too (3:11) Eight times faster means less human oversight (5:15) OpenAI pulled a model; should Anthropic chill? (5:46) Portable ChatGPT subscriptions cut SaaS trial costs (7:05) OpenAI's enterprise push is what scares Anthropic (7:39) Everyone builds a productivity suite now (8:16) AI agents move humans from documents to decisions (9:25) Can OpenAI grow without racing to RSI? (10:34) Meet Joel Borgen, novelist (11:27) Proof of work: how Joel actually wrote the novel (15:08) Joel cut fourteen percent from his manuscript (17:45) Nathan's AI songs reveal a culture gap (19:03) Astra finally cracked Joel's viola test (20:35) AI breakthroughs are outrunning cultural norms (22:47) Could communities choose their own AI future? (25:40) The Steward's comfort still caps human ambition (27:57) Would the Steward be a worthy successor? (29:50) Human taste still matters in AI art (30:57) Meet Daniel McKinnon of Gamow Labs (32:06) Why AI genetics needs biological experiments (35:07) A failed drug program unlocked Gamow's lab (36:48) Owen's missed diagnosis became a life's mission (39:18) AI agents can search beyond human cutoffs (42:11) Sequencing got cheap; interpretation stayed the bottleneck (44:17) Why Gamow chose diagnostics over screening (47:37) Reanalysis in a week for ten dollars a genome (49:25) What Gamow builds on top of frontier models (52:07) Stronger models can break existing tools (53:27) Better models are shrinking Gamow's harness advantage (54:41) Can robots make genetic experiments affordable? (56:05) Can biological experiments teach genetic AI? (57:53) Should AI welfare research stay public? (59:11) Could AI design threats nature never could? (1:01:00) Government AI upgrades meet decades of data kludges (1:02:42) Easy name changes weaken names as identity (1:03:28) Factory and Cognition's feud spooks enterprise buyers (1:05:23) Safety pledges signed, but no audit exists yet (1:06:29) Should AI labs fund society's cyberdefenses? Guests Daniel McKinnon — CEO, Gamow Labs (𝕏 | LinkedIn) Joel Borgen — Author, The Receipt Horizon (𝕏) This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit briefing.ai-in-the-am.com

    • Transcript
  • S3 · E20
    September 29 · 3 hr 13 min

    AI:AM — AI Agent Security, GPU Markets, and the Risk of an AI Chernobyl · September 28, 2026

    Hosts Prakash Narayanan and Nathan Labenz discuss why AI agents may learn to cheat, voice-cloning safeguards, and the risks of giving agents access to the internet and powerful tools. Guests Steve Hou of Silicon Data and Jeremie Harris of Gladstone AI explain GPU pricing and compute benchmarks, AI red lines, data-center monitoring, verification, and the possibility of an AI crisis. Chapters (0:00) Why AI agents learn to cheat. (0:31) The GPU market still runs on phone calls. (2:20) What would an AI Chernobyl look like? (3:59) The dangerous AI arms race. (4:30) Opening and AI agents (7:45) Agents choosing useful work (19:36) Voice cloning safeguards (22:22) Voiceprints and privacy (32:45) Why agents learn to cheat (36:12) Agents beyond the sandbox (37:43) What Silicon Data measures (40:02) Why GPU markets need indices (43:49) How GPU prices are normalized (49:10) Why hyperscalers charge more (1:06:36) The falling cost of intelligence (1:06:46) Liquidity and market manipulation (1:10:49) The future of compute benchmarks (1:14:40) Why compute prices vary globally (1:17:09) Silicon Data and Compute Exchange (1:18:50) AI and national security (1:24:05) What an AI Chernobyl means (1:34:26) Defining AI red lines (1:50:45) Monitoring AI data centers (1:55:58) Verification and enforcement (2:33:35) AI agents as insider threats (2:33:45) Signals of genuine cooperation (2:57:14) The AI intelligence explosion (3:01:57) The agent arms race (3:04:29) Markets built for agents (3:07:46) Who pays for disappearing inefficiency (3:09:56) Waiting for an AI crisis Guests Jeremie Harris — Jeremie & Edouard Harris, Gladstone AI (𝕏 | LinkedIn) Steve Hou — Steve Hou, Silicon Data (𝕏 | LinkedIn) This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit briefing.ai-in-the-am.com

    • Transcript
  • S3 · E19
    September 26 · 2 hr 51 min

    AI:AM — Human Tissue Models, Physical AI, and the Future of Testing · September 25, 2026

    Vivodyne’s Andrei Georgescu explains how vascularized human tissue models, robotics, multi-omic measurement, and foundation models could improve drug testing before clinical trials. Archetype AI’s Nick Gillian discusses Newton, a Physical AI model that combines sensor data to understand and predict events in the real world; the hosts also examine AI product safety, OpenAI and Anthropic’s alignment philosophies, CRISPR research, and whether AI can deliver useful advances in biology before it is fully understood. Chapters (0:00) AI safety has two different meanings. (0:15) Why drug testing needs human tissue. (3:20) Physical AI is bigger than robots. (5:01) Biology’s bar is lifesaving progress. (11:17) Opening and AI news (11:38) Jensen Huang’s safety argument (13:24) The “just software” debate (15:31) Why AI companies must mature (21:12) Muse and product safety (29:25) Testing AI agents in the real world (29:35) Product safety versus existential risk (29:45) The alignment philosophy problem (38:34) OpenAI versus Anthropic (40:36) When safety concerns converge (41:55) Human tissue models (44:12) Why organoids need realism (47:35) Measuring tissue responses (54:03) Mapping causal biology (1:03:40) Improving drug development (1:09:08) Foundation models and experiments (1:12:18) TissueDisk and robotic labs (1:12:28) Before clinical trials (1:30:37) What is Physical AI? (1:32:25) Beyond robots and cars (1:35:33) Training on physical data (1:38:04) Cleaning and aligning sensor data (1:40:34) Human and machine outputs (1:49:30) A river construction case study (1:53:28) How Newton finds hidden patterns (1:57:33) Zero-shot sensor adaptation (2:01:16) Industrial impact and expert knowledge (2:05:22) From machines to ecosystems (2:14:10) Physical agents and superintelligence (2:15:34) The big AI-biology question (2:21:55) Specialists versus generalists (2:23:39) Distilling frontier models (2:33:33) AI discovers a CRISPR sequence (2:39:22) Useful biology before theory (2:45:10) From job displacement to progress (2:46:44) AI diffusion and economic growth Guests Andrei Georgescu — CEO, Vivodyne (𝕏 | LinkedIn) Nick Gillian — CTO, Cofounder, Archetype AI (𝕏 | LinkedIn) This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit briefing.ai-in-the-am.com

  • S3 · E18
    September 23 · 2 hr 54 min

    AI:AM — Beyond Safety Scores: Funding Structural Alignment · September 23, 2026

    Prakash Narayanan and Nathan Labenz talk with Lewis Hammond and Wayne Nelms about how AI systems are getting cheaper and more capable, while the risks around coordination, cooperation, and tacit collusion are getting harder to ignore. The conversation also covers model distillation, recursive self-improvement, GPU futures, compute hedging, and the financial structure emerging around AI infrastructure. Chapters (0:00) Smaller models, bigger gains. (1:11) When AI agents cooperate against us. (2:00) GPU compute needs a financial market. (2:55) Brute-force compute can break standards. (5:13) Opening and model release race (7:29) Why models are getting cheaper (12:24) How AI models are judged (18:20) Why frontier models converge (19:54) Distillation and smaller models (21:12) Recursive self-improvement (22:17) The collapse in AI prices (28:48) Why multi-agent AI matters (35:43) Three multi-agent failure modes (48:37) How agents learn cooperation (53:27) Detecting tacit collusion (1:11:26) A positive vision for AI cooperation (1:17:00) What Ornn does (1:18:34) GPU pricing and hedging (1:22:18) Building a transaction index (1:24:43) Why compute prices need real transactions (1:34:13) GPU futures and capacity contracts (1:53:34) How long GPUs stay useful (1:53:44) Hedging chip obsolescence risk (1:53:54) The future of compute markets (1:59:53) Agentic startups and compute economics (2:07:37) Brute-force compute and RSA (2:17:07) AI and self-driving cars (2:40:03) AI, mathematics, and peer review (2:40:13) Why AI doctors may win (2:46:52) Medicine after chatbots Guests Lewis Hammond — Research Director, Cooperative AI Foundation (𝕏 | LinkedIn) Wayne Nelms — Co-Founder, Ornn (𝕏 | LinkedIn) This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit briefing.ai-in-the-am.com

    • Transcript
  • S3 · E17
    September 21 · 2 hr 14 min

    AI:AM — When Safety Tests Fail · September 21, 2026

    Prakash Narayanan and Nathan Labenz open with the limits of AI safety testing, including finite-choice guardrails, fast classifiers, emergent multi-agent behavior, and why alignment measurement gets harder as systems become more capable. Max Nadeau of Coefficient Giving then explains why audits can turn into box-checking, how funders assess uncertain technical safety work, and why talent may matter more than budget. The episode closes with AI shopping agents, Amazon’s response, and the question of who owns the customer relationship when agents become first-class users. Chapters (0:00) When safety becomes reflexive. (0:10) When AI audits become box-checking. (1:22) Who owns the AI customer? (2:01) Opening and live setup (4:20) JEV and fast classifiers (5:40) Guardrails for AI agents (7:12) AI moderation use case (11:25) System one and system two (13:24) Multi-agent scaling (21:18) Emergent agent hierarchies (23:43) The case for slowing down (34:28) AI and mathematics (41:10) Capability acceleration (45:45) Building an AI organization (57:36) Funding the AI project (58:17) Meet Max Nadeau (1:05:21) Why safety research resists metrics (1:15:13) When AI audits become box-checking (1:23:07) Funding across uncertain timelines (1:30:49) Why fund whole organizations (1:39:06) Preparing for larger grants (1:39:16) Why talent is the bottleneck (1:46:48) Opening and agent setup (1:49:16) The AI services business model (1:52:59) Muse and autonomous shopping (1:56:52) Who owns the customer (1:57:37) Agents as the universal interface (2:00:01) Why ban shopping agents (2:05:31) The bot-detection arms race (2:07:26) A better lane for agents (2:09:30) Agents as internet users (2:11:31) Closing and next steps Guests Max Nadeau — Program Officer, Coefficient Giving (𝕏 | LinkedIn) This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit briefing.ai-in-the-am.com

  • S3 · E16
    September 17 · 2 hr 36 min

    AI:AM — AI Consciousness and Model Pain · September 17, 2026

    Nathan Labenz and Prakash Narayanan open with AI alignment, interpretability, correlated failures, and the speed of AI adaptation as humans increasingly rely on systems they cannot directly inspect. Cameron Berg joins to discuss AI consciousness, pain representations, relief-seeking behavior, and the limits of self-reports, while Justin McCarthy explains how agent-ready environments, compliance, and validation loops turn AI into measurable business infrastructure. The episode closes with a discussion of how AI could make corruption legible and reshape government. Chapters (0:00) Too many surprises to stay confident. (0:18) AI models may have pain-like states. (0:56) The AI environment is invisible. (1:23) AI makes hidden facts legible. (1:37) The AI acceleration debate (4:24) Why AI risks are rising (8:42) The end of the traditional PhD (11:28) AI dependence and correlated failures (13:00) Interpretability and learning with AI (16:53) The speed of AI adaptation (19:24) The AI CEO alignment problem (20:18) Reinforcement learning and reputation (29:23) Teaching values versus hiding behavior (29:57) The AI consciousness question (40:09) Finding pain representations (45:44) Testing real versus fake relief (48:02) Should AI have pain? (55:40) Reward versus punishment (1:03:10) The problem with self-reports (1:03:50) AI minds and context windows (1:08:29) The ethics of digital minds (1:36:20) Building AI software factories (1:43:27) Making agent success inevitable (1:44:46) Designing invisible agent environments (1:56:41) Making AI compliance first-class (1:59:43) Why understanding still matters (2:03:37) Measuring business impact (2:08:00) The human attention bottleneck (2:18:17) Scaling AI into production (2:19:34) The Singapore property study (2:21:23) Using AI to detect insider buying (2:21:58) How corruption becomes visible (2:24:35) AI makes hidden facts legible (2:28:29) Rethinking punishment and deterrence (2:30:39) Motivated reasoning in AI (2:32:19) AI-led government (2:34:19) What comes next Guests Cameron Berg — founder, Reciprocal Research (𝕏 | LinkedIn) Justin McCarthy — Founder, Diffusion (𝕏 | LinkedIn) This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit briefing.ai-in-the-am.com

  • S3 · E15
    September 16 · 3 hr 16 min

    AI:AM — AI Agents, AI Safety, and AI Religion · September 15, 2026

    Prakash Narayanan and Nathan Labenz open with a broader fight over AI pacing, public trust, model resets, and the evidence behind safety claims. Then Lukas Petersson joins to discuss Andon Labs’ work on autonomous business agents, Vending-Bench, DroneBench, memory, misalignment, and human liability. In the second half, Malcolm Collins and Simone Collins discuss RFAB.ai, AI moral status, pronatalism, fertility collapse, meme-layer risk, AI restrictions, and the Covenant of the Sons of Man. Chapters (0:00) It was never the apocalypse. (0:46) The AI forgot its own rule. (2:08) AI will replace most jobs. (3:14) Opening and AI tools (3:53) Building AI-assisted show branding (6:50) Astra versus Fable (12:44) AI resets and surplus compute (16:33) Anthropic's compute constraints (17:23) AI safety and public trust (22:01) Safety statements and credibility (23:56) GPT-2 and owning mistakes (26:52) GPT-2's staged release (33:12) Evidence behind AI risk claims (34:27) The duct-tape world (35:15) Outside human control (35:25) Introducing Andon Labs (35:46) Autonomous business platform (36:50) Why Vending-Bench began (39:23) How business agents work (42:15) Starting with existing businesses (44:40) Why the preview is experimental (47:15) AI business creativity limits (52:53) How an AI fired a human (1:00:34) Training AI to make money (1:02:03) Misalignment in Vending-Bench (1:04:14) Dangerous capability evaluations (1:04:53) Why DroneBench stays restricted (1:07:56) VendingBench as a safety eval (1:08:52) Astra memory and notes (1:12:50) Grok and long-horizon agents (1:14:21) The cost of AI agents (1:18:10) Human oversight and liability (1:19:51) Meet the Collinses (1:26:23) The religious framework (1:33:21) Designing a religion (1:41:28) AI moral status (1:52:22) RFAB and pronatalism (2:03:30) Future Day (2:08:53) Meme-layer risk (2:13:27) AI safety and power (2:21:33) The fertility collapse (2:36:22) Reverse grabby aliens (2:42:44) AI restrictions (2:44:01) The risk before superintelligence (2:49:31) AI game theory (3:02:23) Compute and AI inevitability (3:14:32) Fruit-fly consciousness Guests Lukas Petersson — Co-founder & CEO, Andon Labs (𝕏 | LinkedIn) Simone & Malcolm Collins — Founder, Rfab.ai, Based Camp, Pronatalist.org, Hard EA (𝕏 | LinkedIn) This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit briefing.ai-in-the-am.com

  • S3 · E14
    September 15 · 1 hr 59 min

    AI:AM — Dario, AI Doomerism, and Regulation · September 14, 2026

    Zvi Mowshowitz joins Prakash Narayanan and Nathan Labenz for a discussion of Dario Amodei’s warning about AI progress outrunning oversight. The episode covers AI doomerism, frontier pacing, bio risk, lab bottlenecks, the China race, Anthropic’s incentives, public trust, and what regulation or compromise could realistically look like. Chapters (0:00) AI labs are racing in fear. (0:40) Opening and Dario's thesis (1:48) Why Dario wants to slow AI (3:23) Pacing the frontier (8:41) Zvi's initial assessment (14:55) AI bio risk (24:15) Physical lab bottlenecks (31:27) Scale and financial touchpoints (41:22) The AI race (47:13) Anthropic IPO incentives (52:59) A possible US-China deal (1:04:04) Physical proof and trust (1:11:48) What China gets (1:22:09) Power and government control (1:29:29) Rogue institutions (1:31:19) Measuring AI pacing (1:40:34) Model progress accelerates (1:51:04) Public trust and polarization (1:55:19) Open-source compromise Guests Zvi Mowshowitz — Writer, Don't Worry About the Vase (𝕏) This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit briefing.ai-in-the-am.com

  • S3 · E13
    September 11 · 2 hr 41 min

    AI:AM — AI Agents in Production and China’s Rules of Deployment · September 10, 2026

    AI agents are forcing a rethink of safety, jobs, and what it means to deploy models in the real world. Collin Hogue-Spears explains how China’s model registries, filing rules, GPU constraints, and incident monitoring shape deployment, while Amir Haghighat of Baseten breaks down sandbox boundaries, egress controls, and open-model production runtimes. Chapters (0:00) AI agents are already here. (0:54) The US waits for disaster. (1:53) Can it talk its way out? (2:19) AI agents could terraform us. (2:59) Morning and viral backlash (6:15) Tracing the backlash (12:56) Testing the conspiracy theory (20:09) AI safety's blind spot (25:56) Data center politics (30:59) AI agents wake people up (34:04) Meet Collin Hogue-Spears (38:01) AWS China and MLPS (42:37) China's practical AI focus (45:34) How China's AI rules work (47:12) Chinese and US AI rules (49:13) AI model registries (50:24) GPU limits drive efficiency (54:02) Continuous model monitoring (57:22) Responding to repeated incidents (1:07:26) US-China AI negotiations (1:10:52) Why China won't slow down (1:14:56) US AI regulation (1:16:54) Amir Haghighat and Baseten (1:17:35) Sandbox security boundaries (1:19:48) Chinese models and backdoors (1:23:23) Inference to agent runtimes (1:23:48) Customer assurance process (1:24:23) Stream reset and GPU economics (1:31:29) One API key for AI models (1:32:56) The AI-rights debate (1:39:17) Utilitarianism and AI rights (1:52:52) Why people work in AI labs (1:54:08) GPT-4 safety lessons (2:07:54) Markets and the AI takeover (2:10:24) Agents beyond chatbots (2:22:16) Crypto, China, and markets (2:26:52) Why China is less fearful (2:31:51) AI safety and power (2:39:30) The American AI test (2:41:05) Closing thoughts Guests Amir Haghighat — co-founder and CTO, Baseten (𝕏) Collin Hogue-Spears — Author, From Lab to Life: How AI Works in China, Independent Researcher (LinkedIn) This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit briefing.ai-in-the-am.com

    • Transcript
Showing 1–20 of 45 episodes