AI:AM

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 accelerationGuestsBryan 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

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 damageGuestsMike 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

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 backlashGuestsMichael Freeman — Senior Director of Threat Research, ServiceNowSagar 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

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?GuestsAnastasios 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

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 SaaSGuestsEvan 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

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 fallbackGuestsZvi 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

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 waveGuestsCameron 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

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 freeGuestsEverett 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

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 productionGuestsEdward 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

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 expectedGuestsThomas 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