PJFP.com

Pursuit of Joy, Fulfillment, and Purpose

Tag: OpenAI IPO

  • Jensen Huang Says the AI Apocalypse Is ‘Complete Nonsense’: NVIDIA’s CEO on AI Jobs, China, Open Source Models, the AI Bubble, and the Trillion-Agent Future (Axios Behind the Curtain)

    Sitting on the floor of a brand new chip factory in Fort Worth, Texas, NVIDIA CEO Jensen Huang gave Axios reporter Mike Allen one of his most combative and quotable interviews yet. In this episode of Behind the Curtain, the head of the world’s most valuable company dismisses AI doom scenarios as “complete nonsense,” argues that AI is creating jobs rather than destroying them, defends Chinese open source models like Kimi and DeepSeek, explains why the AI build out is not a bubble yet, and calls for Anthropic’s most powerful model to be made available to everyone.

    TLDW

    Huang covers the full sweep of the AI moment: Chinese export control threats and why he wants open research flows in both directions, why the world needs both closed models (Anthropic, OpenAI) and open models (Kimi, Qwen, DeepSeek, NVIDIA’s own Nemotron), why Wall Street misread the Kimi selloff exactly as it misread DeepSeek, the sovereign AI argument that no company or country should “outsource its alpha,” his evidence that AI is increasing jobs for radiologists, paralegals, and manufacturing workers, a sustained attack on AI doomers and the “made up” narratives of singularity, simulation, and machine consciousness, the CapEx-heavy economics of manufacturing intelligence via tokens, his claim that the bubble is not coming in the next five years because physical constraints (chips, memory, power, construction workers) are pacing the build out, his warm relationship with President Trump and his warning against knee-jerk regulation, his position that Claude Mythos should be available to all users, the coming era of a trillion AI agents, the “ChatGPT moment” for robots having already arrived, and closing life lessons on pain, suffering, practice, immigration, and why he refuses to wear a watch because “now is the most important time.”

    Thoughts

    The first thing to hold in mind while watching this: every single position Huang takes, without exception, maps to selling more GPUs. Open models are good (more diffusion, more compute). Closed models are also good (more services, more compute). Chinese models are good (more use, more compute). Doom talk is bad (fear slows adoption, which slows compute). The bubble is far away (keep buying compute). That perfect alignment between worldview and order book does not make him wrong, but it means his arguments deserve scrutiny on the merits rather than deference to his position. He is the most effective anti-doomer in the industry partly because he is the person with the most to lose if the world gets scared.

    That said, his strongest material is empirical, and it lands. The radiologist example is a direct rebuttal to one of the most famous predictions in AI history, Geoffrey Hinton’s 2016 claim that we should stop training radiologists. Huang’s version of events, that automating the scan-reading task let radiologists see more patients and demand for them grew, is a textbook case of what economists call the Jevons effect applied to labor. Whether his specific numbers (20 percent more radiologists, 10 percent more paralegals, 50 percent more manufacturing jobs) survive fact-checking, the structural argument that automating a task can grow the profession around it is historically well supported, and it is the single most useful reframe in the interview: your job is not your task, and when the task gets automated, the purpose remains.

    The open source security argument is the most intellectually serious part of the conversation and the one most directly aimed at his own customers. Huang praises Anthropic and OpenAI as businesses in one breath and then dismantles the “closed models are safer” position in the next: Linux runs the world’s digital infrastructure precisely because millions of people can inspect and harden it, and a world defended by one closed model is a world with a single point of failure. His call for “massively distributed, diverse defense” via open models in the hands of cybersecurity experts everywhere is a real policy position with real stakes, and it puts him closer to Meta’s historical stance than to the labs he supplies.

    The bubble section is where the skeptic should lean in. Allen hands him the most famous cursed phrase in financial history, “this time is different,” and Huang takes the bait enthusiastically: it is different, he says, because the demand is industrial rather than cyclical. Every bubble in history was justified by exactly this argument, including the railroads and the dot-com fiber build out that Huang implicitly invokes as precedent. But his supply-side observation deserves weight: bubbles pop when supply overshoots demand, and right now everything (chips, memory, packaging, power, land, construction labor) is short. A market that cannot build fast enough is at least not overbuilt yet. His own concession that “the bubble will come someday” and his refusal to vouch for years five through ten is more honest than the rest of the answer.

    Finally, notice the tension he never resolves. He says warnings about AI’s power are “well heeded,” that safety is the leaders’ responsibility, and that Anthropic must fix jailbreaks fast. He also says consciousness, singularity, and existential risk are “all made up,” and shrugs off the referenced Mythos jailbreak with “everything was fine, you and I are here having a conversation.” Those two postures, take the technology seriously enough to harden it but never seriously enough to fear it, are held together mostly by confidence. It is a bet that capability and controllability scale together. The doomers he mocks are making the opposite bet, and nothing in this interview actually settles which one is right.

    Key Takeaways

    • On reports that Chinese regulators may tighten export controls on AI models and semiconductors to keep them from the West: Huang hopes it does not happen, notes half the world’s AI researchers are Chinese, and says both sides should de-escalate and let the technology advance.
    • He opposes any US ban on Chinese models like Kimi: American companies should absolutely be allowed to use them, because downloaded open models can be fine-tuned, guardrailed, and run inside secure sandboxes and harnesses, and the “back door” fear is a misconception.
    • The world needs both closed and open models: use closed services (Anthropic, OpenAI) as much as possible because they are excellent and convenient, but science, cybersecurity, and sovereignty require open models.
    • Regulate applications of AI (medicine, transportation, autonomous vehicles), not the underlying technology, which is dual use and should advance as fast as possible.
    • NVIDIA’s China sales are “approximately zero today” and he has told investors to expect none; he would consider it an honor to return if both governments allow it.
    • The market misunderstood DeepSeek and is now misunderstanding Kimi the same way: great open models, wherever they come from, drive more AI use, which drives more NVIDIA computers, more data centers, and more services.
    • Open models are not adversarial to closed models: the most likely customer to upgrade to Anthropic or OpenAI is someone who already uses AI and wants it more convenient and better.
    • NVIDIA’s Nemotron open model exists for companies that must build their own AI for sovereignty, regulatory, privacy, or IP reasons. “We don’t have to be the frontier. We have to be at the frontier.”
    • The large language model is the brain; a harness (he names OpenClaw and Claude Code as examples) turns it into a working agent. With the right harness, Nemotron can be world-class for specific skills.
    • Cheap or free open source tokens are “fantastic” for the proprietary labs: free AI grows the population of people who realize they need AI, and running even a free model yourself usually costs more than renting a service.
    • Echoing the viral Palantir CEO interview: “Nobody should outsource their alpha.” Companies and countries should rent AI wherever they can but must build their own AI for domain-specific, proprietary, sovereign, secret, or regulated work.
    • For non-differentiating work (marketing automation, legal department productivity), outsource to the frontier labs as much as possible.
    • Nothing AI has done has truly surprised him; what society needs to realize is that automating tasks is increasing the number of jobs the world needs.
    • His jobs evidence: radiologists up roughly 20 percent because AI-automated scan reading lets them see far more patients; paralegals up roughly 10 percent for the same reason; US manufacturing jobs up roughly 50 percent in recent years because AI data centers require industrial might.
    • On the demonstrated ability of Anthropic’s Mythos to break into hardened systems: “it surprised me that people were surprised.” An AI that can write and debug software can necessarily find vulnerabilities; the same capability powers cyber defense.
    • His security architecture argument: one single model is one single point of attack and failure. Open models in the hands of cybersecurity experts worldwide create “massively distributed, diverse defense,” the same reason Linux is trustworthy.
    • Whether China has “caught up” does not matter: the race-with-a-finish-line framing is wrong, China manufactures more AI researchers than the rest of the world combined, holding China back is ill-conceived, and neither side can hold back the other.
    • “AI is not going to destroy all of our jobs. Someone who uses AI is going to take our jobs.” The biggest risk to the US is scaring industries and society out of adopting AI.
    • On doomer AI CEOs: warning is fine, warning with a solution is better, and making things up is “absolutely inappropriate.” End-of-humanity and half-of-jobs-destroyed claims are “complete nonsense” contradicted by all the evidence.
    • Asked why Asia loves him while America is anxious: “the doomers spend too much time theorizing about these science fiction outcomes, maybe it makes them sound smart.”
    • OpenAI and Anthropic are not in trouble from Chinese competition: “zero possibility” China runs US companies off the road, both labs are thriving, and their IPOs will be the most successful in human history.
    • On chip stocks down 18 percent after Kimi dropped: free AI is great for hardware, chips, and data centers; the market got it wrong with DeepSeek (NVIDIA fell about 30 percent) and is getting it wrong again.
    • AI cannot have peaked because diffusion into society and industry has barely begun; useful AI has finally arrived, and useful AI is profitable AI, citing coding agents companies happily pay hundreds of millions a year for.
    • The new IT industry is CapEx heavier than software because intelligence must be manufactured: machines produce the tokens behind every answer, image, protein, and robot maneuver, and the resulting productivity will more than pay for the build out.
    • A token is an embedding of knowledge and intelligence, and unlike pi it gets smarter over time; smarter tokens are more valuable, which is why token economics keep improving.
    • On the bubble: “The bubble will come someday. It’s just not today.” Very unlikely in the next five years; five to ten years depends on how fast the industry can build.
    • The build out is constrained in every direction (chips, memory, land, power, construction workers), and that constraint is healthy: it pushes out the day supply exceeds demand.
    • This cycle is “industrial-driven,” not seasonal or consumer-demand-driven: the world needs a new intelligence infrastructure layer on top of energy, internet, roads, and railroads, and the semiconductor industry needs to be 5 to 10 times larger within ten years.
    • He is not worried about customers issuing hundreds of billions in debt to buy his chips: these companies generate enormous cash, the compute platform shift is real, and the ROI question has been answered because AI is now demonstrably profitable.
    • He would use Kimi himself, with fine-tuning, guardrails, sandboxing, and access control, the same way the world already trusts open source software like Linux.
    • On Trump: they text, the president “remembers everything” including H20, H200, Blackwell, and Rubin, and the Fort Worth factory they are sitting in is a direct result of their first conversation about reindustrializing America.
    • His warning to the administration: do not over-correct based on science fiction narratives about AI consciousness; talk to many CEOs and scientists, not one or two, and take time to be informed before regulating.
    • On the government taking an equity stake in NVIDIA: unnecessary, because the US already has a stake via $10 billion in taxes paid last year, job creation, and the stock market holdings of most Americans.
    • Claude Mythos should “absolutely be available to everyone,” not just selected institutions; it is Anthropic’s job to harden it and patch jailbreaks fast, and he notes that when it was jailbroken “everything was fine.”
    • On distillation of closed models: learning from other intelligence is fundamental (soon the internet will be 99 percent AI-generated content anyway), but violating terms of service or privacy is not okay and should be handled through existing legal channels.
    • NVIDIA has 6,500 employee families in Israel he is concerned for; he remains bullish on the UAE reinventing itself from an oil economy into an AI hub.
    • NVIDIA runs about 50,000 employees and may reach only 75,000 in ten years, “as small as possible,” because strategy means maximizing impact per unit of resource.
    • Jobs that are a single task (customer service call centers) will be automated; jobs with purpose survive because purpose does not change when the task is automated. “Don’t mistake your task for the job.”
    • In 10 to 20 years, photos of people typing at keyboards will look like old photos of typing pools with IBM Selectrics: typing was never the job, solving problems and creating value was.
    • The ChatGPT moment for robots has already arrived (a robot can reason through “put the apple in the drawer,” including opening the drawer first); useful robots in ordinary life within 3 to 4 years would not surprise him.
    • The agentic era’s capability has arrived and diffusion is next: the future holds 100 billion to a trillion agents running constantly, and agents will not become computers, they will use computers, which is why compute demand explodes.
    • $300 billion has been invested into US venture capital startups in the last six months, and he tells his nieces and nephews that great fortunes will be created on a laptop.
    • Life lessons: greatness requires “plenty of pain and suffering” and practice when nobody is watching; under maximum stress, time slows down the way athletes describe, and that comes from repetition.
    • He advises every bright mind in the world to come to America, the country built by immigrants that will need amazing immigrants in the future.
    • He wears no watch and refuses to let Outlook manage his life: “now is the most important time.” His perfect Saturday: dogs, work, family dinner, a cocktail, and he notes every weekend is exactly like that.

    Detailed Summary

    Export Controls Cut Both Ways

    The interview opens on a Financial Times report that Chinese regulators are considering export controls of their own, restricting Chinese AI models and semiconductors from reaching the West. Huang’s response is de-escalation in both directions: half the world’s AI researchers are Chinese, groundbreaking research flows from both countries, and once one side reaches for export controls, everyone starts thinking in those terms. He is confident the US will continue to lead as long as government supports rather than constrains its companies. Asked whether the US should ban Chinese models like Kimi, he rejects the premise: downloaded open models run inside harnesses and sandboxes with security, privacy, and access controls, and the idea of hidden back doors phoning home to China is a misconception. His China sales, he notes pointedly, are approximately zero today, so his position is not about protecting revenue he does not have.

    Open and Closed Models Both Win

    Huang’s framework is consistent: rent closed models (Anthropic, OpenAI, which he personally uses along with Perplexity) whenever you can because they are excellent and convenient, and build on open models only when you must, for sovereignty, regulation, privacy, or proprietary domain reasons. This is the pitch for NVIDIA’s own Nemotron open model family, which he positions not as a frontier competitor but as raw material for companies that need custom AI: “We don’t have to be the frontier. We have to be at the frontier.” He describes the modern stack in plain terms: the large language model is the brain, and a harness (he cites OpenClaw and Claude Code) turns it into a working agent. Open, cheap, and free models are on-ramps that grow the total population of AI users, which is why he insists the labs should not fear them: the person most likely to pay for Claude is someone already using AI who wants it better and easier.

    Kimi, DeepSeek, and Wall Street’s Repeated Mistake

    Chip stocks fell 18 percent in the month after Kimi dropped, echoing the roughly 30 percent NVIDIA drawdown when DeepSeek landed. Huang says the market got it wrong both times and for the same reason: free and open AI is great for hardware, because great models drive use, use drives data centers, and data centers drive chips. He runs through the models he considers extraordinary (Kimi 3, Qwen, Nemotron, GPT 5.6, Codex, Claude Code) and lands on his core claim about this moment: useful AI has finally arrived, and useful AI is profitable AI. Companies like NVIDIA happily pay hundreds of millions of dollars a year for coding agents doing high-value work, which funds more AI, which he describes as a flywheel that has now started.

    Don’t Outsource Your Alpha

    Allen raises the viral Palantir CEO warning about handing your intellectual property to frontier labs, noting Huang’s unique position as both a top customer and top supplier of those labs, including using their models for chip design. Huang agrees with the principle without hesitation: nobody, no company, no country should outsource its alpha or its intelligence. His dividing line is specificity: work that is domain-specific, proprietary, sovereign, secret, or regulated must be done in-house on your own models, while generic productivity work like marketing automation or legal department support should be outsourced to the labs as aggressively as possible. The same logic scales to nations, which he says cannot outsource their fundamental intelligence to a third party.

    The Jobs Evidence

    Asked what AI has done that scared or awed him, Huang says essentially nothing surprised him, including the demonstrated ability of Anthropic’s Mythos to penetrate hardened systems (“it surprised me that people were surprised,” since an AI that debugs software can obviously find vulnerabilities). What he wants the world to notice instead is the labor data. Radiology reading has been substantially automated, and the number of radiologists is up roughly 20 percent because they can now see the enormous backlog of patients. Paralegals are up roughly 10 percent by the same mechanism. Manufacturing jobs are up roughly 50 percent in recent years because AI data centers require industrial construction. His formulation of the real risk: AI will not take your job, someone who uses AI will, and the worst thing America could do is scare its own industries out of adopting the technology.

    Against the Doomers

    This is the section that gives the interview its title. Huang says warning people is fine, warning with a solution is better, and making things up is absolutely inappropriate. The end of humanity: complete nonsense. Half of American jobs destroyed: complete nonsense. The singularity, living in a simulation, machine consciousness: “all made ups,” fun science fiction he enjoys hearing from “many of those leaders and my friends,” but Hollywood, not ground truth. Asked why he is mobbed by fans in Asia while the American mood is hostile, he suggests the doomers theorize about science fiction outcomes because “maybe it makes them sound smart.” His prescription for the industry is to tell the factual story, that AI is creating millions of jobs, rather than a made-up narrative that frightens the public and, more dangerously in his view, frightens policymakers. His closest thing to a concession: the closest thing to true AI is R2-D2 and C-3PO, “and who doesn’t want R2-D2 and C-3PO?”

    CapEx, Tokens, and the Bubble Question

    Huang’s economic argument for the build out runs through the token. Unlike the CapEx-light software era, intelligence must be manufactured: machines generate the tokens behind every answer, every image, and eventually every protein, chemical, and robot movement. A token is an embedding of knowledge, and unlike a static number it gets smarter over time, which makes it more useful, more valuable, and worth paying more for. On the bubble, he does not deny one is possible: “The bubble will come someday. It’s just not today.” He rules it out for roughly five years and hedges on five to ten. His reasoning is that this cycle is industrial-driven rather than consumer-cyclical: the world is adding an intelligence layer on top of energy, internet, roads, and railroads, the semiconductor industry needs to be 5 to 10 times larger within a decade, and everything (chips, memory, optical interconnects, packaging, TSMC capacity, land, power, construction workers) is short. Those constraints pace the CapEx and push out the day supply overtakes demand. As for customers issuing hundreds of billions in debt to buy his chips, he says the companies are extraordinary cash generators and the ROI question has been settled by profitable coding agents.

    Trump, Washington, and the Over-Correction Risk

    Huang describes a genuinely warm relationship with President Trump: they text, the president remembers chip model numbers (H20, H200, Blackwell, and next-generation Rubin), and the Fort Worth factory hosting the interview traces directly to their first conversation about restoring American manufacturing. He praises Susie Wiles, Secretary Bessent, and Secretary Lutnick. But his message to the administration is a warning: signs point toward more restrictive AI policy, and he fears policymakers falling for science fiction narratives (consciousness, an imminent finish line in a US-China race) pushed partly by companies hoping regulation will advantage them. His advice: talk to many CEOs and scientists, not one or two, take time, and do not over-correct. He rejects the 100-meter-dash framing of the China race entirely, arguing the win is diffusion, not invention: America did not invent electricity or manufacturing, it applied them with more enthusiasm than anyone, and that is what made the country. Asked about the government taking equity stakes in AI companies, he calls it unnecessary: the US already holds a stake in NVIDIA through $10 billion in annual taxes, job creation, and the stock market.

    Mythos for Everyone, and the Distillation Question

    In the most newsworthy exchange, Allen asks whether the world is ready for Anthropic’s most powerful model, Claude Mythos, to be available to everyone rather than selected institutions. Huang’s answer is unambiguous: it should absolutely be available to everyone, it is Anthropic’s responsibility to harden it, and jailbreaks are the nature of software, to be patched as fast as they are found. He points to the referenced jailbreak incident and observes that “everything was fine,” while noting that holding Anthropic back serves no American interest, especially since open models are available regardless. On distillation, he splits the question: AIs learning from other AIs is fundamental and inevitable (within a few years, he predicts, the internet will be 99 percent AI-generated content, so every model is distilling other AIs anyway), but violating terms of service or privacy is not acceptable, and aggrieved providers should pursue the conventional legal remedies that already exist.

    Robots, Agents, and the Next Era

    Huang argues the ChatGPT moment for robots has already happened, on his definition: the 2022 ChatGPT moment was not when AI became useful (that took four more years) but when it did something surprising, and a robot that can reason through “put the apple in the drawer,” including opening the drawer first, clears that bar today. Useful everyday robots within three to four years would not surprise him. On the agentic era, capability has arrived and diffusion is what comes next: where perhaps 100 million humans use computers at any given moment today, the future holds 100 billion to a trillion agents of every kind running constantly. His line: agents are not going to become computers, agents are going to use computers, and that is the deepest driver of compute demand.

    Life Lessons from 33 Years at the Helm

    The closing stretch turns personal. On keeping NVIDIA at roughly 50,000 employees (maybe 75,000 in ten years, “as small as possible”) while peers run six figures, he says strategy is using limited resources with maximum precision, a craft he has practiced longer than any CEO in tech history: “this is my kung fu.” On which jobs disappear, he distinguishes task from job from purpose: call center tasks will be automated, but a radiologist’s purpose (ending human suffering) survives the automation of scan reading, and typing was never the job in the first place. Born in Taiwan and sent to a rough American boarding school at nine, he calls America the greatest country in the world because open discourse and freedom let it work through its disagreements, and he urges bright minds everywhere to come. On greatness: no athlete just happens to be great, it is practice when nobody is watching, setbacks, losing, and “plenty of pain and suffering” that elevate craft, character, and resilience. He wears no watch because now is the most important time, and his perfect Saturday (dogs, work, family dinner, a cocktail) is, he says, exactly what every weekend already looks like.

    Notable Quotes

    “And so the fact that this is going to be the end of humanity, it’s complete nonsense. The fact that this is going to destroy half of the American jobs. It’s complete nonsense. And all of the facts, all of the evidence point exactly to the opposite.”

    Jensen Huang, on AI doom predictions from fellow tech leaders

    “AI is not going to destroy all of our jobs. Someone who uses AI is going to take our jobs, and so we have to make sure that we adopt AI, diffuse AI into the industries as quickly as possible.”

    Jensen Huang, on the real employment risk of the AI era

    “Nobody should outsource their alpha. Nobody should outsource their intelligence. No country should.”

    Jensen Huang, agreeing with the Palantir CEO’s warning about handing IP to frontier labs

    “We don’t have to be the frontier. We have to be at the frontier.”

    Jensen Huang, on NVIDIA’s Nemotron open source model strategy

    “The bubble will come someday. It’s just not today.”

    Jensen Huang, on whether the AI build out is a bubble

    “It is made up that there’s going to be a singularity. It’s made up that somehow we’re living in a simulation. These are all made ups.”

    Jensen Huang, on science fiction narratives he says are scaring the public and policymakers

    “The closest thing to true AI is R2-D2 and C-3PO. And who doesn’t want R2-D2 and C-3PO?”

    Jensen Huang, on how to inoculate the public against fear of AI

    “These two companies will be the most successful IPOs in human history.”

    Jensen Huang, predicting the public debuts of OpenAI and Anthropic

    “If your job is the task, then it’s very likely that when that task is automated, your job will be eliminated or changed.”

    Jensen Huang, on which jobs disappear in an industrial revolution

    “Because now is the most important time. I refuse to let Outlook manage my life, and I refuse to let a watch manage my life.”

    Jensen Huang, on why he does not wear a watch

    Watch the full conversation between Jensen Huang and Mike Allen on Axios Behind the Curtain here.

    Related Reading

  • Tim Ferriss and Kevin Rose Random Show: Mortality and Grief, Zen Insights, Rock Climbing at 50, LSD for Anxiety (MM120), AI Smart Homes, and Why You Should Buy the Company Instead of the Product

    Tim Ferriss and Kevin Rose reunite over tequila for another Random Show, and this one swings from the heaviest material they have covered in years (the death of their friend Om Malik, aging parents, dementia, and what grief actually is) to Zen retreat breakthroughs, rock climbing as a post-50 obsession, a phase 3 LSD trial for anxiety, AI-powered smart homes, the coming wave of AI IPOs, and the single investing lesson both keep relearning: let your winners run, and when you love a product, buy the company.

    TLDW

    Kevin reframes the loss of Om Malik and his father through a simple equation: grief is love with nowhere to go, and the sorrow is proof of how lucky you were. Tim adds Tim Urban’s “The Tail End” math (you have spent roughly 95% of your lifetime hours with your parents by high school graduation) and Sam Harris’s “The Last Time” meditation. Kevin recounts a micro-awakening at a five-day silent Zen retreat (“nothing lacking”), both plug their meditation app The Way with Henry Shukman, and Tim declares multi-pitch climbing in Yosemite his next deliberate-practice obsession, complete with hangboard protocols and grip-training gear. The health segment covers A2 whey, venison organ-meat sticks as a multivitamin, the 1,3-butanediol ketone controversy, ketones temporarily unlocking speech in relatives with dementia, terminal lucidity, a JAMA phase 3 trial of MM120 (lysergide) showing 12 weeks of anxiety relief from a single dose, and the Norwegian 4×4 protocol whose hippocampal benefits may persist for five years. The AI segment runs from Kevin’s Claude-coded camera system that opens his gate via license plate recognition, to Tim’s 20-year angel investing retrospective built with Claude Code and the Gmail API, to their handicapping of Google versus Anthropic versus OpenAI, China’s open-source push, local inference boxes, and why buying at IPO and holding may match venture returns.

    Thoughts

    The emotional spine of this episode is the best thing in it. Kevin’s formulation, that the gap left by a death “is just love at the end of the day,” is not new philosophy, but it lands differently coming from someone actively managing a dying dog, a mother with dementia, and a friend’s fresh death, all in the same month. The practical corollary the two keep circling is time-boxing: Tim Urban’s Tail End math and Sam Harris’s “last time” framing both convert vague mortality awareness into a scheduling problem. Tim credits one short blog post with causing years of family trips that his emotionally reserved family would never have taken otherwise. That is about as strong an endorsement as content can get: it changed the calendar, not just the mood.

    The health middle of the show is classic Random Show in that the interesting part is the epistemology, not the products. Tim flags that the loudest critics of 1,3-butanediol ketones sell competing ketone salts, applies a shelf-life heuristic to processed meat instead of memorizing ingredient lists, and treats organ-meat sticks as a dosed multivitamin rather than a diet. The MM120 discussion is the meatiest science: a five-arm randomized trial where a single 100 microgram dose of lysergide produced roughly twelve weeks of relief in generalized anxiety disorder, which Tim, who has been diagnosed with GAD and OCD, reads as a plausible future where anxiety treatment is episodic rather than daily. The unresolved tension they name honestly: the promising dementia signals (ketones, psilocybin case reports, microdosing) all crash into the consent problem. A person who cannot consent cannot sign up for a hallucinogen, and “it might give you half a day of real conversation back” is both a miracle and an ethical minefield.

    The AI section quietly contains one of the more useful predictions frameworks going: Kevin’s argument that Google’s confusing high-bandwidth TPU architecture only makes sense as a bet on continuous learning, where models stop shipping as discrete releases and start improving around the clock like a child. If self-improving models are really 12 to 18 months out, the “model drop” news cycle this episode itself participates in (new Sonnet today, Mythos tomorrow) is a temporary artifact. Tim’s counterweight is human-scale and more sobering: an AI trained on your own writing produces in 30 seconds what takes you 30 hours, and he compares the demoralization to Lee Sedol retiring after AlphaGo. His book sales chart, stable for a decade and then compounding downward every year since ChatGPT launched, is the receipts. The tension between “AI made my 20-year retrospective possible” and “AI is draining my motivation to write” is the honest version of the AI discourse most podcasts flatten into one direction.

    The investing segment is the most immediately actionable. Three ideas stack neatly: let winners run (Tim has lost more money selling early than he made buying), the venture-returns myth (a famous firm’s own analysis found that buying at IPO and holding a decade roughly matched their gains from early rounds through lockup), and buy-what-you-use (the friend who spent $100k on a top-of-the-line Tesla instead of Tesla stock forfeited roughly $15 million; teenage Tim bought Pixar after seeing Toy Story). None of this is sophisticated, which is the point both make explicitly: with Anthropic and OpenAI racing to IPO, ordinary people who use these tools daily will get a shot the private markets never gave them, and the discipline that matters is holding, not access.

    Key Takeaways

    • Kevin and Tim lost their friend and colleague Om Malik of True Ventures within the past week; Kevin found out mid-way through a five-day silent meditation retreat.
    • Kevin’s reframe on grief: the severe sense of loss is “just love at the end of the day.” The gaping hole his father’s death left is love manifested through sorrow, and recognizing that converts anguish into gratitude for having crossed paths at all.
    • Tim credits Matt Mullenweg twice: for organizing the Antarctica trip where he got days of uninterrupted time with Om (including a visit to an emperor penguin colony), and for sending him Tim Urban’s blog post “The Tail End.”
    • The Tail End’s core math: by high school graduation you have used up roughly 90 to 95% of the total in-person hours you will ever spend with your parents. Reading it drove Tim to organize regular family trips, awkwardness be damned, before his father’s mobility declined.
    • Sam Harris’s short meditation “The Last Time” pairs with it: for many activities you will do a last time without knowing it was the last time.
    • Kevin’s 15-year-old dog Toaster had a violent shaking episode (a stress syndrome after standing six hours at a vet visit, not a terminal event), and Kevin’s takeaway from being covered in the aftermath was that when you love an animal that much, none of it matters.
    • At a traditional Zen sesshin with Henry Shukman and his visiting Japanese teacher Yamada Roshi, Kevin had a two-second micro-insight while working his koan: a felt sense of “nothing lacking,” where nothing could be added or taken away because everything was already fully present. Not an emotion, a steady state.
    • Both are investors in The Way, Henry Shukman’s single-path guided meditation app, which they frame as an ideological investment like their funding of the dog aging study on rapamycin. Tim’s favorite sessions: “Whole Earth is Medicine” and “This Too is Me.”
    • Tim’s practical meditation pitch: you do not need a retreat; 10 minutes twice daily works, and there seems to be real alchemy in the twice-a-day rhythm. Kevin, once the guy who quit everything in two weeks, is coming up on five years of consistent practice.
    • A physiology aside: Henry’s instruction to drop the jaw slightly mirrors how Tim’s mandibular snoring device works. Dropping the jaw an eighth of an inch down and forward opens the airway. The ancients found it by trial and error.
    • Kevin, approaching 50, wants to stop saying “one day” about his bookmarked obsessions (Japanese woodworking, ships in bottles) and actually commit to things in the next two decades.
    • Tim’s next deep dive is rock climbing: his surgically repaired right elbow finally allows it, and his stretch goal is multi-pitch climbing in Yosemite despite being, in his words, deadly terrified of heights, sweaty palms included.
    • Tim’s philosophy of training: “training to not die sooner than is necessary” is not a sufficient goal. He needs a concrete deadline event, the way the Lancaster Classic structured his archery, to make deliberate practice worth it.
    • What sold Tim on climbing longevity: the 60-to-almost-80-year-olds at Salt Lake City gyms climbing 5.11+ on weekday mornings, out-performing what he could imagine doing, plus women who cannot do five pull-ups climbing 5.13 and 5.14 on pure technique.
    • Climbing is also social in a way archery never was: bouldering routes are literally called “problems,” and strangers trade beta. After decades of solitary repetition, Tim has hit his quota.
    • Training tools discussed: Michael Eckert’s finger-strength course (the multiple-time pull-up world champion Kevin just bought into), the Nug (a pocket-size wooden grip trainer Tim travels with), and Abrahangs, Emil Abrahamsson’s protocol of moderate partial-bodyweight hangs, 10 seconds on and 50 seconds off for 10 minutes twice a day, which produces outsized forearm and finger gains.
    • Tim’s fantasy recommendation: The Blade Itself, whose treatment of the randomness of death (a friend of Tim’s just died in a plane crash) doubles as a gratitude practice. The audiobooks are exceptional.
    • Protein talk: Kevin likes Pioneer Pastures A2 whey (30 grams a shake, lactose removed, no investor relationship); Tim gets roughly 40% of his protein from Maui Nui wild-harvested axis deer venison and treats the liver-and-heart pepper sticks as a two-or-three-a-week multivitamin.
    • On processed meat and nitrates, Tim’s heuristic is shelf life: if an ultraprocessed meat lasts three years on a shelf, raise an eyebrow. Minimally processed meat almost definitionally does not keep.
    • Exogenous ketones containing 1,3-butanediol may carry liver toxicity risk, though Tim notes many people pushing that claim sell competing ketone salts. His personal policy: use them intermittently, not daily.
    • The startling ketone anecdote: given to relatives with dementia, sentence length roughly 5xed within 20 minutes. Caveats: it tastes like gasoline, and 1,3-butanediol can affect balance, a serious concern when a broken hip is often the beginning of the end for older adults.
    • Kevin moved his mother, who has non-Alzheimer’s (likely vascular) dementia, into a new home equipped with an AI radar orb that detects falls instantly. She cannot recall breakfast but knows who he is, which he will take all day long.
    • The exercise-for-brain-health protocol Tim assembled with neuroscientist Dr. Tommy Wood: Norwegian 4×4 VO2 max intervals (4 minutes on, ~3 minutes off, 4 rounds) three times weekly for five to six months produces volumetric changes in the hippocampus that appear to last up to five years.
    • The only bike Tim can tolerate for it is the Kaiser M3i indoor bike, because the handlebars raise enough to spare his lower back. Kevin’s sustainable alternative: incline treadmill walking while playing Duolingo chess until 40 minutes disappear. Tim’s version of don’t-let-perfect-be-the-enemy-of-good: a 5-minute, three-set gym session still counts.
    • The JAMA study that grabbed Tim: a phase 3, five-arm randomized trial of MM120 (lysergide, essentially LSD, from the company formerly known as MindMed) for generalized anxiety disorder. Effects were dose-dependent, with 100 micrograms (a standard full trip) as the apparent minimum effective dose, and relief persisting through 12 weeks after a single treatment.
    • Mid-conversation they discover the trial ran at Neuroscape at UCSF, their friend Adam Gazzaley’s lab, which Kevin helped fund. Tim, clinically diagnosed with GAD and OCD, finds 12 weeks of relief from one dose remarkable.
    • Related dementia signals: a case report of an elderly Japanese woman with dementia who took a five-gram “heroic dose” of psilocybin mushrooms, slept 19 hours, and woke temporarily capable of full expositional conversation instead of monosyllables; Tim has also seen an unpublished case report of LSD microdosing producing similar verbal fluidity.
    • Both note the hard ethics: hallucinogens for someone who cannot consent, the devastation of a bad trip you inflicted, versus the possibility of half a day of real connection or slowed decline.
    • Terminal lucidity, the well-documented phenomenon of vegetative or unresponsive patients becoming fully lucid in their final days, leaves both baffled: if cognition is fully localized in a structurally deteriorated brain, where is the lucidity coming from? Kevin’s analogy: we assume nothing is backed up to the cloud.
    • Tim’s caffeine pacing hack: Nutonic nootropic toothpicks (a gift from Chris Williamson), roughly 20 to 25 milligrams of caffeine each, a hard ceiling per toothpick that prevents his chain-refill coffee problem.
    • Kevin’s AI smart home: his Ubiquiti camera system has a full API, so with Claude writing the glue code, the cameras now recognize individual people (and Toaster, who gets a dog emblem), play deterrent audio at loiterers in his alleyway, and open his gate automatically when they read his license plate. The camera costs about $200; anyone can do this now.
    • Tim’s flagship AI project: a 20-year retrospective of his angel investing, built with Claude Code and the Gmail API, testing his own stories about his batting average against hard data. Doing it manually would have taken a year of full-time work by multiple people.
    • The humbling adjacent stat from Kevin: friends with always-on AI wearables report that about 70% of what we confidently remember is what actually happened. Startup genesis stories are the same phenomenon, a five-minute bit polished until the teller believes it.
    • Tim’s most valuable everyday AI use: holistic health cross-checking (contraindications between medications and supplements, could A explain D), hallucination-limited by fact-checking across multiple LLMs.
    • Tim’s contrarian AI take: for most people the honest impact is small “because most shit isn’t worth doing in the first place.” Doing something well does not make it worth doing, and AI is skyrocketing the volume of efficiently produced BS.
    • The demoralization is real, though: an AI trained on your writing produces in 30 seconds what takes 30 hours. Tim compares it to the top Go player who lost the joy of the game after AlphaGo, and his all-format book sales have compounded downward every year since ChatGPT launched (roughly -5%, then -28%, then -49%, tracking toward -67%).
    • The prompt experiment both loved: with cross-conversation memory enabled, ask your model “What are three to five rewarding paths I might explore in the next five years?” Tim sent the answers to close friends who called them outstanding, including a non-book business idea Kevin urged him to build. Ask AI open-ended questions the way you would ask a close friend, not robot questions.
    • Kevin is prototyping “Bond,” an app built from scanned values-card decks: swipe to surface your core values, form explicit agreements with partners and friends that both sides “shake” on, weight the damage of a broken bond, and accumulate a trust ledger. He calls the underlying idea dark information: real relational data (trust, reliability, empathy) that exists everywhere but has never been given physical form.
    • Tim’s writing unlock for the blank page: dictate a rambling brain dump into Wispr Flow while walking, drop it into Claude to clean up, and uncomfortable procrastinated emails come together in minutes. Gear notes: Shokz OpenMeet bone-conduction headset (open ears for traffic, recommended by Exploding Kittens co-founder Elan Lee) and a Sennheiser lav mic plus the Ferrite app as a pocket recording studio that beats studio mics in echoey hotel rooms.
    • State of AI, per both: the big three are Google, Anthropic, and OpenAI, with X/Grok never count-out-able (though Anthropic and Google buying excess Colossus capacity suggests weak Grok demand; Kevin still values Grok’s X-API grounding and uses it heavily for Digg). Meta has phenomenal assets but, Kevin thinks, not the talent to keep pace. Apple is quietly a couple of years out.
    • Kevin’s Google thesis: they own the full stack (TPUs, data centers, models, Android’s install base), and their confusingly high-bandwidth chip architecture is a bet that the future is continuous learning, models improving 24/7 like a child rather than shipping as discrete releases. Consensus estimates put self-improving models 12 to 18 months out.
    • Kevin’s insider color: touring Google X with Sergey Brin and Bill Maris a decade-plus ago, he saw Waymos years before the public knew. Google is sitting on roughly five years of undisclosed deck and holds back frontier models partly for cost and partly to avoid government intervention. In 12 months we will know where Google really stands.
    • Counterweights: ChatGPT owns consumer mindshare and OpenAI must crack advertising, which is very hard; Anthropic is reportedly the fastest-scaling enterprise business ever but keeps taking hits from the administration; no frontier lab will remain unconstrained by government; and China is releasing open-source models on par with the frontier (“doing it the American way”), while AMD’s ~$4,000 local inference box can run massive models at home, eight months behind the frontier, which for many users is fine.
    • The investing lessons: let winners run (Tim: “I’ve lost more money by selling stocks early than I’ve ever probably made buying the original stock”); a famous venture firm’s internal analysis found buying at IPO and holding roughly 10 years matched their gains from early-stage investing through post-lockup; and buy the company, not just the product. Kevin’s friend David Prager spent $100k on a maxed-out Tesla instead of Tesla stock, forgoing roughly $15 million. Tim’s first stock, at about 15 years old, was Pixar, bought because Toy Story convinced him animation was the future.
    • Kevin relaunched Digg: from 20,000 weekly users to nearly 500,000 and millions of monthly page views, pulling the zeitgeist from X and other feeds with heavy AI curation rather than trying to build another social network.

    Detailed Summary

    Grief, the Tail End, and the Last Time

    The show opens with banter about alcohol taxes and ketamine before turning serious: Toaster, Kevin’s 15-year-old dog, just had a terrifying (ultimately survivable) collapse, and the pair lost their friend Om Malik of True Ventures within the week. Kevin, who got the news at a silent retreat, offers the episode’s emotional thesis: the loss and sorrow are the shape love takes when the person is gone, and he would not trade the chaos of caring for people and animals for a calmer, emptier life. Tim thanks Matt Mullenweg for the Antarctica trip that gave him days of psychologically naked time with Om, and for sending him Tim Urban’s “The Tail End,” the post whose parents-time math pushed Tim into years of deliberate family trips before his father needed a wheelchair. Sam Harris’s meditation “The Last Time” extends the theme: you rarely know a last time is the last time. Kevin’s response is to do the thing one more time anyway, bouncy-house backflips at 49 included.

    Zen, Nothing Lacking, and The Way

    Kevin describes his five-day traditional Zen sesshin with Henry Shukman and Yamada Roshi: wall-gazing with eyes open, koan practice on the out-breath, and private interviews with the Roshi. His micro-insight, about two seconds long, was a non-emotional steady state of “nothing lacking,” everything fully present with nothing to add or subtract, what Zen calls the removal of the veil. Tim relays his favorite sessions from The Way (the app both back as a philosophical investment, like the rapamycin dog aging study): “This Too is Me,” which dissolves the burden of a squirrel-chasing mind by including everything experience serves up as you, and Henry’s small physical instructions, like dropping the jaw, which Tim connects to his mandibular snoring device: an eighth of an inch down and forward opens the airway. His bottom line: 10 minutes twice a day captures most of the benefit, and watching the formerly two-weeks-and-out Kevin sustain five years of practice has been deeply reassuring.

    Rock Climbing as the Next Decade’s Project

    Kevin, marching toward 50, wants to stop bookmarking dreams (Japanese woodworking, ships in bottles) and start doing them. Tim’s answer is rock climbing: his repaired right elbow finally allows it, and his stretch goal is multi-pitch in Yosemite despite sweating through his palms at the mere thought of heights. What converted him was the Salt Lake City gym crowd at 11 a.m.: retirees in their 60s and 70s climbing 5.11+, inverted on overhangs, evidence that this sport rewards technique and consistency over youth (women who cannot do five pull-ups climb 5.13). After archery, which he loved but found definitionally solitary, climbing’s social “beta”-trading culture is the draw. The training stack: Michael Eckert’s finger-strength course, the Nug pocket grip trainer, and Abrahangs (Emil Abrahamsson’s 10-seconds-on, 50-off, 10-minute, twice-daily hang protocol). A darker aside grounds the ambition: a friend of Tim’s just died in a plane crash, and The Blade Itself keeps teaching him that life-or-death is often dumb luck.

    Protein, Ketones, and the Dementia Frontier

    The supplements run: Kevin’s new favorite is Pioneer Pastures A2 whey (30 grams, lactose removed, gut-friendly); Tim, disclosure-forward as always, travels with Maui Nui venison and treats the liver-and-heart sticks as a twice-weekly multivitamin. On processed meat, Tim’s heuristic is shelf life over ingredient forensics. The exogenous ketone conversation is more fraught: 1,3-butanediol may stress the liver (though the claim’s loudest advocates sell competing ketone salts), so Tim doses intermittently. The astonishing part: given to relatives with dementia, ketones 5xed sentence length within 20 minutes, going from non-answer answers to full paragraphs, “offline to online.” Balance risks make it dicey in exactly the population that needs it. Kevin’s mother’s new care home uses an AI radar orb for instant fall detection. For prevention, Tim’s protocol from conversations with Dr. Tommy Wood: Norwegian 4×4 VO2 max intervals three times a week for five to six months, whose hippocampal volumetric changes appear to persist up to five years, done on the one bike (Kaiser M3i) that does not wreck his back. Kevin’s sustainable version: incline treadmill plus Duolingo chess.

    MM120, Psilocybin Case Reports, and Terminal Lucidity

    Tim walks through the JAMA-published phase 3 trial of MM120 (lysergide, effectively LSD) for generalized anxiety disorder: five arms (placebo, 25, 50, 100, 200 micrograms), dose-dependent response, with 100 micrograms reading as the minimum effective dose and relief lasting through the 12-week measurement window from a single supervised treatment. Kevin clicks through mid-show and discovers it ran at Neuroscape at UCSF, their friend Adam Gazzaley’s lab, which Kevin helped fund. For Tim, clinically diagnosed with GAD and OCD, episodic rather than daily treatment is the headline. The dementia thread continues: a case report of an elderly Japanese woman who took five grams of psilocybin mushrooms, slept 19 hours, and woke into temporary full conversation; an unpublished LSD microdosing report with similar verbal fluidity. Both wrestle with consent ethics. And then terminal lucidity, the documented phenomenon of unresponsive patients becoming fully lucid days before death, which neither can explain: as Kevin puts it, if it is all localized in a deteriorated brain, where is that coming from?

    AI at Home and AI on Yourself

    Kevin’s Ubiquiti camera system, glued together with Claude-written code against its API, now recognizes faces (and Toaster), scolds loiterers through a speaker, and opens his gate when it reads his license plate, all on a $200 camera. Tim’s project is introspective: a Claude Code plus Gmail API retrospective of 20 years of angel investing, checking who made which introductions, what he passed on, and whether his stories about his batting average survive contact with data (they mostly did; he missed fewer explicit opportunities than he feared). Kevin cites friends with always-on AI wearables: about 70% of what we confidently remember is accurate. Tim’s daily-driver use is health: cross-referencing medications, supplements, and symptoms across multiple LLMs. His caution: most tasks AI accelerates were not worth doing, and the volume of efficient BS is skyrocketing. His countervailing enthusiasm: the “what should I do in the next five years” prompt with cross-conversation memory produced ideas good enough to deeply inform his next chapter. Ask it questions like a close friend. Kevin’s next experiment is “Bond,” a values-and-trust app for making implicit relational agreements (what he calls dark information) explicit, trackable, and reflective. Tim’s practical writing unlock: Wispr Flow voice dumps cleaned up by Claude, especially for procrastinated uncomfortable emails, recorded on a Shokz OpenMeet bone-conduction headset.

    The AI Landscape and Where the Money Goes

    Recorded the day a new Anthropic Sonnet launched, with Mythos due the next day, the forecasting segment lands on a big three of Google, Anthropic, and OpenAI. Kevin’s Google case: full-stack ownership (TPUs whose high-bandwidth architecture only makes sense as a bet on continuous, 24/7 self-improving learning, expected within 12 to 18 months), Android distribution, data center expertise, billion-dollar engineer retention, and a five-year hidden deck he glimpsed touring Google X with Sergey Brin and Bill Maris before Waymo was public. Google holds frontier models back for cost and regulatory reasons; within a year we will know what they have. OpenAI owns consumer mindshare but must solve ads; Anthropic is crushing enterprise ARR while absorbing slaps from the administration; no lab escapes government constraint; China’s open-source frontier-parity models and AMD’s ~$4,000 local inference box threaten the subscription model from below. The investing translation: these companies are going public, and ordinary users will finally get access. The lessons both preach: let winners run, remember that buying at IPO and holding a decade roughly matched one famous firm’s venture returns, and buy the company behind the product you love, the lesson of Prager’s $15 million Tesla and teenage Tim’s Pixar shares. Kevin closes with Digg’s relaunch (20,000 to nearly 500,000 weekly users) and Tim with the sobering chart of his AI-era book sales, compounding downward since ChatGPT.

    Notable Quotes

    “I realized that that gap is just love at the end of the day because I wouldn’t have it unless I loved this man so much. I cared for this person so much. How lucky am I to have crossed paths with this person to get to know them?”

    Kevin Rose, on losing Om Malik

    “When I lost my dad, like that is just a gaping hole of love manifested through sorrow and sadness.”

    Kevin Rose, on grief as a consequence of deep love

    “I had a sense of nothing lacking. Nothing needed to be added and nothing even possibly could be added and nothing possibly could be taken away because everything at that moment was full in the way that it should be.”

    Kevin Rose, describing his micro-insight at the Zen retreat

    “Training to not die sooner than is necessary is not sufficient for me.”

    Tim Ferriss, on why he needs concrete physical goals like multi-pitch climbing in Yosemite

    “Doing something well does not make it important or worth doing in the first place.”

    Tim Ferriss, on AI’s honest impact when most tasks were never worth doing

    “I can still write, but what they can do in 30 seconds is what would take me 30 hours. And I’m just like, it really drains the motivation for me to put in those 30 hours.”

    Tim Ferriss, on AIs trained on his own writing

    “It’s not about those new models dropping. It’s about just like a child learning. Tomorrow it’ll be better than today for forever.”

    Kevin Rose, on Google’s bet that continuous learning replaces the model-release cycle

    “You got to let your winners run as long as possible. I’ve lost more money by selling stocks early than I’ve ever probably made buying the original stock.”

    Tim Ferriss, the takeaway from his 20-year angel investing retrospective

    “You find something that you love and you buy said object when you should actually buy the company.”

    Kevin Rose, on the $100k Tesla that should have been $15 million of Tesla stock

    Watch the full conversation between Tim Ferriss and Kevin Rose here on YouTube.

    Related Reading

    • The Tail End (Wait But Why) the Tim Urban post that quantifies how little time you have left with the people you love.
    • The Way Henry Shukman’s single-path guided meditation app that both Ferriss and Rose back and use daily.
    • Terminal lucidity (Wikipedia) background on the end-of-life phenomenon neither host can explain.
    • LSD (Wikipedia) context for MM120/lysergide and the history behind the generalized anxiety disorder trial.
    • The Botany of Desire by Michael Pollan, the book Tim cites on how dogs (and plants) co-domesticated us as much as we domesticated them.

  • OpenAI’s Leaked 2025 Financials: $34 Billion in Spending, a $38.5 Billion Net Loss, and a $17 Billion Microsoft Bill Ahead of Its IPO

    Infographic summarizing OpenAI leaked 2025 financials: $13.07B revenue, $34B total costs, $20.92B operating loss, $38.53B net loss, where the $34B went, the $17.2B paid to Microsoft versus $303M paid back, inference costs, and IPO valuation context

    OpenAI’s audited 2025 financials leaked this week, and they are the clearest picture yet of what it actually costs to run the company behind ChatGPT. Independent journalist Ed Zitron first published the documents, and the Financial Times independently confirmed them. The headline: OpenAI spent $34 billion last year, booked $13.07 billion in revenue, and reported a net loss attributable to the company of $38.5 billion. The disclosure lands just days after OpenAI confidentially filed for an IPO that could value it north of $1 trillion.

    TLDR

    OpenAI’s audited 2025 numbers, leaked by Ed Zitron and confirmed by the Financial Times, show revenue tripling to $13.07 billion while total costs reached $34 billion, producing a $20.92 billion operating loss and a $38.53 billion net loss attributable to the company. The much larger net loss is inflated by a one-time $41.55 billion non-cash charge tied to OpenAI’s October 2025 conversion from a nonprofit to a public benefit corporation; strip the non-cash items and the loss is closer to $8 billion. R&D alone was $19.18 billion, cost of revenue (inference) was $7.5 billion, and sales and marketing ballooned to $5.73 billion. OpenAI paid Microsoft $17.2 billion in 2025 while Microsoft paid OpenAI only $303 million, exposing a deep Azure dependency. The company burned $1.60 for every dollar of revenue, down from $2.37 in 2024, and gross margin slipped from roughly 40% to 33% as more capable models consumed more compute per query. The leak arrives as OpenAI files a confidential S-1, targets a listing as early as September 2026 at up to a $1 trillion valuation, and races rival Anthropic, which is more valuable on paper and claims it is already turning an operating profit.

    Thoughts

    The most important thing to understand about these numbers is that there are two loss figures and the press will conflate them. The $38.53 billion net loss is the scary headline, but $41.55 billion of it is a non-cash accounting charge from converting investor convertible interests into equity during the for-profit restructuring. That charge is real on the audited statement and it will show up in the eventual S-1, but it is a one-time artifact of OpenAI’s unusual corporate history, not money that left the building. The number that describes the actual business is the $20.92 billion operating loss. That is the one to watch, and it is still enormous.

    The genuinely encouraging line in the whole release is the loss-per-dollar ratio. In 2024 OpenAI spent $2.37 to generate a dollar of revenue. In 2025 that fell to $1.60. A company that is still losing $1.60 on every dollar is not a healthy business, but a company whose efficiency improved by a third in a single year while tripling its top line is at least pointed in a defensible direction. The bull case for OpenAI lives entirely in the slope of that line. If it keeps improving at that rate, the math eventually crosses over. If it stalls, the valuation is a fantasy.

    The Microsoft relationship is the single most revealing disclosure, and it is wildly asymmetric. OpenAI paid Microsoft $17.2 billion in 2025. Microsoft paid OpenAI $303 million. That is a 56-to-1 ratio, and it reframes the partnership: Microsoft is not really a peer or even just an investor, it is OpenAI’s landlord and primary supplier, collecting rent on every model trained and every query answered. The April 2026 renegotiation that capped revenue-share payments at $38 billion through 2030, down from a projected $135 billion, suddenly looks less like a favor and more like OpenAI desperately trying to lower its single largest cost. The dependency cuts both ways, but right now Microsoft holds the better hand.

    The structural problem hiding inside the cost of revenue line is inference. Training a model is a fixed, one-time cost. Serving it is a recurring cost that scales with every one of ChatGPT’s roughly 800 million weekly users. OpenAI spent $5.02 billion on Azure inference in the first half of 2025 alone, and the more capable its reasoning models get, the more compute each answer burns. That is why gross margin went down even as revenue went up. It is the opposite of how software is supposed to work, where the marginal cost of one more user trends toward zero. OpenAI’s marginal cost is real, large, and growing. The counterargument is that per-token inference costs have been falling roughly tenfold a year, so the unit economics could still flip. That is the entire wager.

    Finally, the timing matters more than the numbers. OpenAI’s confidential S-1 means these audited figures were going to become public regardless, since the SEC requires the full prospectus at least 15 days before a roadshow. What the leak changes is who gets to study them first. Prospective IPO buyers, enterprise customers signing multi-year API contracts, and competitors now have the audited books weeks or months early, and they are reading them against Anthropic, which filed at a higher valuation and claims an operating profit. For a company asking the public markets to underwrite a $1 trillion bet on a monopoly outcome that does not yet exist, losing control of the narrative this early is not a small thing.

    Key Takeaways

    • OpenAI’s audited 2025 financials were first published by independent journalist Ed Zitron and independently confirmed by the Financial Times, the first verified look at the company’s books before its planned IPO.
    • Revenue grew from $3.7 billion in 2024 to $13.07 billion in 2025, more than tripling year over year, making OpenAI one of the fastest-growing businesses in history.
    • By the end of 2025 OpenAI was generating roughly $2 billion in monthly revenue, up from about $1 billion a quarter at the end of 2024.
    • Total costs and expenses hit $34 billion in 2025, up from $12.48 billion in 2024.
    • Research and development was the single largest expense at $19.18 billion, up from $7.81 billion, and exceeded total revenue on its own.
    • Of that R&D spend, $10.59 billion went to Microsoft, almost certainly the GPU compute cost of training frontier models on Azure.
    • Cost of revenue, the expense of serving ChatGPT responses (inference), rose from $2.65 billion to $7.5 billion.
    • Sales and marketing jumped from $1.11 billion to $5.73 billion, a 418% increase.
    • General and administrative costs rose from $907 million to $1.57 billion.
    • The operating loss, the truest measure of day-to-day economics, grew from $8.78 billion to $20.92 billion.
    • The net loss attributable to OpenAI was $38.53 billion, up nearly eightfold from $5.09 billion in 2024.
    • The bulk of that jump was a one-time, non-cash $41.55 billion charge from OpenAI’s October 28, 2025 conversion to a public benefit corporation, reflecting the changing fair value of convertible interests and warrant liabilities.
    • Stripping out the restructuring charge and other non-cash items such as stock-based compensation and Microsoft computing credits, the underlying loss was about $8 billion.
    • Including all factors, gross net loss reached $60.35 billion, lowered to the $38.53 billion attributable figure by removing $21.82 billion attributed to noncontrolling and redeemable noncontrolling interests.
    • OpenAI burned $1.60 for every $1 of revenue in 2025, an improvement from $2.37 in 2024, the clearest data point in the bull case.
    • Measured as a percentage of revenue, the operating loss improved from 237% in 2024 to 160% in 2025.
    • In total, OpenAI paid Microsoft $17.2 billion in 2025: $10.59 billion in R&D fees, $6.047 billion in cost of revenue, $527 million in sales and marketing, and $42 million in G&A.
    • Microsoft paid OpenAI just $303 million in the same year, a 56-to-1 imbalance underscoring OpenAI’s Azure dependency.
    • SoftBank paid OpenAI $867 million in 2025.
    • At year-end OpenAI carried $3.64 billion in outstanding payables to Microsoft, plus tens of millions more in accrued and non-current liabilities.
    • OpenAI spent $5.02 billion on Azure inference in just the first half of 2025; Azure inference from 2024 through Q3 2025 totaled $12.43 billion.
    • ChatGPT serves roughly 800 million weekly users, meaning billions of queries a week, each one burning GPU time at Azure’s pricing of about $6.98 per H100 GPU-hour.
    • Gross margin fell from roughly 40% in 2024 to 33% in 2025, because more capable reasoning models consume more compute per query.
    • Research firm Sacra estimates OpenAI’s inference costs reached $8.4 billion in 2025 and will rise to $14.1 billion in 2026, a 68% increase.
    • At year-end OpenAI held just over $50 billion in assets, with almost half in cash.
    • The April 2026 Microsoft renegotiation ended exclusivity and capped revenue-share payments at $38 billion through 2030, down from a projected $135 billion, potentially saving OpenAI up to $97 billion over five years.
    • OpenAI filed a confidential draft S-1 with the SEC around May 22, 2026 and confirmed it publicly on June 8, naming Goldman Sachs and Morgan Stanley as underwriters.
    • The company is targeting a listing as early as September 2026 at a valuation that could exceed $1 trillion, though Sam Altman has said a public offering “may be a while.”
    • OpenAI raised $122 billion earlier in 2026 at a $730 billion pre-money valuation, putting its post-money value around $852 billion.
    • At an $852 billion valuation, OpenAI trades at roughly 65 times its 2025 revenue.
    • Rival Anthropic also filed IPO paperwork this month after raising $65 billion at a $900-$965 billion valuation, making it more valuable on paper than OpenAI, and says it expects to report an operating profit of $559 million in the June quarter.
    • HSBC analysts estimate OpenAI may need more than $207 billion in additional capital through 2030 even under optimistic projections.
    • OpenAI projects profitability by 2029 or 2030; independent analysts put the more likely date at 2031 or later.
    • Bridgewater partner Greg Jensen reportedly told clients the implied revenue multiples price OpenAI for “a monopoly outcome that does not yet exist.”
    • Zitron separately reported OpenAI had a negative 122% non-GAAP operating margin in Q1 2026 and that ChatGPT growth has stalled, with the company projecting paid ChatGPT Plus subscriptions to fall from 44 million in 2025 toward cheaper tiers in 2026.

    Detailed Summary

    How the leak happened and why it matters now

    The audited documents were obtained and first published by Ed Zitron on his newsletter Where’s Your Ed At, then independently verified by the Financial Times, which reviewed the same materials. That dual sourcing matters: this is not a rumor or a model, it is OpenAI’s actual audited financial statement. The timing is the story. OpenAI filed a confidential draft S-1 with the SEC around May 22, 2026 and confirmed it publicly on June 8. Under SEC rules the full prospectus must be released at least 15 days before an investor roadshow, so the 2025 numbers were going to be public soon regardless. The leak simply moved that disclosure forward, handing prospective investors, enterprise customers, and competitors an early look at the books.

    Revenue tripled, costs grew faster

    OpenAI’s revenue rose from $3.7 billion in 2024 to $13.07 billion in 2025, and monthly revenue reached nearly $2 billion by year-end. By almost any normal standard that is spectacular growth. The problem is that costs grew faster, reaching $34 billion against $12.48 billion the year before. The gap between what OpenAI earns and what it spends has widened every year since its founding, and 2025 is the starkest example yet. Revenue alone was outpaced by research and development as a single line item in both of the last two years.

    Two loss numbers, and why both matter

    There are two figures that get cited interchangeably and should not be. The operating loss of $20.92 billion is what the business spent beyond what it earned from operations: training models, serving ChatGPT, paying engineers, running marketing. The net loss attributable to OpenAI of $38.53 billion is far larger because 2025 was the year OpenAI completed its conversion from a nonprofit to a for-profit public benefit corporation, finalized on October 28, 2025. That restructuring triggered a $41.55 billion non-cash charge reflecting the changing fair value of convertible equity interests and warrant liabilities. Before the conversion, investors held convertible interest rights treated as liabilities under US accounting rules and revalued upward as OpenAI’s valuation climbed, creating the charge. It is not expected to recur. Including all minor items, gross net loss reached $60.35 billion, reduced to the $38.53 billion attributable figure after removing $21.82 billion tied to noncontrolling and redeemable noncontrolling interests, primarily the OpenAI Foundation’s stake. Strip the non-cash noise and the underlying loss was about $8 billion.

    Where the $34 billion went

    The spending breaks into four lines. Research and development was $19.18 billion, the largest category, with $10.59 billion of it flowing to Microsoft for training compute. Cost of revenue, the expense of serving responses to users, was $7.5 billion and captures inference, the compute consumed every time someone prompts ChatGPT or calls the API. Sales and marketing reached $5.73 billion, up 418% year over year, a striking jump for a product that grew largely by word of mouth. General and administrative costs added $1.57 billion. The shape of the spending tells you OpenAI is simultaneously racing to build better models, serve a massive and growing user base, and aggressively defend market share through marketing.

    The Microsoft dependency

    The most striking single disclosure is the scale of the Microsoft relationship. OpenAI paid Microsoft $17.2 billion in 2025: $10.59 billion in R&D fees for model training, $6.047 billion in cost-of-revenue for inference serving, $527 million in sales and marketing, and $42 million in G&A. Microsoft paid OpenAI just $303 million the same year. SoftBank paid OpenAI $867 million. The 56-to-1 ratio between what OpenAI pays Microsoft and what Microsoft pays back makes the structural reality plain: Microsoft is OpenAI’s largest landlord. The dynamic began shifting in April 2026, when the two renegotiated, ending Microsoft’s exclusivity and capping revenue-share payments at $38 billion through 2030, down from a projected $135 billion. That could save OpenAI up to $97 billion over five years, though Microsoft keeps its IP license through 2032 and remains the primary cloud partner.

    Why inference is the core problem

    Training happens once. Serving happens billions of times a day. When OpenAI releases a model it spends months and billions on training compute, a fixed cost that falls away when training ends. Inference is the opposite: every ChatGPT message runs through the model on Azure GPU hardware, consuming electricity and compute to generate a response. With roughly 800 million weekly users, that is billions of queries a week, each burning GPU time at roughly $6.98 per H100 GPU-hour on demand. OpenAI spent $5.02 billion on Azure inference in the first six months of 2025 alone. Sacra estimates full-year inference costs of $8.4 billion in 2025, rising to $14.1 billion in 2026. This is why gross margin fell from about 40% to 33% even as revenue tripled: more capable reasoning models consume far more compute per query, and revenue has not kept pace with the cost growth that capability generates.

    What it means for the IPO and the race with Anthropic

    OpenAI was last valued around $852 billion post-money after raising $122 billion in early 2026, which puts it at roughly 65 times 2025 revenue. It has named Goldman Sachs and Morgan Stanley as underwriters and is targeting a listing as early as September 2026 at up to a $1 trillion valuation, though Altman has hedged that it “may be a while” and that staying private might be the better course. HSBC estimates the company may need more than $207 billion in additional capital through 2030. The race is with Anthropic, which filed paperwork the same month after raising $65 billion at a $900-$965 billion valuation, making it more valuable on paper, and which says it expects a $559 million operating profit in the June quarter. The contrast is sharp: the two leading AI labs heading toward public markets at the same time, one bleeding cash at scale, the other claiming profitability, both asking investors to bet on a future that has not arrived.

    Notable Quotes

    “The financial condition of OpenAI is deeply concerning. $38.53 billion in losses are astronomical, and far higher than most believed it would be. Losses also appear to be mounting year-over-year at a dramatic rate, and I’m not sure how this company finds a way toward any kind of sustainability or profitability.”

    Ed Zitron, the independent journalist who published the leaked audited financials

    “It’s unclear what this means, nor how OpenAI reconciled the removal of $3.74 billion in costs. I will not speculate further.”

    Ed Zitron, on a discrepancy he found in the restated 2024 figures

    “OpenAI’s two biggest expenses are R&D and marketing. Budget cuts there, coupled with an ability to raise prices or win new sources of revenue, could see the company move into the black over time. Cutting R&D would be the most difficult part of that, given that AI companies can only hold onto their customers by generating the best-performing models.”

    Jim Edwards, Fortune, on whether OpenAI has a realistic path to profitability

    “What the audited documents make impossible to argue is that the path to profitability is short, clear, or cheap.”

    TechTimes analysis of the leaked OpenAI financials

    The implied revenue multiples price OpenAI for “a monopoly outcome that does not yet exist.”

    Bridgewater partner Greg Jensen, reportedly telling clients how to read OpenAI’s valuation

    “OpenAI spent $34bn last year as the ChatGPT maker poured money into a race to dominate the fast-growing AI market ahead of a planned stock market listing.”

    George Hammond and Bryce Elder, Financial Times, framing the audited 2025 spend

    Read Ed Zitron’s original reporting with the full breakdown here, and the Financial Times confirmation here.

    Related Reading

    • Ed Zitron, Where’s Your Ed At the primary source that broke the audited 2025 financials with the full line-by-line breakdown.
    • OpenAI (Wikipedia) background on the company’s history, structure, and the nonprofit-to-for-profit conversion that drives the non-cash charge.
    • Inference (Wikipedia) on the recurring compute cost that explains why OpenAI’s gross margin shrinks as usage grows.
    • Anthropic the rival lab that filed IPO paperwork the same month at a higher valuation and claims it is already operating at a profit.
    • SEC on confidential filings context for why OpenAI’s audited numbers were headed for public disclosure regardless of the leak.