Summary

Corgi Insurance’s latest funding round valued it at $2.5B; CEO Nico’s one-word culture is “winning”: seven days a week, work trials, and “if your days off happen to be Saturday and Sunday every week, then you will not have a place at Corgi.” He sleeps 3-4 hours on an office mattress and has paid in psoriasis and heart palpitations: “I would rather measure my lifespan in victories than years.”
The most tradeable datapoint in the episode: Corgi spends ~$400,000/month on Anthropic and $0 on OpenAI. “Anthropic frankly speaking has been out-executing OpenAI. Their product’s much better… OpenAI needs to lock in” — but switching “wouldn’t be that hard,” and a big enough token discount would make them consider a contract, so the spend is not locked in.
Fundraising playbook worth stealing: Brian Chesky told him “never take the highest price” and Corgi always takes the second or third highest; rounds close in “a couple days at most,” because a long process means “you start to optimize for selling equity instead of selling your goods and services.” The operating creed he adopted from an insult: “good companies get deals done.”
His market-structure view: private markets are a vehicle for valuing growth, public markets for valuing cash flow — a public Corgi “would probably be worth less.” Against Harry’s future-free-cash-flow framing: “brand is worth something and identity is worth something — why do you buy a Rolex instead of a Chinese copy?” Consistently, he has never sold a secondary: “I believe our equity is going up and I’m voting with my feet.”
The intellectual spine is a legitimacy thesis: startups are “a crisis of legitimacy” (his Roman Empire analogy), tier-one VC brands are rented legitimacy, and fraud — including the ARR frauds Harry raises — is “downstream of credentialism or that quest for legitimacy,” passing muster “especially if a round is moving quick, which most good companies’ rounds do.”
Very long London: exceptional British and European talent plus the US “visa situation” give London “a phenomenal opportunity to become a center of AI” — Corgi opened London before New York (“people that move their companies to New York do so because they want to date”). Meanwhile the under-$100k, 24/7 Corgi Cafe that investors hated is always full at 3am, slightly profitable with Brex and Deel sponsorships, and is expanding to London.
Forward AI calls: AI makes sales and marketing more important, because “organic network effects around B2B product launches just don’t really exist” anymore; there are “trillions of dollars” in extracting retiring workers’ knowledge “inside of computers that can think and talk”; and he expects “more ChatGPT moments… maybe a couple years from now,” with biology a candidate for category-defining outcomes.

Deep dive

1. Winning as a culture: asymmetric upside, no weekends, and “yeah, easy” at 1,000 people

Nico’s philosophical base is anti-loss-aversion: “losses are beautiful because you can have asymmetric upside with winning.” His touchstone is Bezos’s baseball analogy — in baseball upside is capped at four runs, so swinging for the fences is often a losing strategy, but in business “a home run might result in actually infinite upside… something like AWS popping out of nowhere.” Napoleon was “the best general of all time” and still lost battles; fear of losing turns you into “this creature afraid of doing anything.”
The seven-day week follows mechanically: “Whatever you can get done in 5 days, I promise you you’ll get more done in six and seven.” An occasional day off is fine, but “if your days off happen to be Saturday and Sunday every week, then you will not have a place at Corgi” — and why stop at defending five days? “Why not four or three or two or one?” Asked for culture in one word: “winning” — “maybe in the past I would have answered ex-founder, hackers… all that’s actually kind of fake.”
He rejects the claim that he’s alone in this: visit high-growth AI companies in San Francisco and “the offices are pretty full on the weekends… I don’t think it’s a coincidence.” A hypergrowth startup not working weekends is “basically just quiet quitting.”
Can the intensity hold at 1,000 people? “Yeah, easy.” Lawyers and accountants won’t grind like engineers — “you just need to hire more of them for the same work output… a return on investment calculation.” The only reason he can think of for an off-ramp is accepting lower growth for less volatility, which he currently declines: “volatility is something that I think can produce alpha.”

2. A mattress in the office, 3-4 hours of sleep, and lifespan measured in victories

He literally lives in the office — a mattress in what staff call “Nico’s room,” showering at the Equinox one street over (which closes at 8pm on Fridays: “that was unpleasant”; the London office has a shower, SF doesn’t). The justification is symbolism: governments and religions care about flags for a reason, and there’s a difference between “leading the troops from the front line” and preaching hard work “from some yacht somewhere.”
The costs are stated candidly: psoriasis and heart palpitations from being “so all in,” averaging 3-4 hours of sleep while — without irony — encouraging employees to sleep a lot. His resolution: “I would rather measure my lifespan in victories than years.”
Harry’s would-you-rather — trillion-dollar Corgi but dead at 50, or failure and living to 80: “I think the answer to that is pretty easy… if I’m dying either way.” Harry matches it with the Olympian study where 98% would trade ten years of life for a gold medal.
On the backlash — “the death threats and the DMs, people think I’m nuts” — he doesn’t care, though he notes young men “don’t have that much support within society.” To the exclusionary charge: plenty of parents work at Corgi, “life’s about trade-offs and sacrifices,” and if you want to clock in and check out “there’s infinite other jobs you can get… Corgi is not for everyone.”

3. Startups are a crisis of legitimacy — and tier-one VCs are rented brand

His biggest not-aggressive-enough regret: going to college. “If you want to do big things, you need to go out in the wilderness a little bit and be humbled.” Credentials are supposed to convey legitimacy but “it’s a bit too naked, you know, and on the nose.”
The framework: startups face a crisis of legitimacy, like the Roman Empire’s non-hereditary monarchy — “the emperors weren’t inheriting any legitimacy, so if they were a bit crap you could just kill them.” No blood ties, no inherent loyalty — so founders borrow legitimacy from “tier one investors,” which is why “perhaps capital isn’t so fungible after all.”
His tell that something changed: it would be very unlikely that Steve Jobs would have flexed Apple as “a Sequoia company,” yet today companies are routinely described by their investors — while VCs admit “out of the other side of their mouth” that post-investment they can’t fix an incompetent team.
Is Corgi legit yet? “We’re more regulated than a bank and we make a lot of revenue… I don’t think we’re quite legit yet.” The ladder never ends — pension funds, endowments, then government, “the ultimate form of legitimacy… probably because the government can just take people’s money.” Even likely Andreessen, Kleiner, and Sequoia “at some point were just a couple guys in some shitty room.”

4. The $100k cafe investors hated — and being “very long London”

The 24/7 Corgi Cafe was an accident of real estate: taking the SF building’s signage meant taking a derelict barber shop, and the financial district “closes really early… you can’t even find a place to eat past 6:00, 7:00 p.m. — clearly most restaurants don’t have a growth mindset.” It cost under $100,000 to open and is now always full at 3am; over 20 people have told him they submitted YC apps or signed term sheets there.
Economics: he says they lose money on the cafe, then says sponsorships mean it is “probably slightly profitable.” Investors’ verdict: “They hated it at first… I got a ton of calls and texts.” He’d already dropped the seven-day chef on the same logic: food in the office matters (going out is “extra cognitive load”), but non-core operations “lead to logistical burdens” — and he loathes pampering culture, “Google is like the worst of that.”
Now a “cafe expansion head” is being hired for London, where Corgi is “very long”: exceptional British and European talent, and “given the whole visa situation in the United States… a phenomenal opportunity for London to become a center of AI.” Corgi opened London before New York — “the people that move their companies to New York do so because they want to date.” SF remains special, “almost a spiritual quality… as if everyone were in exile”; London’s possible handicap is “too many pubs.”

5. Hiring and comp: work trials, low cash, equity top-offs, and cutting the bad fast

Every candidate does a work trial — mock work for a day or several, ideally over a weekend so they see the office full and “very quickly learn that we’re not joking around.” Bonus: candidates can trial without taking time off, which “does make it easy to poach.” In trials, soft skills matter more than hard skills — you can tell who’s all-in versus “into checking in and checking out.”
His most revealing interview question: “what matters to them and why” — people answer it honestly. “I want money for my family” isn’t alarming (“that normally comes from a certain place”), but pure money motivation “leads you to local maximums that don’t actually result in the most money being made.”
On pay: “if you’re working for a salary at Corgi, you’re doing something wrong.” Low cash comp is good; generosity comes as ad-hoc equity “top-offs” for performance, so that staying is “the highest EV thing they can do” — and if you can’t offer that upside, good people should leave. But comp alone retains no one: “if you’re not solving big problems, no matter how much you pay people, you’re not going to retain them if they’re actually good.”
Any org splits into three: the really good, the fungible/mediocre, and the bad — “getting rid of the bad very quickly is important even if it makes a lot of noise.” On the mediocre: people where “you’re not going to be sobbing and punching the wall if they’re no longer with you.” Harry’s own heuristic, which Nico doesn’t dispute: ask for more money → usually very good; ask for more title → usually very bad.

6. Fundraising: never take the highest price, close in days, “good companies get deals done”

The origin story undercuts any guru pose: his first Corgi pitch (likely First Round Capital), raising five on 28 in 2024, off a red-eye with genuinely red eyes, needing $5M between Friday and Monday. The pitch “was a lot of like, just trust me” — he wouldn’t even name the insurance carrier for fear VCs would “call them and ruin it.” Early-stage investor behavior he’s endured: flossing their teeth in meetings, eating barbecue ribs, one on FaceTime “hitting his vape… I thought he was asleep.”
Likely Brian Chesky’s advice, taken to heart: “never take the highest price ever” — Corgi always takes the second or third highest, and “we could raise all of our rounds at higher prices, but we never do.” Rounds close in “a couple days at most” because time in market means “you start to optimize for selling equity instead of selling your goods and services.” To Harry’s disservice-to-investors challenge: “They all got good deals, too.”
The line he now runs the company on came as an insult — an investor’s “good companies get deals done.” “At first I was mad… I’ve actually taken it to heart because I think it’s true. Time kills all deals.” It applies to VC firms too; the venture deals he respects most are the rescues — backing a company whose “metrics might be going in the wrong direction… doing the hard thing in venture is rare, unfortunately.”
Harry’s pushback on speed — “I do want to get to know you” — draws blood on both sides. Nico: “VCs always say that, but they could put in the work to get to know you, and they never do.” Harry: “We try and you know what we get? ‘Thanks so much for your email. We’re heads down.’” Nico, unrepentant: “I’m heads up — if you’re working on an important deal that’s going to double your revenue, the last thing you want to do is meet with some VC.” His one concession: “a bad investor is much worse than no investor — and it’s hard to get rid of them.”

7. Private markets price growth, public markets price cash flow — so he sells nothing and skips boards

He has never sold a secondary despite constant offers: “I believe our equity is going up and I’m voting with my feet” — his stake is “the scarcest asset that I could have any meaningful influence over.” Angel investing is out for the same reason: “it’s a distraction… not core to winning.”
A public Corgi “would probably be worth less because the public markets don’t know how to evaluate high growth or potential” — look at listed financial-services companies, “they’re all terrible, they all don’t grow.” His frame: “private markets feel like a vehicle for valuing growth; public markets feel like a vehicle for valuing cash flow” — what we used to call dividend stocks.
Harry’s challenge — aren’t all companies ultimately embodiments of future free cash flow? — gets only partial agreement: “maybe at some terminal point from a theoretical perspective, but brand is worth something and identity is worth something — why do you buy a Rolex watch instead of a Chinese copy? Both tell the time.”
Governance, same instinct: “I don’t like boards… every committee decision I think is BS.” Boards are “a little theatrical,” a check-the-box — “you think Elon Musk or Mark Zuckerberg are caring what their board thinks?” And they don’t prevent fraud: unethical management “is going to do it anyways, probably with or without a board — that’s a management problem.”

8. Few good companies, and fraud as the shadow of credentialism

His under-discussed obsession: “the number of good companies is very small. A good company wants to win, is world historic, will be talked about in a hundred years.” Most don’t want to put in the work — “it’s okay to be average and part of the mediocre masses.” What investors miss is the “metaphysical” character of a company — “Andreal” (likely Anduril) and Anthropic “are super different companies,” both possibly good, but “I don’t think people really understand the philosophy behind the management teams that they’re investing in.”
On Harry’s question of whether VCs are to blame for the rise in fraud, especially around ARR: “the founders lie,” and fraud is “downstream of credentialism or that quest for legitimacy.” VCs are only partially responsible — post-investment “there’s very little you can do” — except crypto investors who pushed founders to take on personal criminal liability while “arbitraging that and pumping the tokens… that was the wrong type of tokens to work on.” The uncomfortable kicker: frauds pass muster “especially if a round is moving quick, which most good companies’ rounds do.”
His ask of the industry: VC fund leaders should take a more active public stance and align deployed capital with the changes they want — funds set the ecosystem’s incentives, and “incentives often drive behavior.”
His own first believer: likely Jared Friedman at YC — “there’s been times where he was our only believer… everyone else is a doubter.” He’d do YC again.

9. Anthropic is out-executing OpenAI — and the next ChatGPT moments

Given OpenAI or Anthropic at the same price of 900: he flags stated-versus-revealed preference, since Corgi spends “maybe $400,000 per month on Anthropic and $0 on OpenAI”. Spiritually, OpenAI is “a little more pure” — it created the AI-lab-as-company back to YC Research, while Anthropic came from employees unhappy with management, “complaints about profit and ethics… the EA movement” — and “when your origins are that impure it’s hard to overtake the king.” But right now “Anthropic frankly speaking has been out-executing OpenAI. Their product’s much better… OpenAI needs to lock in.” The loyalty caveat: switching “wouldn’t be that hard,” and a big enough token discount would get a contract considered.
His structural AI call: AI makes sales and marketing more important, not less. A hardcore engineering team used to win by shipping something obviously better; now “the organic network effects around B2B product launches just don’t really exist in the same way.” B2B sales is good, “marketing is awful” — conferences are “not it” (“hanging out with a bunch of sweaty people in a room”) and B2B should learn from consumer.
His biggest changed mind in 12 months: valuing experience. He used to consider “old people using fax machines moving at 2mph… uniformly bad”; now he sees “trillions of dollars to make getting that knowledge out and putting it inside of computers that can think and talk” — especially from the retiring generations.
The ten-year view is humble by construction: ten years ago the exciting technology didn’t exist — only in late 2022, with ChatGPT 3.5, did LLMs become obviously important. He expects “more ChatGPT moments… maybe a couple years from now that will really shake things up,” with category-defining outcomes in areas where venture hasn’t worked before — “biology is one that comes to mind.”

The Era When Everything Is Vibe: Why Xiaohongshu Is Buying the World Cup? — A Crossover with Flip Sports Anthropic's Digital God, Pope vs AI, Job Loss Narrative Flips, Open Source Crackdown Coming?