Sep 25, 2025 · 1h 25m · 20vc
The Impact of H1B Visas on Startups in the US & NVIDIA Invests $100BN Into OpenAI · 20VC with Harry Stebbings
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
In this episode of 20VC, host Harry Stebbings, Jason Lemkin Lemkin, and Rory O'Driscoll O'Driscoll analyze the AI CapEx boom, monopoly dynamics in tech, and the structural shifts in venture capital portfolio management. The discussion covers everything from the valuation hangovers of the 2021 bubble to H-1B visa changes and IPO liquidity strategies.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Harry holds 14.2% of the talking time here. How this is scored →
Harry as informed peer 4.5 Guest teaching 3.9 Guest disagreement 3.3 Harry pushing back 3.4
speaking balance: gold is Harry, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 1:15:16 Jason rejects Harry's view on pulled term sheets
Jason forcefully rejects Harry's stance against pulling term sheets, arguing that founders offering only one-hour weekend deadlines without disclosure deserve to have deals rescinded if bad facts emerge.
Hardest push from Harry ▶ 2:43 Harry interjects on scaling laws consensusHarry directly interrupts Rory's premise about infinite scaling bets, pushing back with counter-evidence that GPT-V shifted toward efficiency because scaling improvements hit a ceiling.
Biggest teaching moment ▶ 38:15 Rory breaks down portfolio management and risk aversionRory walks Harry through financial literature and the mathematical rationale behind risk aversion, explaining why selling liquid winners to secure net worth overrides simple expected return maximization.
Harry holds his own ▶ 41:59 Harry pulls precise metrics from Navan's S-1 filingHarry showcases deep preparation by citing exact data points from Navan's S-1, including $613M revenue, 32% growth rate, 10,000 customers, and 110% net dollar retention to anchor the discussion.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Harry as informed peer | Guest teaching | Guest disagreement | Harry pushing back | Why |
|---|---|---|---|---|---|---|
| Is NVIDIA and OpenAI's Capital Loop an Infinite Money Machine? | 4 | 4 | 3 | 5 | Harry interjects to challenge Rory's premise regarding scaling laws, noting recent GPT efficiency shifts. Rory and Jason reframe the capital dynamic, emphasizing that OpenAI is allowed to run the experiment regardless of consensus. | |
| Anthropic's Position and the Monopoly Debate | 2 | 3 | 2 | 1 | Harry sets up a prompt about Anthropic's standing relative to OpenAI. Jason and Rory lead the discussion on compute scale and whether OpenAI is deliberately steering clear of formal monopoly backlash. | |
| NVIDIA's Monopoly Risk, Customer Concentration, and the CapEx Boom | 6 | 3 | 3 | 3 | Harry demonstrates strong category expertise by bringing specific data labeling customer concentration facts from Merkur to the table. Rory expands on the CapEx boom disconnect from underlying revenue. | |
| Is the AI Boom Echoing the 1999 Dot-Com Bubble? | 3 | 4 | 4 | 2 | Harry prompts a comparison to historical tech bubbles. Rory forcefully disagrees with corporate stock buyback practices that simply offset employee dilution, calling the practice dumb as rocks. | |
| Public Market Exuberance, Cash Reserves, and 'Frothy' LP Behavior | 3 | 3 | 2 | 2 | Harry asks about personal public portfolio allocation in a booming S&P market. Jason reveals his zero cash position while Rory explains asset allocation and LP behavior in frothy cycles. | |
| Capital Concentration and the Realities of Late-Stage Venture Funding | 5 | 4 | 4 | 5 | Harry brings up capital concentration statistics and debates funding realities in unloved sectors like restaurants. Jason directly dissents from popular VC Twitter commentary regarding required growth rates. | |
| Evaluating Tech Exits, Multiples, and the Psychology of Portfolio Management | 4 | 5 | 3 | 4 | Harry reacts to the valuation multiples of recent exits while Rory provides a structured masterclass on blended fund returns and risk aversion theory using portfolio management principles. | |
| Navan's S-1 Filing and the Debate Over Asset Concentration | 6 | 4 | 3 | 5 | Harry quotes exact metrics from Navan's S-1 filing and presses Jason on asset concentration limits in early-stage funds. Jason and Rory break down GP risk vs LP risk diversification. | |
| IPO Strategy, Lockups, and the Mechanics of Venture Liquidity | 4 | 6 | 3 | 2 | Harry asks about liquidity mechanics and pre-IPO secondaries. Rory and Jason educate on lockup waivers, secondary offerings, and the legal nuance of holding stock while possessing board inside information. | |
| The Impact of H-1B Visa Fee Changes on Tech Startups | 6 | 3 | 2 | 2 | Harry cites specific figures on H-1B visa applications and GDP contributions. Jason shares personal experience of how early H-1B hires were vital to his first startup's survival. | |
| Notion's ARR Re-acceleration and the Reality of SaaS Valuations | 6 | 4 | 4 | 5 | Harry highlights Notion's ARR re-acceleration, defending why growth at scale is noteworthy. Rory pushes back on calling it exceptional while noting SaaS valuation compression from 2021 heights. | |
| Moving On From 2021 Froth, Due Diligence, and Pulled Term Sheets | 6 | 4 | 5 | 6 | Harry criticizes investors pulling term sheets during rushed deals. Jason strongly counters, defending investors who pull back after uncovering omitted facts during compressed weekend timelines. | |
| Redefining 'Founder Friendly' and 'Founder Honest' in Tech | 3 | 4 | 5 | 2 | Harry asks if the guests have become less founder friendly. Jason delivers a passionate reframe arguing that standard founder friendliness is performative bullshit compared to standing by founders during down cycles. |
Statements from this episode (42)
Jason Lemkin is 0% cash, mirroring his 2008 fund allocation
“Well I'm excited because just like 2008 at the moment, I'm zero percent cash.”
Jason Lemkin calls 'founder-friendly' in venture capital meaningless
“Founder friendly has become bullshit, right?”
Jason Lemkin claims hot AI deals involve zero due diligence
“Any hot AI deal, there is no diligence provided, nor is any done. It's just done on Saturday.”
Jason Lemkin calls ChatGPT the 'Standard Oil of tech'
“This is an epic monopoly, like we've never seen. Think how much ChatGBT already dominates our lives. It's the standard oil of tech.”
Rory O'Driscoll: Tech giants will spend into oblivion on AI CapEx
“That's bad news, but the good news is all six of them are determined to spend themselves into oblivion to win the prize.”
O'Driscoll: AI capital loop will continue unchecked for 1-2 years
“No one's going to call time out along the way, at least for another year or two based on this, because everyone's getting, you know, the capital is being made available and everyone involved is getting kind of mentally marked up.”
Lemkin: Sam Altman fulfills predictions much faster than Elon Musk
“Elon Musk will say something, and it happens. He's just like a couple of years off, right? But it always happens. The self-driving cars and the rockets, right? Sam says something, and it happens, like, kind of soon”
O'Driscoll: AI infrastructure returns require unrealistically heroic assumptions
“I'm somewhat more skeptical because I think the level of heroic assumptions you have to start making to make these investments all work is high.”
O'Driscoll: Anthropic Could Raise $10B Tomorrow If It Wanted
“My understanding is they were beating people off with a stick on the recent on Tropic round. And if they decided they want another ten billion, it'll be there tomorrow morning. So it's not capital constrained.”
O'Driscoll: Six customers generate 83% of Nvidia's quarterly revenue
“I think six customers I saw on the last quarterly accounted for something like 83% of the revenue”
O'Driscoll: AI CapEx spend of $600B far outpaces $30B-$40B market revenue
“They've been willing to say, let's spend in aggregate six hundred billion of capex per year on a market that today, depending on how you add up all the revenues, is yielding 30 to forty billion dollars in revenue.”
O'Driscoll: Concentrated AI vendors avoid price pressure due to buyer speed
“If you have a services business and you're only selling to six customers, you can make this intellectual MBA case that, oh my God, they'll get ground down on price. That case is totally wrong because those six customers don't have time to optimize. No one of t…”
Lemkin: Nvidia free cash flow surged from $3.8B in FY23 to $60B in FY25
“NVIDIA fiscal year, 2023, 3.8 billion dollars in free cash flow fiscal 2023. Pretty good, right? 3.8 billion. Fiscal 2024, twenty seven billion. Fiscal 2025, sixty billion”
O'Driscoll: Nvidia repurchased $9B in stock last quarter with $60B authorized
“They bought nine billion of stock back last quarter. I would have, you know, which is just interesting at this period in the cycle, and they've authorized a sixty billion buyback program.”
O'Driscoll: Share buybacks aimed solely at offsetting equity dilution are foolish
“I think the buyback the same as your diluted is as dumb as rocks, right? You should buy back when your stock is cheap and you should sit on your cash when the stock is dear, right?”
Jason Lemkin: LPs bragging on LinkedIn signals peak market froth
“The frothiest sign is now LinkedIn LPs are bragging about their returns on LinkedIn LPs. When LPs are start bragging about their returns to me that that's the most, 20, 21 moment I've seen when LPs are LPs who are usually hide behind the wizard of Oz curtain a…”
Stebbings: 75% of 2025 VC dollars went to 19 companies
“75% of VC dollars in twenty-twenty-five went to 19 companies.”
O'Driscoll: Focus on 18-24 month follow-on rounds, not 5-year returns
“It's far more useful to say to yourself, what does this, cause you don't have visibility on five years, except at a macro level. But what you can say is what's this company going to do in 18 to 24 months? And do I believe it can raise a follow on round at a st…”
Lemkin: 100%+ growing startups face harder funding market than 2 years ago
“It's a lot more work to get funded growing a hundred percent at Twenty million, or a 110% of ten million than it was 24 months ago. It's just a lot more work, because you can't get the meeting.”
O'Driscoll: Startups achieving 'triple, triple, double, double' still get funded
“I think that whole meme is a little overdone, right? The whole, oh, three, three, two, two, double, it's not good enough... And I think that if you really have clarity on that kind of traction, especially at any, you know, and some reasonable scale, right? I t…”
Iconic Growth logged $8.5B Netskope and $1B DX exits in one week
“And Iconic had two billion dollar exits this week, Iconic Growth, okay? It had Netscope, which no one talked about, eight and a half billion dollar exit, and they congratulated themselves on DX, which Atlassian bought for a billion”
Rory O'Driscoll: Large funds must concentrate capital in 5 to 7 deals
“There's very few ways to make a good return on a large amount of money, which is why the bigger the fund size, the more you have to be in only five or seven deals.”
Jason Lemkin: 80% of tech IPOs trade down after listing
“Rory's point, I mean, look, I mean, what, 80% of IPOs trade down, right?”
Harry Stebbings: Navan's S-1 shows $613M revenue and 32% growth
“Six hundred and thirteen million revenues, growing 32% year on year, 10,000 customers, a 110% NDR”
Jason Lemkin invests nearly 10% per company across two checks
“I'm a pretty concentrated investor, so I'm gonna get to 10% almost in two checks into any deal”
Stebbings: Capital concentration limits destroy top venture returns
“Capital concentration limits are the enemy of great venture returns, and that's why they have 33% of their fund in Airbnb.”
O'Driscoll: Navan held operating expenses flat to slightly down year-over-year
“If you actually look in the P&L at Navan, right, they really held OPEX flat to, in fact, slightly down this year versus the prior year.”
O'Driscoll: VCs typically take 18 to 24 months post-IPO to fully exit
“Typically, my go, 18 months from the IPO, plus or minus, maybe 24.”
Lemkin: Board members legally exploit insider information by holding stock
“Hold is your best time to trade on, to legally trade on inside information, though. If you're on the board or close to it, it's your best time to trade on that illegal by not trading. Yeah, you can hold legally with inside information. It's a privileged positi…”
Lemkin: Wealthy investors should hold winning public stocks indefinitely
“When you're up enough personally, it makes sense to hold your, all your winners in the public markets for taxes and other reasons. There's just no reason if you're personally up enough to sell any winner, right?”
Lemkin: First Startup Would Have Failed Without H-1B Visa Holders
“My especially my first startup, which was hard. It had a material science component. It wouldn't have been possible without H-one B at least on its surface. Okay. I had H-one B. I ate two folks with H-one Bs on my, They were transfers, right? I didn't sponsor …”
Harry Stebbings holds Notion equity from portfolio company acquisitions
“And at five hundred million in revenue, you could actually squint And see my ten billion dollar stock, which I have since it acquired some of my companies, reasonably being finally up to the watermark of ten billion dollars.”
Rory O'Driscoll: Notion would price at $4B-$8B in an IPO today
“A 30% grower, I mean, Netscope and Navan, and well, Netscope has priced already, you know, it's seven, eight, nine times NTM, so something like that, four or five billion, maybe more if they afford growth rate is a little better, or if I'm underestimating the …”
O'Driscoll: Startups shouldn't delay IPOs just to reach peak private valuations
“I don't think you should stay private just to earn your way back into twenty billion dollars. I think you can go public and then I think it will boil down to, you know, you might go public and either the stock gets converted or as we've discussed, the preferen…”
O'Driscoll: Ten-year-old tech companies are priced on fundamentals, not sizzle
“Stories at the forefront can be priced on sizzle. Stories that are 10 years old are going to be priced on fundamentals.”
Lemkin: Wrote down all frothy portfolio holdings except one $300M ARR deal
“I wrote down everything. I guess a year and a half ago, whatever I get that I wrote every, everything that had a hint of froth. I have one deal. I'm still carrying myself, right? Because it's over three hundred million in revenue and growing. I'm holding it as…”
O'Driscoll: Sequoia achieved a 7x total return on Klarna
“Its peak was 45, Sequoia, which SoftBank did Sequoia brilliantly did the round at five or six, and obviously two or three X that, and I think overall seven X the overall investment.”
Jason Lemkin: Real founder-friendliness means writing checks when others won't
“Founder friendly is writing the check when no one else does. That's founder friendly. That founder friendly is when no one else is there at the board meeting anymore and you're there and you still have a W on the other side of it. That's founder friendly. Foun…”
Rory O'Driscoll: VC founder-friendliness can only be judged in tough deals
“The only way to a judge who really is founder friendly is how they behave in a tough deal. There's no information on how you do in a good time, right?”
Lemkin: A US-China TikTok deal will be signed within 60 days
“Whatever this crazy deal is, it's going to get signed in the next 60. I think these deals are going to get done this calendar year, and I think it's going to get signed”