Oct 27, 2025 · 1h 13m · 20vc
Sequoia Partner, David Cahn on Who Wins in AI, Defence & The New $0–$100M Playbook · 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, Sequoia Capital Partner David Cahn joins host Harry Stebbings to analyze the physical, economic, and operational dynamics of the AI infrastructure bubble, while outlining why defense tech is poised to become the next major venture capital frontier.
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 20.1% of the talking time here. How this is scored →
Harry as informed peer 5.1 Guest teaching 5.6 Guest disagreement 2.4 Harry pushing back 4.5
speaking balance: gold is Harry, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 35:52 Rejecting the VC King-Making Premise
David directly rejects the popular industry belief that elite VCs can king-make startups, labeling it as a false ego trap that leads to poor investment decisions.
Hardest push from Harry ▶ 35:52 Direct Rejection of King-Making DenialHarry explicitly tells David 'I love you, but I respectfully disagree,' insisting that Sequoia's brand and follow-on capital give portfolio companies an undeniable unfair advantage.
Biggest teaching moment ▶ 27:28 Debunking the Debt Unwind MythDavid corrects the consensus media narrative that AI is a debt-fueled bubble, explaining that because it is equity-funded, any crash will hit retail equity portfolios directly rather than triggering a 2008-style credit freeze.
Harry holds his own ▶ 58:30 Calling Out Sequoia's Defense AI MissesHarry forcefully challenges Sequoia's record in defense AI, asking directly if the firm was asleep at the wheel for missing category leaders Helsing and Anduril.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Harry as informed peer | Guest teaching | Guest disagreement | Harry pushing back | Why |
|---|---|---|---|---|---|---|
| Challenging Sequoia's Role in Defense AI | 5 | 4 | 3 | 7 | Harry opens with an aggressive provocation asking if Sequoia was asleep at the wheel by missing defense leaders Helsing and Anduril. David deflects gracefully by setting up his broader thesis on the physical constraints of AI infrastructure and power. | |
| The $600 Billion AI Revenue Question | 6 | 6 | 2 | 5 | Harry demonstrates familiarity with Cahn's well-known $600B AI revenue question and asks if US GDP growth contradicts it. David clarifies the underlying math, updating the figure to $840B and highlighting supply chain bottlenecks. | |
| Unforeseen Trends: Scaled Pay Packages and Talent Desperation | 3 | 6 | 2 | 3 | David explains how AI talent packages have ballooned to $100M-$1B per individual. He schools the host on how venture probability math gets abused to justify these astronomical figures. | |
| Re-evaluating Meta and Founder-Led Moats | 5 | 5 | 2 | 4 | David candidly acknowledges his 12-month prediction on Meta was wrong, while defending founder-led companies long-term. Harry pushes on whether labs like OpenAI and Anthropic disprove the vertical integration thesis. | |
| Surviving the AI Bubble: Commodity vs. Monopoly | 4 | 5 | 2 | 3 | David details how consensus has shifted toward acknowledging an AI bubble. He distinguishes between short-term capital destruction in market cycles and multi-decadal technological transformations. | |
| Producers vs. Consumers of Compute | 7 | 7 | 3 | 6 | Harry offers a strong counter-argument by citing cloud providers like GCP, AWS, and Azure as non-commodity compute producers. David responds with a historical reframe on anomalous monopoly eras versus market transparency in AI. | |
| Capital Allocation and Companies That Don't Need Funding | 5 | 6 | 3 | 5 | Harry points out that the compute consumer thesis is now accepted wisdom. David points out the gap between accepted narrative and actual pitchbook data where over 80% of VC capital still flows to compute producers. | |
| Noticing Fragility: Circular Deals and Hyperscalers | 4 | 7 | 3 | 4 | David educates the host on market fragility using Nassim Taleb's anti-fragility framework. He outlines how risk shifted from hyperscalers like Microsoft and Amazon to smaller players like CoreWeave and circular vendor financing by chipmakers. | |
| The Scale of the AI Power Question | 5 | 6 | 1 | 3 | Harry presses on global capital constraints for multi-gigawatt buildouts. David calculates that 100 gigawatts translates to an $8 trillion question, illustrating the massive scale of unfunded commitments. | |
| Debt vs. Equity: How the AI Bubble Will Unwind | 6 | 7 | 3 | 6 | Harry pushes on Oracle's high debt-to-equity ratio and Mag 7 concentration risks. David corrects the press narrative by explaining that this bubble is equity-funded rather than debt-funded, meaning an unwind will hit consumer equity portfolios. | |
| AI's Realistic Impact on GDP and Productivity | 5 | 7 | 3 | 4 | Harry brings up Masa Son's prediction of a 5% GDP impact from AI. David agrees on the GDP volume but refutes Masa's 50% profit margin assumptions using global McKinsey macroeconomic data. | |
| The Myth of King Making in Venture Capital | 6 | 5 | 5 | 7 | David explicitly rejects the concept of VC king-making. Harry forcefully pushes back, citing real-world examples like Profound where Sequoia's brand creates immediate network moats. | |
| Do Gross Margins Matter in AI? | 5 | 5 | 2 | 4 | Harry queries whether software gross margins still matter in AI. David explains that initial low margins can expand over time as compute costs fall, referencing historical examples like Snowflake. | |
| The Zero to $100 Million Playbook and Hyper-Growth | 5 | 5 | 2 | 4 | Harry questions extreme VC growth expectations like reaching $2M ARR in 10 days. David defines the modern $0 to $100M ARR playbook exemplified by companies like Harvey, Clay, and Juicebox. | |
| The Experience Curve and Scar Tissue of Early Founders | 5 | 6 | 2 | 3 | Harry outlines his benchmark for time taken to scale from 1 to 50 million ARR. David highlights portfolio companies like Clay and Juicebox that spent years in the wilderness building scar tissue before breaking out. | |
| Quick Successive Rounds and Overcapitalization | 6 | 6 | 2 | 4 | Harry quotes Pat Grady regarding the dangers of quick successive rounds. David agrees that overcapitalization creates false winning signals, explaining that anything multiplied by zero is still zero. | |
| The AI Talent Revolution: Hiring Young Generalists | 5 | 7 | 4 | 5 | David advocates hiring 23-year-old AI generalists over experienced staff engineers because no one has over five years of ChatGPT experience. Harry challenges him on the emotional maturity risks of young hires. | |
| Career Advice: Breaking the Memetic Algorithm | 5 | 6 | 2 | 4 | David explains the recursive memetic algorithm young graduates use to choose careers and how AI disrupts it. Harry shares his frustration that UK students still chase traditional investment banking roles. | |
| Why Defense is the Next AI | 7 | 6 | 3 | 8 | Harry confronts Sequoia on missing Helsing and Anduril in defense AI. David admits Sequoia was late with humility, then lays out his thesis comparing current defense AI to two years post-Transformer paper. | |
| National Champions vs. Traditional SaaS | 7 | 4 | 1 | 6 | Harry argues defense is not a broad venture category like SaaS because it can only support a few national champions. David agrees with Harry's framing and outlines Sequoia's focused national champion thesis. | |
| Quick Fire Round: Personal Growth and Investment Philosophy | 3 | 4 | 1 | 2 | In a quickfire round, David discusses getting his driver's license, parenthood, and his biggest venture miss in Datadog, which taught him to focus exclusively on top priority opportunities. | |
| The Undervalued Future of AI Voice Interfaces | 3 | 4 | 1 | 1 | David highlights AI voice interfaces as wildly undervalued technology, pointing to Sequoia's investment in Sesame. Both end on an optimistic note about AI's generational impact. |
Statements from this episode (67)
Cahn: Infrastructure oversupply in an AI bubble benefits compute consumers
“Consumers of compute benefit from a bubble, because if we overproduce compute, prices go down, your cogs goes down, and your gross margin goes up.”
Cahn: Venture investors cannot force a startup to succeed
“The lesson that punches you in the stomach in venture is you can't make a company succeed.”
Cahn: Power generators for data centers are sold out until 2030
“Generators are sold out until 2030”
Cahn: AI power infrastructure was the best trade of 2025
“The best trade of 2025 was the AI power trade.”
Cahn: AI is one of the largest drivers of US GDP growth
“AI is now one of the biggest contributors to GDP growth in the United States”
Cahn: AI infrastructure investment requires $840B in end-user revenue
“If you invest one hundred fifty billion in NVIDIA chips, that's about three hundred billion of data center investments. And to pay that back, the person using the compute needs to earn a 50% gross margin. So there's about six hundred billion of revenue that ne…”
Cahn: Whether real end-user demand exists for AI compute remains unanswered
“Is there actually an end user for this compute? I don't think that's been answered.”
David Cahn: AI industry is at the start of data center construction delays
“One of the predictions I made last year, in addition to saying it was going to be the year of the data center in 20, 25, I said, Hey, we're going to have these construction delays. We're going to have issues now in building out these data centers. And the info…”
David Cahn: Physical data center construction capability will be a competitive moat
“One of my core perspectives that I've been developing over the last 18 months of writing about this is that construction itself is going to be a moat. The ability to build things is hard. And I think we underestimate that.”
Cahn: Top AI researchers command $50M to $1B compensation packages
“If you're a 25 year old recent grad from an elite university who is perceived to be an AI expert, you can get a fifty hundred million dollar pay package right now. And if you are a brand name that everyone recognizes your name, you can get a billion dollar pay…”
Cahn: Enormous AI compensation packages reflect desperation to prove investment ROI
“I think they're symbolic of this sort of desperation in the ecosystem where it's like, we need to eke out progress. We need to prove that all these investments are worth it.”
Cahn: Investors overestimate how much individual researchers drive AI progress
“That said, I think we are psychologically biased to overestimate what that percent contribution is. And it may be the case that there's these broader macro variables, which we'll talk about, I'm sure later in this discussion, there's these broader macro variab…”
Cahn: Meta's $100M AI packages stem from underperformance
“And I think that Meta, you know, these hundred million packages are coming in large part from Meta because they haven't performed as well as they thought they were going to.”
Cahn: OpenAI and Anthropic are vertically integrating into physical infrastructure
“Well, I think the simple version would be OpenAI and Anthropic are now steel servers and power companies. And that's like a big change that's happened in the last 12 months. And so I actually think we're, you know, in many ways, OpenAI and Anthropic are becomi…”
Cahn: Competitive pressures will force all AI model providers to vertically integrate
“I think that competitive pressures will push all of the model providers to spend more time on this and to have teams focused on this. So I think the answer is yes. I do think that this is a trend that is going to be durable.”
Cahn: Sequoia Targets Only One or Two AI Investments Per Year
“I don't need to find 10 investment opportunities this year. I'd like to find one or two investment opportunities a year that I really love.”
Cahn: Compute-consuming AI products will outperform compute producers long-term
“So I think if you're a producer of compute, you're fundamentally in a commodity business, just like an oil company is in a commodity business, and that is going to trade a different way, and that is going to have more cyclicality than if you're in a non-commod…”
David Cahn: Seven Tech Companies Account for 40% of S&P 500
“We're living in an anomalous monopoly era. And it's funny because there's so many comparisons to industrial revolution, and in some ways we're living in this new gilded age. And we have these seven companies, and they represent 40% of the S&P 500, which is jus…”
David Cahn: AI Era Competition Will Prevent Monopoly Profits
“The difference is that these monopolies are not hiding in plain sight. We all now know that if you build an amazing tech company, it can be worth a trillion dollars. In 2000, if you told people that they can have a trillion dollar tech company, they would have…”
Cahn: Over 80% of AI venture capital goes to compute producers
“And I think probably 80% plus of the dollars in AI are still going to producers of compute. Not consumers of compute.”
Cahn: The best venture investments are companies that don't want capital
“And I think some of the best investments are those companies that don't want to raise capital.”
Cahn: Zoom was profitable and didn't want capital when Sequoia invested
“When Sequoia invested in Zoom, they didn't want to raise capital, right? They were profitable. They were doing really well.”
Cahn: AI market consists of roughly 10 recursive major players
“There's like 10 players around this big chessboard, and they're extremely powerful, and each of their moves affects the other people's moves, so it's kind of recursive.”
Cahn: Big Tech AI spending will not stop without incentive changes
“And so I think until the incentives change, the behavior is not going to change. And so there is no coordinating mechanism.”
Cahn: Microsoft walked away from two data centers in early 2025
“Microsoft walked away from two data centers and it sent a message to the market like, hey, we're not stepping up.”
Cahn: Chip makers finance AI buildouts to book revenue circularly
“The chip companies are now stepping up and saying, okay, we'll absorb some of the risk. We'll put in the capital to finance this build out where the demand on the other side is not so clear because of course the chip companies also get to book this as revenue.”
Cahn: Building out one gigawatt of AI power costs $40B to $60B
“A gigawatt is forty billion dollars to build out. Jensen says it's 50 or 60 if you use the next generation Vero Rubin chips. So let's say it's somewhere between 40 and sixty billion.”
Cahn: AI infrastructure build-out is mostly equity and cash funded
“I actually think that what's interesting about this AI build out is that for the most part, and let's put Oracle aside, which maybe has some debt, but for the most part, the AI build out today has been equity funded and cash funded.”
Cahn: Any AI bubble unwind will hit household stock portfolios, not banks
“What I think is going to be interesting if to the extent that the bubble unwinds at some point, it's going to be an equity unwind. And what that looks like is 40% of the S&P 500 is basically a bet on AI. And so to the extent that the bet unwinds, stock prices …”
Cahn: Market concentration in 'Mag Seven' leaves stocks vulnerable to AI narrative shifts
“And so I think you have a similar dynamic here where the mag seven are just a humongous portion of the market. Now these companies are great. They have cash machines, like they're going to do fine, but I do think we should be concerned that these companies rep…”
Cahn: AI will disrupt at least 5% of global GDP ($9T+)
“I actually agree with him fundamentally that AI is going to affect five percent of GDP. Probably where I disagree with Massa, so I think he used the nine trillion dollars. I think that's the number he used. It's going to Disrupt nine trillion of GDP. And then …”
Cahn: Monopolistic dominance is not the steady state of business
“People over estimate the monopolistic nature of businesses, and that we're living in this sort of unique Gilded Age monopolistic era, and that that Is, is not the steady state of business.”
Cahn: The AI timeline is the most overestimated factor in tech
“I think that we are, I think there's a number of things that are over being overestimated. I think the most important one is the timeline.”
Cahn: AI newcomers expect immediate AGI while pioneers project long timelines
“Like the path we're on was invented by these people who are raising the most concern or saying the timeline is longest. And it's the people who've been in AI the shortest who I think are saying like, Hey, it's going to come tomorrow.”
Cahn: Venture capital firms cannot 'king-make' startups with capital alone
“I don't believe in King making and that's maybe a controversial thing to say. I think one of the lessons, you know, you'd think like, oh, Sequoia should be able to King make companies and like, that's so great. And that would be, by the way, if that was true, …”
Cahn: Sequoia's primary value to portfolio startups is talent recruiting
“I think that's the number one way that companies do benefit from having Sequoia on the cap table, is that, is talent and recruiting, and we can talk more about that, and I'm fascinated by recruiting, and recruiting dynamics, so I do think Sequoia helps with th…”
Cahn: Low early gross margins do not prevent long-term software success
“Plenty of companies that get critiqued for having low gross margins end up being super healthy businesses in the long run. You know, Snowflake was one of the big indicts on Snowflake in the early days was that it had a low gross margin. Obviously it's a very g…”
Cahn: Overanalyzing gross margins can impede venture capital returns
“At the end of the day, our job is to invest in companies that become really successful, not to be like super smart about analyzing them. And so I think sometimes the instinct to criticize a gross margin can get in the way of money making.”
Cahn: Harvey, Open Evidence, Clay, and Juicebox are reaching $100M ARR rapidly
“Companies that are on that trajectory or have crossed that trajectory are companies like Harvey and Open Evidence and I think, and Clay and Juicebox.”
Cahn: Scaling $0 to $100M ARR signals true product-market fit in AI
“Everybody's on the internet and everybody wants to buy AI. So if you have something really good, it's going to get adopted really fast. And so I do think to the point of playing the game on the ground and adapting to what you see in the market, the biggest thi…”
Stebbings: Speed from $1M to $50M ARR matters far more than initial timeline
“I always say I don't care how long you take to get to a million in revenue, but I care desperately about how long it takes for you to go from one to 50.”
Cahn: Growth speed from $1M to $50M ARR accurately predicts company valuation
“There's a lot of data that indicates that that is a very good leading indicator of what it's worth. The data I've looked at suggests that that is a historically good algorithm.”
Stebbings: UiPath took nine years to hit $550k ARR
“You know, one of yours is UiPath, and he's a dear friend of mine, Daniel. And I mean, it took nine years to get to 550 K of ARR.”
Cahn: Raising Seed through Series B in 12 months is a false narrative
“Most, there's this false narrative, I think, that, like, all the good companies, they, you know, they raise the seed, and then they raise the A, and then they raise the B, and it's all in 12 months, and it, Revenue, that's not really how most of these companie…”
Cahn: Led Sequoia's investment in Clay at $1B+ valuation after long pivot
“Sequoia invested at the Series A in, I think, 2019. The company spent three or four years in the wilderness really figuring it out. I look at Kareem, I think the man is, like, enlightened from this experience. Like, a super painful experience. Varun ended up j…”
Cahn: Long-term market volatility is irrelevant unless overextension causes bankruptcy
“Anything multiplied by zero is zero. And I think that's one of the really tricky things in investing, which is just to say that market volatility doesn't matter in the long run if you have a great business. But if you overextend yourself and then some crash ha…”
Cahn: Founders should be maximally aggressive while investors provide perspective
“Your job as a founder is to be maximally aggressive. And you should do that. And then the investor should hopefully be giving some advice, helping think these things through, giving some perspectives you know, understanding the, a broader time horizon perspect…”
Cahn: AI startups will win by hiring 23-to-25-year-old AI generalists
“I think that the new playbook for these AI startups is actually going to be much more about hire the AI generalist, this 23, 24, 25 year old who's really native in AI, really passionate about it, and I think those are the sort of the front lines that are going…”
Cahn: Generative AI breaks traditional recursive career algorithms
“The recursive algorithm of, like, what did the guy a year above me and the person a year above me do is actually breaking because those people didn't have this information. They didn't know that AI was going to change the world. They didn't understand gen AI.”
Cahn: AI gives junior startup hires near-parity with senior peers
“Maybe 10 years ago, if you joined a startup and people didn't join startups that often, 10 years ago, maybe 10 years ago, if you joined a startup, like there's this whole experience curve, you're the junior engineer. There's a lot of people who are smarter tha…”
Cahn: Defense technology is the next AI-scale investment opportunity
“I would say, and I think this ties into our conversation so far, that AI, that defense is the next AI.”
Cahn: Sequoia Capital was late to investing in defense tech
“So I do think, look, there's no way around it. Skoya was late to defense.”
Cahn: Defense tech is under-hyped and awaiting its ChatGPT moment
“And so, anyways, I think that the Transformer paper moment was the was the Ukraine war, and then I think the Chi-GPT moment hasn't happened yet. And so I think that defense is actually, you know, is under-hyped in some ways, or like underestimated in some ways”
Cahn: Defense Tech Adoption Cycle Is Only 1% Complete
“We're like one percent there on catching up. Like we're actually so, so early in this defense cycle because, you know, now we have a few dozen companies, maybe a hundred companies that have sort of new innovations. They're not integrated into the force structu…”
Cahn: Few concentrated defense tech winners will emerge per country
“And I guess my framework, and this is the thesis that I've been investing behind now for the last couple of years is my framework is there are going to be fewer companies that succeed in defense for this reason. Defense is consolidated for good reason. There's…”
Cahn: Anduril is clearly the US national champion in defense tech
“And in my view, Andrew is clearly the national champion in the U S and credit to that team, just really phenomenal company, phenomenal visionaries.”
Cahn: Sequoia has invested in defense startups Kela and Stark
“So the other two national champions that I've invested in, one is a company called Kella, which we think is going to be a national champion. It's based in Israel. The thesis is that Israel has the best people in the world for this, and they can help defend the…”
Cahn: Kela and Descartes are Israel's top talent consolidators
“I think the two big talent consolidators right now in Israel are Kala and Descartes.”
Stebbings: Defense tech is not a broad market category like SaaS
“I don't think defense is a category. And you're like, what? A category is enough that can support an ecosystem with its breadth and depth. I don't think defense is. I think there is your andrills and maybe two to three more in the U.S., and I think there's, yo…”
Cahn: Sequoia will only invest in defense startups every couple years
“I mean, my objective, I probably invested in a dozen AI companies in my career. I hope to invest in 20 more. My objective is not to invest in 20 more defense companies throughout my career. I think it's going to be a very small handful of companies. Maybe we'l…”
Stebbings: Most venture dollars invested in defense tech today will be lost
“I think so many of the dollars going into it today will be lost.”
David Cahn: Missing Datadog deal to Dragoneer was major financial loss
“One big financial miss is Datadog, and I worked on this before joining Sequoia, but I remember, you know, Datadog was this amazing company. The numbers were incredible. It was proper. Like, it was one of these businesses where you just, like, your mouth waters…”
Cahn: Voice as an interface for AI is wildly underestimated
“I think that people are underestimating voice as an interface for AI.”
Sequoia Capital invested in AI voice startup Sesame
“We just, I think today right before this podcast, we announced our investment in a company called Sesame, which is an AI voice company, an AI conversation company.”
Cahn: Sesame hit 1M users and 5M minutes within weeks of launch
“They launched this product, this AI voice product that you can talk to, got a million users in a few weeks, five million minutes, like just tremendous product market fit.”
Cahn: People will talk to and form relationships with AI within 10 years
“I think the idea that we're going to be sitting here in 10 years talking to our AI, having a relationship with our AI, I think that's very likely, and I think it's a little bit sci-fi right now and I think it's going to get less so in the coming years.”