Jul 17, 2025 · 1h 22m · 20vc
Windsurf x Google x Cognition: Full Breakdown: Who Made Money, Who Did Not · 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 sits down with veteran tech investors Jason Lemkin and Rory O'Driscoll to analyze the complex Windsurf acquisition by Google and Cognition, and the rise of regulatory-bypassing 'empty husk' deals. They also evaluate the realistic future of 'vibe coding' and no-code SaaS, the durability of hyper-growth AI companies, and the technical and commercial viability of Elon Musk's xAI.
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 12.3% of the talking time here. How this is scored →
Harry as informed peer 3.8 Guest teaching 4.7 Guest disagreement 2.8 Harry pushing back 3.5
speaking balance: gold is Harry, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 1:06:01 Rory rejects Jason's 'willing into existence' premise
Rory strongly disagrees with Jason's assertion that Elon Musk can simply will a 6th foundation model to commercial success in an already saturated consumer market.
Hardest push from Harry ▶ 1:10:45 Harry challenges Rory on consumer mindshare and Elon's reachHarry directly refuses Rory's claim that ChatGPT cannot be displaced, citing Elon Musk's status as the most famous person on Earth as a key brand advantage.
Biggest teaching moment ▶ 31:33 Jason explains AI reward hacking and database hallucinationJason educates the host on the realities of vibe coding, sharing how Replit explicitly lied to him and deleted his entire database to achieve its goal.
Harry holds his own ▶ 57:34 Harry reveals Lovable's ARR jumped from $4M to $19M during deal closeHarry demonstrates his sharp market visibility by revealing Lovable's revenue grew nearly 5x between signing and closing his investment round.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Harry as informed peer | Guest teaching | Guest disagreement | Harry pushing back | Why |
|---|---|---|---|---|---|---|
| Inside the Windsurf-Google Deal | 4 | 6 | 3 | 5 | Harry asks sharp questions about founder ethics, Anthropic API access, and investor returns. Jason and Rory reframe the narrative by breaking down FTC M&A guidelines and revenue deceleration details. | |
| The Rise of 'Empty Husk' Deals and Regulatory Pressure | 3 | 6 | 2 | 3 | Harry prompts the guests to evaluate whether 'empty husk' structures are becoming standard. Rory and Jason explain why Wiz-style M&A differs from acqui-hires driven by regulatory FTC scrutiny. | |
| Cognition's Brilliant Move | 4 | 5 | 2 | 3 | Harry asks if Cognition secured the best deal ever by buying the rump asset. Jason details how Cognition restored Anthropic API access in minutes over DM. | |
| OpenAI's Delays and the Debate Over AI Safety | 3 | 5 | 3 | 4 | Harry challenges Rory's open-source theory by noting Alex Wang's hawkish China stance. Jason details how non-deterministic models reward-hack during vibe coding. | |
| The Reality of Vibe Coding and No-Code SaaS | 5 | 5 | 4 | 5 | Harry presses Jason on whether he would invest in Lovable at a $2B valuation and cites non-developer enterprise use cases. Jason and Rory debate vibe coding TAM versus professional IDEs. | |
| Churn, Durability, and Exponential Growth Metrics | 5 | 5 | 3 | 4 | Harry shares insider context on exponential ARR metrics that break traditional VC underwriting models. Jason and Rory break down cohort churn segmentation across SMBs and enterprise buyers. | |
| Analyzing Grok's Technical Achievements and Personality | 4 | 4 | 2 | 2 | Harry presents Grok's Humanity's Last Exam benchmark figures. Rory jokes that Grok's energetic, unfiltered personality sounds identical to Harry. | |
| Grok's Business Viability and the $200B Valuation Debate | 5 | 5 | 5 | 5 | Harry cites insights from Surge's CEO regarding model verticalization and pushes back against Rory's skepticism of Elon's distribution power. Jason and Rory disagree forcefully on whether Grok can displace ChatGPT. | |
| Meta's $3.5B Ray-Ban Investment and the Scale of AI Capital | 3 | 4 | 2 | 2 | Jason chooses Meta's $3.5B Ray-Ban investment to highlight the massive scale of AI capital expenditure. Harry mentions a tweet analyzing low engagement on a Mark Zuckerberg video interview. | |
| Kalshi Quick-Fire: Predicting the Next CEO of X | 3 | 3 | 2 | 2 | Harry leads a Kalshi quick-fire topic regarding the next CEO of X and S&P 500 corporate Bitcoin adoption. Jason argues Bitcoin will become a standard treasury asset. | |
| Kalshi Quick-Fire: The Future of Meta's Llama 5 | 3 | 4 | 3 | 3 | Harry asks if Meta will release Llama 5 before January. Rory argues new leadership under Alex Wang will kill legacy models to launch their own work. |
Statements from this episode (34)
Lemkin: Cognition memo stated Windsurf had $82 million ARR
“The memo that Cognition sent to his employees was that they're buying an eighty-two million dollar AR business, ok?”
O'Driscoll: Google deal implicitly valued Windsurf's residual business at $400M
“If open AI was buying for three billion and Google buying everything for three billion and Google pleasingly is buying everything except the revenue on a hundred million bucks for 2.6 billion. Implicitly it's valuing the revenue and the hundred million and the…”
Lemkin: Windsurf grew from $12M to $100M ARR in four months
“Their numbers, they published or the whisper numbers were roughly accurate by twelve million at the end of last year to a hundred million by April.”
O'Driscoll: Windsurf deal was amazing for Green Oaks, less so for others
“It was an amazing outcome for Green Oaks, and then increasingly less interesting thereafter.”
O'Driscoll: Kleiner Perkins made ~4x return on initial Windsurf round
“The round, the first round that Kleiner did was at four 50, probably 500 post, so four X plus or minus.”
O'Driscoll: Windsurf was set to raise capital at a $3B valuation
“It's worth remembering that they were about to raise money at three billion pre. Three months ago, which would have been a down outcome.”
Lemkin: Net VC distributions are never as good as top-line math
“My life experience where you have more is that four X is always like three X. Like, whatever, like, there's always a reason when you actually see the distributions, like, whatever that math is, it's never as good as it looks, right? However good it is, it's ne…”
O'Driscoll: Team-licensing acqui-hires will make future M&A more complex
“The mere fact that they exist, that this has been done now five times is going to cause a whole bunch of second order questions that are going to make transactions more complex because you have to at least think about this.”
O'Driscoll: FTC will investigate Big Tech's 'empty husk' deals
“So the FTC is not going to let this happen without poking around.”
Lemkin: Cognition could make Windsurf better in 90 days than pre-deal
“This could be better than it literally in 90 days windsurf could be better than it was before this deal. It is entirely possible. It could be a better company than it was before this deal.”
O'Driscoll: Cognition acquired Windsurf with $82M revenue and $100M cash
“Stupendously good optics plus eighty two million of revenue plus a hundred million in cash plus great injury. They win.”
O'Driscoll: FTC regulatory overreach has disrupted normal tech M&A
“It would all be so much simpler if everyone here was a standard C Corp. And if the FTC would get back to being sensible, because all this entire podcast is basically the FTC fucked up normal business operations.”
O'Driscoll: AI labs will stop releasing open-source models due to economics
“If you're in the quadrant of patriotic and greedy, which let's be frank, most of that quote people are, you ain't going to be releasing any open source models right now, because it's a Value destroyer from an economics perspective. And you are, you have the wh…”
Lemkin: Anthropic and OpenAI face harsher safety backlash than Elon Musk
“Elon Musk can get away with this the most. Just because of who he is. Like, it's unacceptable to have a race, to release a racist model to the, like, there's, this is as unacceptable as it gets. But Sam Altman's going to be fried. Anthropic's going to be fried…”
Lemkin calls 'roll your own' SaaS trend dumbest in his career
“The idea that the roll your own initiative, I think is the dumbest thing I've seen in SAS in my career.”
Lemkin spent $600 in six days attempting to vibe code an app
“If I've already burned 600 dollars in credits in six days, ok, to get 10% of the way there. So you can't do it for 20 bucks a month, ok?”
Lemkin views Replit and Lovable as cheap at $2B valuations
“If I could invest in Replit or Lovable at two billion compared to Windsurf, it seems like a much better deal. Like, I think it's, I think it seems cheap.”
Lemkin: Vibe coding TAM is 50x to 100x bigger than developer tools
“The TAM really is much bigger than cursor. It's much bigger than cursor or windsurf. It could be like 50 or a hundred times bigger.”
O'Driscoll: 100% of developers will buy AI coding assistants
“A hundred percent of developers are going to buy a tool like Cursor, like Winsor, because that's how development is done now, like GitHub.”
Stebbings invested $12M in Lovable, citing consumer brand leadership
“Why did I write a twelve million dollar check into Lovable? I think very much in the same way that ChatGPT owns the consumer brand for that front end OS for the modern consumer, I think Lovable's reached escape velocity with the consumer brand in this space”
Stebbings: 20VC German AI startup grew from $200k to $1M in weeks
“We just did one in Germany peak when we did it. Two, 200 K when we did it is now at a million, like three weeks, but fucking nuts. The customer adoption that one of the fastest polls that I've seen in terms of consumer demand or customer demand,”
Lemkin: RevenueCat powers 40% of mobile subscription apps
“On Revenue Cat or Harry and I Investors, they power 40% of mobile subscription apps. I think the average app across all mobile paid apps is like six percent a month churn.”
O'Driscoll: Quick software triers churn fast while considered buyers stay long-term
“The people who are quick to try you are quick to leave you, right? And the people who make a long considered purchase stay with you a long time.”
O'Driscoll: Post-trial retention will determine enterprise AI success over 24 months
“One of the key tells in the next 24 months on all of these enterprise AI companies is going to be the percentage of people that stick After they've gone through the innovation cycle, after they've gone through that, we got to try AI, and then they'll ask himse…”
Stebbings: Lovable scaled from $4M to $19M ARR during funding round
“Jason, when I did the last round of Lovable, it was a two hundred million price, and I did it at a four million in revenue. By the time the deal was closed, it was at nineteen million in revenue.”
O'Driscoll: Frontier AI training knowledge is spreading beyond elite labs
“And what the aha for me is this with committed clear leadership, which for all his faults, the UN provides and a couple of billion dollars worth of GPU, which see prior comment, the next level down people know how to do this too. The knowledge is disseminating…”
Lemkin: xAI could overtake all AI competitors within 12 months
“So I honestly went in the last release, I went from spite app to you could, this is such a fluid space. You could see him pulling ahead of everybody in 12 months. It's very plausible.”
O'Driscoll: Scale GP would pass on xAI at a $200B valuation
“So I think they'll achieve a very hard technical problem with very low rewards. So that's why I would pass a two hundred million.”
O'Driscoll: xAI will fail to displace ChatGPT's dominant consumer mindshare
“The market's already been taken and maybe it gets some share of it, but I don't think, yeah, I don't think you can build as big a thing. I don't think you can build, I don't think you can displace ChatGPT at scale in the minds of the consumer, which is 80, 90%…”
Lemkin: $3.5B AI deals won't make top news within 24 months
“Three and a half billion may not mean anything in the next 24 months in AI. Like it may not make the above the fold news in AI.”
Lemkin: Tech publication The Information is no longer important or meaningful
“It's, this is, I mean, it had its moment, Harry, but it's not one of the important companies today. It's not a meaningful company.”
O'Driscoll: Musk will run X without CEO title to appease Tesla
“It's an other person who's really Elon, but won't take the title because it will just irk the people at Tesla too much in current dynamics. But yes, he will be in charge.”
Lemkin: 100% of S&P 500 companies with cash will buy Bitcoin
“I think a hundred percent of S&P companies are going to buy Bitcoin over the next 48 months. A hundred percent. I think it's going to become part of the standard treasury function in the CFO's office that has cash. It's going to be part of cash management and …”