Jun 26, 2025 · 1h 16m · 20vc
Navan Files to Go Public and Canva Pulls the Brakes: Why and What Happens? · 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
Host Harry Stebbings and tech veterans Rory O'Driscoll O'Driscoll and Jason Lemkin Lemkin analyze the shifting dynamics of Silicon Valley, exploring the talent wars, private versus public market valuations, and the reality of AI-driven labor displacement.
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 12.8% of the talking time here. How this is scored →
Harry as informed peer 4.8 Guest teaching 4.5 Guest disagreement 3.3 Harry pushing back 3.5
speaking balance: gold is Harry, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 14:50 Rory rejects Jason's VC loyalty argument
Rory aggressively counters Jason's moralizing about founders dumping LPs, arguing that returning $15B through high-value exits is the free market working as intended.
Hardest push from Harry ▶ 45:48 Harry refuses Rory's market pricing framing on IPOsHarry directly confronts Rory, calling his pricing thesis naive and reductionist while citing a first-hand conversation with a $10B public CEO on public market misery.
Biggest teaching moment ▶ 57:33 Rory explains the Leverage Beta strategyRory educates the group by reciting a breakdown of why AI startups lie about product capability today to lock down market position before underlying LLMs catch up.
Harry holds his own ▶ 31:36 Harry articulates market unbundling theoryHarry demonstrates deep venture expertise by breaking down how niche AI vertical applications unbundle legal services and shrink the addressable market size, winning Rory's explicit agreement.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Harry as informed peer | Guest teaching | Guest disagreement | Harry pushing back | Why |
|---|---|---|---|---|---|---|
| Meta's Existential AI Bet and Acquisition Strategy | 2 | 4 | 2 | 1 | Harry tees up the Meta acquisition discussion citing a text exchange with Rory, then hands off to the guests. Rory and Jason outline Meta's defensive AI bet, comparing it to Oculus VR insurance and citing ChatGPT mobile download metrics. | |
| The 'Magic Room' Talent and California's Non-Compete Advantage | 4 | 5 | 3 | 3 | Rory explains why early OpenAI talent holds value and how California non-competes enable talent mobility. Harry pushes back on commoditizing magic and reveals a $75M acquisition offer for 20VC, which Jason jokingly dissects. | |
| The Death of Loyalty in Silicon Valley | 1 | 4 | 5 | 1 | Jason laments the lack of institutional loyalty among Silicon Valley founders leaving LPs and startups. Rory forcefully pushes back, arguing that free markets optimize talent allocation and VC returns speak louder than moralizing. | |
| Underwriting Harvey's $5 Billion Valuation | 5 | 5 | 3 | 4 | Harry pushes the guests to explain how Harvey justifies a $5B valuation. Rory breaks down legal market economics and Westlaw software spend vs labor budgets, while Harry presses whether AI can truly capture labor budgets. | |
| Crosby and the Rise of AI-Native Professional Services | 6 | 4 | 3 | 4 | Jason shares his personal workflow replacing junior counsel with Claude. Harry articulates a venture framework on market fragmentation in legal tech, pointing out how niche tools unbundle the TAM. | |
| Navan's IPO, Circle's Valuation, and Public Market Dynamics | 5 | 5 | 3 | 4 | Harry introduces Navan's IPO filing and Circle's $68B market cap, questioning whether Circle is peak meme stock. Rory and Jason break down retail trading behavior and interest rate exposure. | |
| Why Canva is Delaying its IPO and the Larry Ellison Model | 7 | 5 | 4 | 7 | Rory argues that public vs private cost of capital dictates Canva's IPO delay. Harry forcefully challenges this framing as naive and reductionist, citing a discussion with a $10B public CEO about public market misery. | |
| Cluley's Aggressive Marketing and the 'Leverage Beta' Playbook | 6 | 5 | 5 | 7 | Jason praises Cluley's consumer grade product experience, but Harry interrupts to challenge its viral marketing tactics like strippers and police cars. Rory reads a memo explaining the 'leverage beta' playbook of marketing ahead of LLM capability. | |
| Slack's API Lockdown and the Agentic Threat to Salesforce | 5 | 6 | 2 | 3 | Harry asks how to evaluate Slack locking down its API against agentic AI threats and prompts for questions to ask Marc Benioff at dinner. Rory provides specific metrics on AI SDR and service agent resolution rates. | |
| Kalshi Quick-Fire: OpenAI vs. Microsoft Antitrust | 6 | 3 | 4 | 3 | Harry introduces a Kalshi prediction market on OpenAI suing Microsoft. Harry analyzes Sam Altman's calculated media strategy on podcasts to disrupt Meta's recruiting efforts. | |
| Kalshi Quick-Fire: US Government AI Control & Sacks Divestment | 5 | 4 | 2 | 2 | Harry reviews odds on US government AI takeover before asking whether David Sacks benefited financially from divesting assets to enter government. Rory analyzes post-election market swings. | |
| Kalshi Quick-Fire: Trump Mobile Smartphone Prediction | 5 | 4 | 3 | 3 | Harry brings up the Trump mobile smartphone Kalshi bet. Jason highlights the lack of supply chain leaks, while Harry suggests it could be a white-labeled AT&T partnership with a branded sticker. |
Statements from this episode (34)
O'Driscoll: Effective LLM implementations will be boring
“When LLMs finally work at something, the implementation will be boring as f***.”
O'Driscoll: Legal AI startup Harvey is ChatGPT in a law costume
“Harvey isn't some breakthrough in legal AI. It's chat GPT with a law costume.”
O'Driscoll: Meta won't build a major business selling LLaMA APIs
“It's not that they want to open source Lama and make money off it. They won't. It's not that they want to even have an API offering of Lama, something like, you know, a, an anthropic offering that just won't be big.”
O'Driscoll: Meta's Oculus acquisition and $60B VR spend was awful
“In one cynical sense, you know, the Oculus acquisition was awful. It was like you spend two billion, which wasn't that bad. And then you spend another sixty billion on top of it, which is pretty dreadful and you didn't get much”
Lemkin: ChatGPT mobile downloads rival TikTok, Meta, and X combined
“It was ChatGPT over the last 28 days, 29.5 million downloads. ChatGPT, 29.5 million. Just mobile. That's just its mobile app. TikTok, Facebook, Instagram and X, thirty two million. So ChatGPT mobile downloads just about equal to all the social media guys combi…”
Stebbings rejected a $75M acquisition offer for 20VC Media
“Spoiler alert that no one knows. I was offered in the last year, seventy five million dollars for 20 VC media company. Of which I own a hundred percent of it.”
Lemkin: Scale AI returned $15B to its venture capital investors
“At least Alexander from Scale gave his VCs back fifteen billion.”
O'Driscoll: Accel likely generated a $2 billion gain on Scale AI
“If Alexander Wang's highest and best use is as a senior employee at Meta, then isn't it wonderful that the free market system was able to pay everyone else, fifteen billion dollars, all power to excel who probably booked a two billion dollar gain here.”
Lemkin: Harvey was indistinguishable from 11 other AI wrappers at seed
“If I'd met Harvey at the seed stage, I would have said this is great, but I don't know that I could have told the difference from the 11 other ones that were doing this amazingly as essentially, I hate the term, but as wrappers.”
O'Driscoll: Law firms spend 5x more on legal data than software
“For every dollar lawyers spend on software, they spend five dollars on Westlaw Thomson Reuters for actual legal information.”
Lemkin: Public B2B market caps cannot justify current AI deal prices
“There aren't enough 50 to a hundred billion plus B to B public companies today to justify these deals.”
O'Driscoll: AI vendors won't capture labor value due to competition
“If there's three people competing for the same Making the same kind of software. Then the law firm gets the value. Like we don't get paid. Excel doesn't charge 60 grand a year because it replaced an analyst. It charged 60 bucks because that's what you get for …”
Lemkin: I run all legal work for SaaStr through Claude before legal counsel
“Well, all the legal work I do, both for the fund and for Saster Inc., everything, I run through my AI and Claude, and then I run it by my counsel to see if it's correct.”
Stebbings: Legal AI market will be smaller than expected due to fragmentation
“When you look at the, you know, when I was doing the diligence for Solve, patent creation, there were like, 15 competitors. As in patent creation, in every different legal adjacent, there is the same. And I think it is as fragmented and unbundled as it is. Whi…”
O'Driscoll: AI will grind down and replace mediocre knowledge work
“This is going to grind down knowledge work mediocrity. Anyone who's just really recycling stuff that's Easily known and it's just slow and unresponsive.”
Lemkin: Tech IPO volume is up 62.5% this year
“IPOs by number are up 62 and a half percent this year, apparently, based on data I saw today, right?”
Lemkin: Every tech startup with strong metrics will IPO within 12 months
“Anyone that's got the numbers is gonna go public, right? In the next 12 months, they're all gonna go public.”
O'Driscoll: Every recent tech IPO had an overpriced private funding round
“On the last, you know, five or six IPOs, there's been at least one round that was in the private side that was clearly priced wrong.”
O'Driscoll: No analytical investor believes Circle is worth 50x run-rate revenue
“I don't for a second thinks any, I don't think anyone believes who's done any kind of math analysis that it's worth 57 times run rate 50 times run rate revenue. I don't think anyone thinks that.”
O'Driscoll: Circle won't trade at 57x revenue after 6-month lockup expiration
“And I've no doubt in my mind that when the other 80% of the stock comes off in six months, I don't think it'll be trading at 57 tons revenues.”
O'Driscoll: Canva delays IPO because it generates massive cash flow
“They are, they're kicking off cash, which means that they're net buyers of their security from boys, not sellers. So they don't need to raise money.”
O'Driscoll: Larry Ellison expanded Oracle ownership to 41% via stock buybacks
“He now owns 41% of Oracle, right? What's he done? Every year he's run that business superbly. It's got 43% operating margins, and he's used that cash to buy back shares.”
O'Driscoll: Oracle abandoned stock buybacks to fund massive AI CapEx
“One is the stock really popped 40%, right? And he got a lot of cloud credit, but the interesting thing is this is the year he actually abandoned the buyback strategy, because this is the first year where instead of taking all that cash and buying shares back, …”
Lemkin: Replit grew from $10M to $100M ARR in 5.5 months
“Replit, 10 to a hundred million in 5.5 months, right, announced yesterday”
Lemkin: Clay requires a multi-month deployment and $50k agency fee
“Clay is like a six, you know, multi-month deployment period with an agency that you pay 50 grand to, right?”
Lemkin: MCP is an existential threat to B2B SaaS within 12 months
“I think MCP is an existential threat within 12 months to every B to B company.”
O'Driscoll: Slack restricting customer data access will fail over the medium term
“I don't know if you can get away forever in a horizontal app like Salesforce with a channel like Slack denying access, the denying the customer access to their own content over the medium term. I don't think it stands.”
O'Driscoll: Enterprise API lockdowns signal a decaying software empire
“I just think it's a sign of a decaying empire. It's a sign that you can compete on the merits, and it's, I, it's a little bit of a dangerous sign.”
Rory O'Driscoll: OpenAI will publicly accuse Microsoft of antitrust in 2025
“Accused, by the way, is a wonderfully vague word. Will they file a lawsuit and prevail? Maybe not, but will Sam throw words out? Yeah. That two and a half x on a yes, I'd take that chance.”
Lemkin: Sam Altman's casual public statements reveal actions OpenAI has executed
“Everything Sam says that seems off the cuff or like on the side or a little futuristic, he's very clearly telling you what's going, he's very direct. And when he says, when you hear that they're thinking about it, He's done it.”
Lemkin: OpenAI and Microsoft will settle antitrust tensions privately
“I don't think Microsoft wants to be sued by any trust. So I think it's going to get worked out.”
Rory O'Driscoll: US government will not seize control of AI companies in 2025
“I still think no. I think, I just think, you know, ah any, I think the, just the push from the folks on the tech side has been very much AI for good, not AI to control it. I think all the very active tech people from David Sachs to Andreessen Horwitz, their ap…”
O'Driscoll: US regulation targeting Chinese AI companies is highly likely
“It's far more likely to see some kind of regulation of Chinese AI companies, not obviously taking control, but some kind of pushback there. I think that's highly likely, but not US.”
Lemkin: Trump Mobile smartphone will not release before September 2025
“It's impossible. There's no supply chain evidence of any phone in production. There haven't been any leaks of an actual phone in production. And there haven't been any leaks to sites of true product development other than a mock-up of a golden phone.”