Apr 16, 2026 · 1h 30m · news
SpaceX's Financials Leaked: Is it Worth $2TN | Meta Debuts Muse Spark: Are They Back in the AI Race? · 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 tech investors Jason Lemkin and Rory O'Driscoll to dissect major shifts in the AI and tech landscape, ranging from the OpenAI vs. Anthropic rivalry and the decline of traditional SaaS to SpaceX's leaked financials and the rising impact of corporate token maxing.
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 6.9% of the talking time here. How this is scored →
Harry as informed peer 4.5 Guest teaching 5.7 Guest disagreement 3.1 Harry pushing back 3.1
speaking balance: gold is Harry, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 7:41 Jason forcefully rejects Dario Amodei's safety messaging
Jason aggressively attacks Dario Amodei's repeated apocalyptic warnings as 'endless Boy Who Cries Wolf' marketing, declaring he has completely tuned out.
Hardest push from Harry ▶ 1:16:16 Harry refuses the PE portfolio recovery narrativeHarry directly rejects soft PE explanations, interjecting that software buyout portfolios feel like a clusterfuck of pain while explicitly calling out Coupa, Anaplan, and Medallia.
Biggest teaching moment ▶ 17:47 Rory corrects Harry's premise on Amazon Tranium vs NvidiaRory cuts off Harry to correct his fundamental premise, clarifying that Amazon's Tranium is used for internal cloud hosting rather than competing as merchant silicon against Nvidia.
Harry holds his own ▶ 1:03:15 Harry breaks down SpaceX valuation multiplesHarry demonstrates strong financial fluency by detailing SpaceX's $18.5B revenue against its $2TN valuation target to calculate an exact 108x revenue multiple.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Harry as informed peer | Guest teaching | Guest disagreement | Harry pushing back | Why |
|---|---|---|---|---|---|---|
| Anthropic's Claude Mythos and Cybersecurity Impact | 3 | 6 | 2 | 3 | Harry sets up the prompt asking about market fear around Anthropic's Mythos release. Rory and Jason educate the host on agentic code scanning, contrasting rifle versus machine gun security risks. | |
| Debating Dario Amodei: Marketing Stunts vs. The Power of Vision | 5 | 5 | 6 | 5 | Harry presses Jason on his cynicism toward Dario Amodei, asking specifically what he finds uninspiring. Jason strongly rejects Dario's safety warnings while Rory pushes back defending visionary grandiosity. | |
| Amazon's Tranium Chips and Nvidia's Stronghold | 3 | 8 | 5 | 2 | Harry posits that Amazon's Tranium chips are weakening Nvidia's market dominance. Rory abruptly interrupts to correct Harry, explaining Amazon's internal cloud consumption model versus merchant silicon competition. | |
| Code Generation Competition: Anthropic vs. Cursor, Replit, and Lovable | 3 | 6 | 4 | 4 | Harry asks if Claude Code threatens vibe-coding tools like Lovable. The guests lightheartedly tease Harry for mistaking product announcements for shipped software. | |
| The '60% Solution' and the Decline of Incumbent SaaS | 5 | 7 | 3 | 3 | Jason lays out his thesis on incumbent SaaS death spirals due to '60% solutions'. Harry anchors the market discussion by providing exact figures on Wix's stock buyback and subsequent share drop. | |
| Meta's Muse Spark: Is Zuckerberg Back in the AI Race? | 4 | 5 | 2 | 2 | Harry introduces Meta's Muse Spark model post-Scale AI acquisition. Guests analyze Zuckerberg's $14B strategic move to remain competitive in foundation models. | |
| OpenAI's Ad Revenue and the Shift to Enterprise | 6 | 5 | 3 | 3 | Harry demonstrates strong preparation by citing OpenAI's internal ad revenue projections and pilot figures. Rory reframes consumer ad revenue limits relative to enterprise market cap requirements. | |
| Compute Scarcity and the Future of AI Development | 6 | 4 | 2 | 3 | Rory explores compute allocation and consumer vs enterprise utility. Harry offers consumer research findings showing younger demographics seeking emotional support from ChatGPT. | |
| Token Maxing, CIO Budgets, and Enterprise Standardizing | 5 | 5 | 2 | 2 | Jason outlines Aaron Levy's insight regarding CIO token maxing. Harry highlights his booking responsiveness by mentioning he scheduled Levy for the podcast following the thread. | |
| SpaceX's Financials Leaked: Is it Worth $2 Trillion? | 6 | 6 | 3 | 4 | Harry presents leaked SpaceX metrics, accurately calculating the 108x revenue multiple on a $2TN valuation. Rory breaks down Elon discount rates and probability of failure accounting. | |
| Lean Tech: Employee Efficiency and the Rise of AI Agents | 5 | 6 | 3 | 3 | The group discusses employee efficiency metrics like AppLovin's revenue per head. Harry shares his team evaluation framework, which Jason reframes around replacing team members with agents. | |
| Understanding Thoma Bravo and Private Equity Software Models | 3 | 7 | 2 | 2 | Harry asks about Thoma Bravo shutting down its growth equity business. Rory educates both Harry and Jason on PE fund structures and the difference between LBO control bets and minority growth deals. | |
| The Private SaaS Crisis and AI Integration Playbook | 5 | 7 | 4 | 5 | Harry challenges the optimistic PE narrative by highlighting troubled software buyouts like Coupa and Anaplan. Jason and Rory analyze why legacy SaaS portfolios are failing to execute on AI agents. | |
| The Race to IPO: Anthropic, SpaceX, and OpenAI's Alignment | 5 | 5 | 3 | 4 | Harry asks direct, binary questions regarding IPO order and executive retention at OpenAI. Rory explains executive alignment using a police interrogation room analogy. | |
| Managing Executive Dysfunction and Board Politics | 4 | 4 | 2 | 1 | Jason warns VPs against reaching out to board members to complain about CEOs. Harry smoothly wraps up the episode by citing a fitting Shakespeare quote. |
Statements from this episode (45)
Lemkin is burnt out on Anthropic CEO Dario Amodei crying wolf
“I don't buy Dario anymore. He may well be the second greatest founder of all time behind Elon, but I am just so burnt out of the boy who cries wolf.”
O'Driscoll: AI race is two-way fight between Anthropic and OpenAI
“I think it's a two-way fight. Antropic has the advantage of clarity and focus. OpenAI has the advantage of the consumer business.”
Lemkin: Subpar AI agents face a slow death spiral
“If your agents are only 60% as good, you're in a slow death spiral.”
O'Driscoll: SpaceX's $2T valuation assumes a zero percent failure rate
“The Elon discount rate Is zero and the Elon probability of failure rate is zero to get to two trillion.”
O'Driscoll: Skeptics are wrong to dismiss Anthropic's Claude Mythos capabilities
“Those people who said it's not a big deal are wrong and entropic is right.”
Lemkin: CalAI suffered a massive breach days after $100M MyFitnessPal acquisition
“My fitness pal bought CalAI, right? Cool story, right? What was it, Harry? A hundred million dollars, 19 year old kid from Miami, something like that, right? A great story. Two days later, it was instantly breached. All the records were stolen. 3.2 million rec…”
Lemkin: Cybersecurity will worsen during a transitional phase as AI accelerates hacking
“As most databases are now built by AI, as more and more apps are built by AI the number of issues is going to explode, and if Mythos and Friends lets bad actors find every site the second it launches with any PAI and steal it, I think we may enter an era later…”
O'Driscoll: Selling off cybersecurity stocks over AI hacking models makes no sense
“The part that made sense was the, this is a big deal. The part that didn't make sense is the cyber stock should go down because I think you're going to want way more defenses because the bad guys are more heavily armed.”
Lemkin: Dario Amodei is the second-greatest founder behind Elon Musk
“What I mean is listen, he may well be the second greatest founder of all time behind Elon. Look what he's done in five years, right? Five years to thirty billion, the grit, the greatest grudge startup of all time.”
Rory O'Driscoll: Scale passed on early Anthropic deal over pricing and doomerism
“What I realized is if I'd met them at the sea, which I didn't, because it was outside our price bracket, but if I'd met them, I would have done exactly what you did, Jess, and I would have listened to the doom warnings, and I said, that's all silly and wrong, …”
Rory O'Driscoll: AI will not displace 50% of white-collar workers
“I don't think we're going to unemployed, 50% of white collar workers. I think it's madness. I think it has some legitimate dangers in cyber security and bioterrorism, but they're manageable.”
Rory O'Driscoll: Evaluate founder vision by motivational power, not literal accuracy
“So what I've learned to do, which is really hard, is literally listen to the idealism Don't say, do I agree or not? Say to myself, will it motivate people enough to do something where there is economic advantage to be obtained?”
Lemkin is ready to write off portfolio startups failing to adopt AI
“Like you know, I'm ready to move on to the new world and I'm almost, and I'm frankly ready to write off a lot of portfolio companies and a lot of public companies. It's time to move on guys. If you're not going to get there in April of 20, 26, then so be it. S…”
O'Driscoll: Amazon's $20B Trainium business diverts nearly 10% of Nvidia's revenue
“But all that said, it's still twenty billion of dollars that didn't go to Nvidia that went to Amazon. So it is at the margin meaningful. It's 10% of kind of Nvidia's revenue, a little less than 10%. So it's, and so it's not a mega competitor. It's just an in-h…”
Lemkin: Incumbent SaaS cannot monetize mediocre check-the-box AI agent features
“You checking the box does not work with agents. The check the box feature cannot be monetized in the AI era.”
O'Driscoll: SaaS revenue re-acceleration requires standalone-quality AI products
“Unless you have a product that's good enough to charge for independently, You won't have revenue re-acceleration of any meaningful scale.”
Lemkin refuses to buy public SaaS stocks with sub-par AI solutions
“No, I wouldn't buy any of them. I wouldn't buy, because I'm looking for products that are more than a 60% solution.”
O'Driscoll: Public SaaS stocks will not recover until AI leaders IPO
“These stocks aren't going to trade in the same fashion until five or six public of these two of the foundation models and four or five other AI gen, AI native companies are public.”
O'Driscoll: Salesforce risks becoming another IBM if its AI agents fail
“Unless it gets its shit together and makes great agents, it's another IBM.”
O'Driscoll: AI market blip within two years will enable major acquisitions
“And there's some chance that in the next two years, there's a blip in the market. All these AI companies are burning money. And if you're sitting there with twenty five billion dollars, you could have been saying, come to daddy. I got money. Let's talk about h…”
O'Driscoll: Meta's pivot to closed-source AI harms the ecosystem
“Also worth noting that they're talking about being much more closed source, which is a significant thing, because at some point someone's going to need the American version of open source, and Lama was that, and now they're pivoting more to be more closed sour…”
Lemkin: Meta is the most ruthlessly competitive company on Earth
“Ruthless, this is as competitive a company as exists on planet Earth, right? This might be the most competitive company on planet Earth is Meta, right? Ruthless.”
O'Driscoll: Meta lost $70B on VR after making $100B+ on Instagram
“Yeah, I mean, if you look at the, if you look at the big play scoreboard for that, it's like, bet a billion on Instagram, win a hundred billion plus, maybe four hundred billion. Bet seventy billion on Meta and VRAR, lose it all.”
Lemkin: Google Search, Facebook, and Instagram are stronger than ever
“Fast forward today. These are record growth businesses now, right? Google search. We started the show talking about AI overviews and other things. Google search is a better business than it has ever been as is Facebook and Instagram.”
Rory O'Driscoll: OpenAI ChatGPT must inevitably be ad-supported
“It's both obvious and inevitable that a consumer product like OpenAI ChatGPT is going to have to be ad supported.”
O'Driscoll: Even $100B in ad revenue will not cover OpenAI's burn
“Consumer alone ain't going to be enough to feed this beast, because your competitor, who's all in on enterprise, Is, you know, already at 30, right? And you're, so that was my, as I say, I almost feel like a jerk saying it's like, hey, congratulations on your …”
Jason Lemkin predicts OpenAI will reach $100B in ad revenue
“I like this as a goal for 2030 because it's clear what people should be doing... And so I think because it is achievable, it will be achieved. I actually think it will be achieved.”
Jason Lemkin: OpenAI's top-down enterprise strategy will prevail
“I'm pretty bullish actually on opening in the enterprise because all this enterprise DNA they have, which probably didn't help a snail's inch the last 12 months in the future when all the models are so powerful and big enterprises are trying to make decisions …”
O'Driscoll: Tech giants won't cut AI spending in 2026, keeping Nvidia sold out
“No one's going to blink on their investments for twenty-twenty-six. You actually know the next four quarters of NVIDIA announcement. You know the next four quarters of every one of the infrastructure, which is every single thing we can make. If you can count t…”
O'Driscoll: AI labs will throttle subscriptions and hike prices for compute
“You're going to see some compute rationally, people are going to start allocating tokens to the highest, effectively the highest bidder. You're going to see a lot of throttling, like, you're going to see these plans, like, that's why they shot Sora. You're goi…”
O'Driscoll: AI value creation will be two-thirds enterprise, one-third consumer
“Enterprise is two thirds of the ball game in AI and consumer is one third or less, which is the flip of the last time.”
Lemkin: CIO-driven AI budgeting will shift enterprise vendor choices
“All the enterprise stuff is still in some sense rogue today. So much of it is rogue. It is out of budget. It is out of band, and as that changes, it will change which vendor we buy from.”
Lemkin: If Box cannot re-accelerate growth, incumbent SaaS is hopeless
“If he can't get Box to 20 to 30% growth I'm giving up on, on the rest of the world.”
Jason Lemkin: Building AI agents will become accessible to everyone in 18 months
“Everyone in 18 months will figure out how to build an AA, because they'll get easier and easier to train.”
O'Driscoll: Cursor loses money despite generating $4M per employee
“Cursor does four million per employee. Salesforce does 700 per 700,000 per employee. Oh, Cursor must be more efficient. Well, it turns out Salesforce has 30% operating margins and Cursor's losing a lot of money. Why? Because they spend a whole ton on tokens, r…”
Rory O'Driscoll: AppLovin is a highly efficient money printing machine
“It's a four and a half million per, whatever it is, per employee business, where unlike Atropic, our open ad, no CapEx, unlike Cursor, no token cost, it's just a money printing machine.”
Rory O'Driscoll: Software companies must continuously raise revenue per employee
“If you're not making progress on that metric, as a software company, you are not with the program.”
Rory: PE firms enter growth venture at market peaks and retreat when it gets hard
“PE guys regularly come in to growth venture at the top of markets thinking this looks easy, and they regularly retreat from those markets when they discover it's hard, right?”
O'Driscoll: Mature private SaaS companies trade at 2x-4x revenue multiples
“The first thing is these kinds of companies are trading, excuse me, at two to four tons of revenues.”
Lemkin: Hiring external AI consultants won't save private equity SaaS deals
“But Stebbins, Driscoll, Lemkin, AI, Consulting is not going to get him there. And that's who PE firms want to do. Bring in these guys. It's not going to work.”
O'Driscoll: SpaceX, Anthropic, and OpenAI will IPO in that exact order
“SpaceX, Anthropic, OpenAI in that order.”
Jason Lemkin: Anthropic adding Novartis CEO signals an imminent IPO
“The fact that Anthropic just added the Novartis CEO to the board, that's a sign they're getting ready to IPO as soon as they can.”
O'Driscoll: Founders should innovate on product but standardize corporate governance
“On the thing where you're unique and different, you know, where you're changing the world, do it as different as you like, but all the boring stuff, just give the market what it wants.”
Lemkin: You must fire any executive who publicly argues with the CEO
“If you have an executive that's truly arguing with the CO in public, in the media, right? And or like Dario back in the day at OpenAI going directly to the board with craziness, they gotta go. Like, it doesn't matter. They gotta go. They, you cannot be out in …”
Lemkin: VPs who complain to VCs about their CEO will be fired
“And whatever, if any VPs get this far on the pod, whatever you do, do not reach out to your VCs to say there's problems with their CEO. You will, you're losing your job. And not only are you losing your job, it doesn't matter if you're right. You're your, at s…”