Aug 7, 2026 · 1h 15m · allin
Google’s AI Brain Drain, SpaceX's Huge Quarter, Airtable’s 90% Collapse, US Data Fuels China AI
David Sacks · 21m spoken Jason Calacanis · 20m spoken Brad Gerstner · 16m spoken David Friedberg · 10m spoken Chamath Palihapitiya · 4s spoken
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
The All-In Podcast hosts analyze major shifts in the technology landscape, focusing on Google DeepMind's leadership shakeup, SpaceX's post-IPO financial performance, Airtable's dramatic valuation collapse, and the geopolitical implications of U.S. data fueling Chinese AI development. Through these topics, the panel evaluates the changing economics of artificial intelligence, enterprise software, and global tech competition.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. The hosts hold 75.9% of the talking time here. How this is scored →
The hosts as informed peer 6.4 Guest teaching 2.4 Guest disagreement 2.0 The hosts pushing back 2.9
speaking balance: gold is the hosts, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 18:55 Gerstner Rebuts Calacanis on Model Commoditization
Gerstner directly counters Calacanis's assertion that open-source models have rendered closed frontier models obsolete, citing Elon Musk and Jensen Huang to argue that the intelligence and cost gap remains wide.
Hardest push from the hosts ▶ 1:11:35 Sacks Refuses Framing on AI Training Data Export BansSacks repeatedly interrupts and challenges Calacanis's premise that selling labeled data to Chinese companies represents a dangerous leak of proprietary national security secrets.
Biggest teaching moment ▶ 40:22 Gerstner Breaks Down Realities of Data Center Debt FinancingGerstner educates the hosts on the financial engineering behind multi-gigawatt facilities, explaining why Nvidia shareholders resist open-ended backstops and how circular vendor financing introduces systemic risk.
The host holds their own ▶ 1:01:34 Sacks Details Enterprise Compliance Moats Protecting Legacy SaaSSacks demonstrates deep domain expertise in enterprise software, citing DoD clearances, FedRAMP authorizations, and Active Directory dependencies to show why enterprises cannot simply replace core SaaS with vibe-coded solutions.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | The hosts as informed peer | Guest teaching | Guest disagreement | The hosts pushing back | Why |
|---|---|---|---|---|---|---|
| Google AI Shakeup & Infrastructure CapEx Shift | 7 | 1 | 1 | 1 | David Friedberg provides an in-depth operational and financial analysis of Google's AI leadership shakeup, explaining the high return on invested capital for CapEx in compute infrastructure versus the high-beta risk of model training. The discussion is collaborative and host-driven with minimal disagreement. | |
| Frontier AI Duopoly & Corporate Channel Conflicts | 7 | 2 | 1 | 2 | David Sacks articulates a structured view of the frontier AI duopoly between Anthropic and OpenAI, drawing parallels to the Apple versus Android monetization dynamic. Brad Gerstner agrees and expands on the channel conflicts faced by cloud hyperscalers attempting to both host compute and build competing models. | |
| Open Source vs. Proprietary AI Models | 6 | 4 | 4 | 5 | Jason Calacanis claims open-source models make frontier models unnecessary for 95 percent of enterprise use cases. Both Sacks and Gerstner push back, citing recent counterarguments from Elon Musk and Jensen Huang regarding frontier capabilities and the true total cost of ownership of closed models. | |
| SpaceX Q2 Earnings & Compute Revenue Growth | 7 | 2 | 1 | 1 | The panel reviews SpaceX's Q2 earnings report and compute revenue growth. Gerstner contextualizes the post-IPO valuation contraction, while Sacks breaks down the unit economics of data center power pricing at thirty to fifty dollars per watt. | |
| Starlink Economics & The TeraFab Vision | 7 | 1 | 1 | 1 | Friedberg details the financial strength of Starlink, highlighting its 2.6 billion dollars in adjusted EBITDA and projecting thirty billion dollars in annual free cash flow. The hosts and guest collaboratively praise SpaceX's vertical integration and manufacturing execution. | |
| AI Data Center Financing & Seller Financing Risks | 7 | 5 | 3 | 4 | Sacks questions how SpaceX and other data center operators can fund massive multi-gigawatt buildouts without severe dilution. Gerstner walks through the mathematical reality of fifty billion dollars per gigawatt CapEx, explaining the risks of circular seller financing and shifting payback horizons. | |
| Announcement and Preview of the Fifth Annual All-In Summit | 2 | 0 | 0 | 0 | The hosts and guest preview the upcoming All-In Summit, announcing speakers including Jensen Huang and Satya Nadella. This is a promotional and housekeeping segment with no debate. | |
| Airtable's Acquisition and the State of the SaaS Market | 8 | 4 | 4 | 5 | The group analyzes Airtable's fire-sale acquisition by Bending Spoons. Sacks argues the acquirer can strip ninety percent of costs using AI-driven maintenance to generate massive EBITDA, while Gerstner cautions that slowing SaaS growth leads to severe customer churn and developer attrition. Sacks counters by citing enterprise compliance moats that protect core SaaS platforms from vibe coding. | |
| US Data Sourcing and Training Chinese AI Models | 7 | 3 | 3 | 7 | Calacanis raises national security concerns over American startups selling high-end RLHF training data to Chinese AI labs. Sacks forcefully rejects the alarmist framing, arguing that expert data labeling is not a proprietary military technology and that China produces sufficient domestic STEM talent to generate equivalent data. |
Statements from this episode (34)
Friedberg: Google committed $200B in CapEx to AI data centers this year
“Google has made a commitment to deploy two hundred billion dollars in capex this year in AI infrastructure data center build out.”
Friedberg: Google owns significant stakes in SpaceX and Anthropic
“They have a significant ownership stake in SpaceX and in Anthropic, and they can work with all the open weights models.”
Friedberg: Data center CapEx is low beta while model development is high risk
“CapEx is high alpha, low beta in data center infrastructure, that capital, and model development theoretically could be high alpha, but it's very high beta. It's a very risky way to deploy capital.”
Gerstner: Nadella reports over 30% ROIC for Microsoft token infrastructure
“Satya is out this week saying, you know, citing Morgan Stanley's report and saying they're seeing over a 30% return on invested capital in tokens as a service, right?”
Sacks: Frontier AI intelligence is now an Anthropic-OpenAI duopoly
“Now we're really down to just anthropic and open AI. So the market for frontier intelligence has become a duopoly.”
Sacks: Only frontier model providers can charge a premium for AI weights
“If you're not at the frontier, you can't charge for the model layer itself. If you are at the frontier, you can charge a premium, and that's where Anthropic and OpenAI”
Sacks: Anthropic surpassed $80B ARR and will beat year-end targets early
“The latest we heard is Anthropic is now over eighty billion of ARR. Started the year at 10. It had forecast a hundred billion as exit ARR for the year, and most people said that that would be impossible to achieve. Now it looks like they're gonna do it with a …”
Calacanis: Google Cloud posted record 82% year-over-year revenue growth
“Well, if you look at Google Cloud, they posted 82% year-over-year revenue growth, which is something we've never seen in the history of these cloud providers.”
Calacanis: Gemini surpassed 950 million MAUs in Q2
“In Q two, Gemini had over 950 monthly active users tripling year over year.”
Calacanis: Google will lead in consumer AI usage by far this year
“They will be the number one AI company in terms of consumer usage by far, I think, this year.”
Gerstner: Open source captures token volume while frontier labs capture economics
“For the vast majority of use cases, I think token consumption is going up for the open source guys while share of economics is going up for the frontier labs.”
Calacanis: SpaceX Q2 revenue hit $7.8B; Elon Web Services hit $2.6B
“7.8 billion in revenue up 92% year over year. Let that sink in. And 67% quarter over quarter AI revenue. Elon web services more than tripled quarter over quarter to 2.6 billion dollars.”
Gerstner: Combined Cursor and Grok could reach $10B–$20B by year-end
“Cursor was already on a path to go from three billion to ten billion by the end of the year. Cursor plus Grock could be at 10 to twenty billion by the end of the year.”
Friedberg breaks down Starlink Q2: $4.3B revenue, $2.6B EBITDA, 12M subscribers
“But the connectivity piece on Starlink, 4.3 billion in the quarter, and 2.6 billion in adjusted EBITDA. He's got twelve million subscribers. That's doubled year over year. 66 dollar ARPU per month what people are paying per month. And he grew 20% quarter over …”
Friedberg: Starlink could generate $30B in annual free cash flow within a year
“Starlink alone could be generating on the order of forty billion dollars of revenue top line with a huge amount of that flowing to free cash. That could be a thirty billion dollar free cash flow within the year.”
Friedberg: Starlink alone could reach a $1T valuation within 18 months
“The Starlink business alone could be a trillion dollar market cap within two years, within 18 months, let's say.”
Friedberg: Direct-to-cell Starlink could double quarterly subscriber additions to 4M-5M
“And if he launches the handset thing, that subscriber growth is going to go, right now he's adding two million subscribers on the consumer side a quarter. You could see that going to four to five million a quarter.”
Calacanis: Every Tesla sold will include built-in Starlink connectivity
“So the piece there that's going to be explained probably in the next year or two is every single Tesla sold is going to have Starlink in it when they get that merger done. What that means is you're going to have wifi networks connecting any phone. To any Tesla…”
Gerstner: SpaceX investors can triple their money over 3 to 4 years
“And now you have a company at 1.4 trillion that I think if you take a three or four year view, you can see yourself tripling your money in this business at a very reasonable valuation on the Morgan Stanley numbers or on the Elon numbers or whatever.”
Sacks: Starship launches will add over 20x more Starlink bandwidth
“So basically, SpaceX has developed a new V-III satellite that has 10 X the bandwidth of its V-II satellite. So currently the Starlink network is powered by V-II satellites They deploy them on the Falcon nine rocket, and they launch about 27 satellites per laun…”
Gerstner: Frontier AI labs pay 3x to 5x market price for compute
“Right now the shortage is so acute, and the willingness to pay from the Frontier Labs is so high because they all recognize they're on the verge of some massive breakthroughs. That they're willing to pay three, four, five X market pricing in order to get at sc…”
Gerstner: Vast majority of AI compute offtake comes from 3 companies
“So the vast majority of the offtake commitments are coming from Anthropic, OpenAI, and NVIDIA. Right. When you hear about the hyperscalers building all of this, you know, this computer, they're building it out to sell to the people that we you know, that we ju…”
Gerstner: Financing capacity ahead of revenue will make any demand drop violent
“Everything will trade down, you know, together because that's just the leverage that you're pumping into the system. You're effectively backstopping people's ability to build ahead of their revenue. So it becomes much more violent if you ever see demand slippa…”
Gerstner: Credit spreads remain wide on AI infrastructure deals
“Credit spreads are blowing, you know, have continued to stay wide on these deals. So there is fear in the market about them.”
Gerstner: Anthropic could potentially IPO around October 2026
“We're gonna be within 30 days of an IPO, you know, potentially of Anthrop.”
Sacks: Airtable spun out Hyper Agent prior to Bending Spoons acquisition
“Airtable spun out its AI agent business, which is known as Hyper Agent. Into a separate independent company prior to this acquisition.”
Gerstner: Software companies cannot keep top talent from fleeing to AI
“It's almost impossible for a software company today to keep any decent sales people to keep any different, decent product development people, because they all want to go work on AI.”
Sacks: AI makes software maintenance easy without institutional code knowledge
“I think maintenance mode becomes way easier with AI because you don't need the historical knowledge anymore. The AI can go in and sort of reconstitute that, that historical knowledge.”
Sacks: No-code software is the most disrupted SaaS category
“No code has to be the most impacted, the most disrupted area of SaaS right now, because, I mean, what is Claude Code really good at? I mean, that's the ultimate no code tool.”
Gerstner: Altimeter passed on Airtable's last three funding rounds
“We passed on this in the last three funding rounds, right? Which I think we're at 2,000,000,005 1,000,000,011 billion.”
Sacks: AI Data Labeling Is a Commodity China Can Easily Replicate
“My sense of data is that it's largely a commodity. I mean, data labeling certainly is. If you basically tell them that they can't use data labeling, I guarantee you there's no shortage of Labor in China that they can use to do the data labeling.”
Gerstner: U.S. Will Only Restrict China AI Data If Lead Vanishes
“The reason I don't think it will cause us to change our stance with respect to China is because we're winning. But if the president asks his advisors, you know, one of these days, six months down the line, are we winning against China? And all of a sudden he g…”