Aug 13, 2026 · 1h 39m · allin
Anthropic's $2T IPO, Zuck's AI Manifesto, Nvidia's $500B AI Bet, Grok's Comeback
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 and guest Gavin Baker discuss Anthropic's potential $2T IPO, Mark Zuckerberg's open-source AI manifesto, Nvidia's $500B GPU financing framework, physical energy constraints in AI scaling, and recent software market developments.
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 67% of the talking time here. How this is scored →
The hosts as informed peer 6.4 Guest teaching 6.0 Guest disagreement 2.4 The hosts pushing back 2.6
speaking balance: gold is the hosts, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 1:12:05 Gavin Slams Macro Bear Investors as Completely Ignorant
Gavin forcefully dismisses macroeconomic and value investors predicting an AI crash, stating their underlying thesis of subsidized token pricing is factually false.
Hardest push from the hosts ▶ 1:23:40 Sacks Dissects Jason's Argument on Amazon DSP RegulationsSacks refuses to let Jason's corporate responsibility framing pass, pointing out the logical contradiction of praising DSP flexibility while simultaneously calling for municipal bans.
Biggest teaching moment ▶ 1:01:10 Gavin Details the Architecture of Nvidia's $500B Financing FacilityGavin educates the hosts on how Nvidia provides residual value guarantees and takes upside royalties, effectively turning Nvidia into the Federal Reserve of AI infrastructure.
The host holds their own ▶ 1:09:40 Sacks Traces the Multi-Tier AI CapEx Contagion CascadeSacks demonstrates deep structural market expertise by breaking down the exact dollar flows per gigawatt from frontier labs down to hyperscalers, Nvidia, and semiconductor foundries.
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 |
|---|---|---|---|---|---|---|
| Show Opening & Rahm Emanuel Discussion | 4 | 2 | 1 | 2 | The hosts and Gavin open with light banter about SpaceX and the recent interview with Rahm Emanuel. Sacks and Jason offer political commentary regarding the moderate versus progressive wings of the Democratic Party in total alignment. | |
| Anthropic's Rumored $2T IPO & Exponential Revenue Growth | 7 | 6 | 2 | 3 | Gavin breaks down why leaked IPO valuations like $2T often stem from bankers attempting to undercut the lead-left underwriter. Sacks adds detailed context on Anthropic's ARR run rate, OpenAI's pivot to coding models, and month-over-month revenue growth. | |
| AI Infrastructure Constraints: Energy, Data Centers & Media Narrative | 7 | 7 | 3 | 2 | Gavin provides granular operational detail on physical power bottlenecks, repurposed jet turbine engines, and media misconceptions regarding data center water and power usage. Sacks and Jason reinforce this with data on local grid economics and gas turbine buildouts. | |
| Mark Zuckerberg's AI Manifesto & Centralized vs. Decentralized AI | 8 | 6 | 4 | 3 | Sacks and Gavin critique the effective altruist AI doomer philosophy, citing historical centralization failures, while praising Mark Zuckerberg's open-source manifesto. Sacks recounts firsthand discussions in Washington regarding regulatory cartelization. | |
| Mark Zuckerberg's AI Manifesto and Direct Communication Strategy | 5 | 5 | 1 | 1 | Jason and Gavin explore the strategic imperative for tech CEOs to communicate directly on X rather than relying on corporate PR teams. Both agree direct communication provides personal and strategic protection against regulatory scrutiny. | |
| Zuckerberg's Physical Rebrand and Public Aura Maxing | 5 | 5 | 1 | 2 | The conversation shifts from Zuckerberg's public rebrand and physical stunts to Gavin's analogy comparing frontier versus open-source models to Oppenheimer orchestrating Nobel laureates and working physicists in the Manhattan Project. | |
| The $500 Billion Anthropic Revenue Over/Under Debate | 6 | 5 | 3 | 3 | Jason presses Gavin and Sacks on whether Anthropic will surpass $500B in ARR by exit 2027. Both Gavin and Sacks take the slight under or near-500B range due to physical compute constraints while noting the sheer magnitude of such growth. | |
| AI Compute Bottlenecks and Human Energy Efficiency | 6 | 7 | 2 | 2 | Gavin highlights the vast energetic efficiency of the human brain compared to megawatt data centers, arguing this economic advantage guarantees long-term demand for human intelligence. Jason and Gavin then debate whether Meta or Nvidia will lead open-source AI. | |
| Nvidia's $500B AI Infrastructure Financing Strategy | 6 | 8 | 2 | 2 | Gavin explains Nvidia's $500B financing framework with Wall Street asset managers, outlining residual value guarantees and GPU lifecycle longevity extending up to nine years. Sacks and Jason validate the mechanics of treating compute as a standardized financial asset. | |
| GPU Glut Risks and Data Center Build-Out Bottlenecks | 8 | 8 | 3 | 3 | Sacks outlines the theoretical risk of a GPU glut and maps out the downstream financial chain reaction if a pace car like Anthropic slows compute purchases. Gavin forcefully pushes back against macro investors who erroneously assume AI token generation is unprofitable. | |
| Legal and Regulatory Challenges to Amazon's DSP Model | 8 | 5 | 6 | 8 | A heated clash between co-hosts ensues over Amazon's DSP delivery model. Jason argues Amazon exploits subcontractors and externalizes healthcare and liability costs onto taxpayers, while Sacks vigorously defends free contract, citing massive consumer cost increases and job losses if DSPs are banned. | |
| Grok 4.6 Performance, Cursor Acquisition, and xAI's Comeback | 7 | 7 | 2 | 2 | Gavin presents benchmark data demonstrating Grok 4.6's Pareto-optimal pricing and capability on CursorBench and Databricks. Sacks and Jason dissect Elon Musk's strategic dual-option of training frontier models while maintaining compute rental call-and-put options. | |
| Workday PE Buyout Rumor and the Software Market Recovery | 6 | 7 | 1 | 1 | Gavin reports on the breaking news of Silver Lake's buyout bid for Workday, explaining how private equity establishing a valuation floor combined with open-source AI proliferation relieves margin pressure across enterprise software. |
Statements from this episode (35)
Sacks: Rahm Emanuel has no chance in modern Democratic Party
“Which means he has no chance whatsoever in the Mondani DSA Democrat Party.”
Gavin Baker: Generally avoids investing in Elon Musk's competitors
“I don't generally invest in Elon's competitors. I don't think it's a wise course of action.”
Baker: Anthropic is losing marginal market share to OpenAI and Grok
“And what's wild to me is that probably on the margin, Anthropic is losing share to OpenAI, to open source, and to Grok, and they're still growing so fast.”
Sacks: OpenAI revenue growth exceeds 20 percent month-over-month
“I've heard that over the past two months, that growth rate is now over 20% month of a month, which if you extrapolate it out over 12 months would be a 10 X growth rate.”
Baker: Software coder job openings are higher today than a year ago
“There are, you know, there are more job openings for software coders today than there were a year ago. There's more demand for coders, and that is the kind of in the epicenter of the blast zone.”
Baker: Data centers are repurposing private jet engines for power
“Residuals for private jets are, like, extremely strong these days. And a big part of the reason is people are taking jet engines off old planes and repurposing them as turbines to power data centers.”
Sacks: New AI data center scaling will rely entirely on natural gas
“The green stuff can't scale in the short term. It's all gas powered for the new data centers. That's the only way to kind of get the amount of scale that you need in the next year or two.”
Baker: Elon Musk is building a $10B solar plant in Texas
“Elon's building a ten billion dollar new solar manufacturing plant in Texas.”
Calacanis: Anthropic's revenue growth will slow and miss $1T next year
“I think Anthropics growth rate is going to slow considerably next year. And it's been obviously on a tear. There's no way they get to a trillion in revenue next year.”
Baker: Banning open-source AI means losing the AI race to China
“If we ban open source, if we let this control be centralized, we will lose AI, we will lose geopolitics, we will lose everything to China. It's just that Simple.”
Calacanis: Meta sued every startup attempting to index its social graph
“Zuckerberg, please explain how this concept is compatible with the fact that you have sued every startup, many of which I've invested in, that has ever tried to index or use the social graph at Instagram or Facebook.”
Sacks: Anthropic's premium pricing relies entirely on a six-month lead
“Their whole pricing power, that premium they're able to charge is because They're six months ahead of those open models, and that six months would be eaten up in the blink of an eye if they were subject to regulatory approval.”
Sacks: Anthropic's Pricing Power May Be Sustainable for Frontier AI
“So now, look, I actually think that their pricing power may be quite sustainable, because I think people are willing to pay, not the whole market, but a decent subset of the market is willing to pay a premium for true frontier intelligence.”
Baker: Open-source AI models make frontier AI tokens more valuable
“Open source makes frontier tokens more valuable.”
Baker: Frontier models will capture 65 to 85 percent of value
“I think a very plausible outcome is one in which frontier tokens Or 65 to 85% of the economic value. Open source tokens are like 80% of the volume, and that's good for everyone, and everyone wins.”
Baker: Anthropic reaching $400B to $500B revenue by 2027 is realistic
“I think four to five, I think four to five hundred billion is within the zone of like realistic and achievable.”
Sacks: A global compute governance agreement with China will never happen
“These EA types thought they're gonna institute global compute governance. Somehow they're gonna get, like, Some global agreement with China. Total pipe dream. Never going to happen. If we slow down, China is just going to race ahead.”
Baker: Approximating a human brain requires the power of 1M homes
“To approximate our brains, you need to have a data center that consumes the power of a million American homes running For probably six to nine months, and then you get our brains, and then every time you want to answer a question, you need the power of like a …”
Baker: Capitalism Will Always Value Human Intelligence Over AI
“Capitalism, there, it's, there's gonna be a use for the most computationally efficient intelligences on planet Earth, and those are human intelligences.”
Calacanis: AI chip inference efficiency will improve 5x-10x every 12-18 months
“If you look at what Cerebrus and Grok and NVIDIA are up to, they're making inference 10 times, five times, every year, and you're the chip master here, Gavin, but they'll be five to 10 times more efficient every 12 to 18 months.”
Calacanis: Zuckerberg will have the best open-source AI model within a year
“He is going to have the best open source model in the next year. I guarantee it.”
Baker: Nvidia might reach the frontier of open-source AI models
“I think you might see Nvidia at the open source frontier.”
Calacanis: Unnamed hyperscaler pivoted a nine-figure AI strategy to open source
“One of the hyperscalers told me they just did a nine figure Double Lindy reverse, where they went from doing a frontier model on 100 billion dollars in infrastructure, and they're like, bing, it's gonna be open source, not frontier.”
Calacanis: Nvidia partnering with Goldman and BlackRock for $500B financing
“NVIDIA is partnering with Goldman, BlackRock, and others to raise five hundred billion in AI compute.”
Baker: Divergent Chinese open-source model architectures favor Nvidia GPUs
“If you look at the underlying architectures of the three, what I call big Chinese open source models, and maybe even through a few, or four, You know, if we have Quinn, if we have Kimmy, if we have DeepSeq, and then we have GLM, they're actually evolving in ve…”
Baker: CoreWeave is profitably renting 2020 Nvidia Ampere GPUs through 2029
“Corweave said that they are renting Amperes at economically profitable rates today for 2029. So an Ampere that came out in 2020 is going to have a nine-year life, and I think that's what Nvidia sees”
Sacks: Political headwinds will prevent an oversupply of AI compute
“All the political headwinds, I think will almost guarantee that there's not an oversupply relative to the exponentially growing demand.”
Baker: Nvidia is functioning as the central bank of AI
“In some ways, NVIDIA is kind of becoming the central bank of AI. The Federal Reserve of AI. That is like a way to kind of conceptualize what they are doing. And in the same way, the federal reserve, like a lot of monetary policy, it works through private banks…”
Baker: Anthropic is profitable and currently generating cash
“Anthropic is generating cash, and like, they kind of, this fact is, I would say, kind of known in the world, and I've actually never really spoken to Anthropic, but it just, you know, it feels like there are enough signals And then if you're to read the tea le…”
Baker: The overwhelming majority of AI tokens are profitable
“The overwhelming majority of tokens are profitable for everyone in the chain. Everyone. And I think the Anthropic S-one is kind of going to make that clear.”
Calacanis: US AI spend will hit $1T as companies adopt tokens
“There is no world in which I don't see corporations and people in the economy spending five or 10% of the salaries of their employees on the equivalent in tokens. In other words, somebody makes 80,000 dollars a year on average, they're gonna, you will spend 80…”
Sacks: Eliminating Amazon DSPs will add $5.20 per package
“What we do know it will do is raise the cost of packages by about five dollars and 20 cents per package, and it will cost families 664 dollars more per year, and it puts something like 5000 jobs at risk.”
Baker: xAI will release Grok 4.7 in a few weeks
“And this is still a relatively small model. It's one .1 trillion parameter models. A larger model, Grok 4.7, that should be significantly more capable, is coming out, you know, in a few short weeks.”
Sacks: Musk's compute deal includes a 90-day cancellation clause
“Both sides have a 90 day cancellation, which means that if the spot price of compute goes up, he benefits. If it goes down, by the way, it could hurt him because either side could renegotiate within 90 days, right? So basically Elon is able to charge spot -90.”