Key Moments
Anthropic S-1 Shocker, Greatest Business Ever? Google’s $500B Wipeout, Midterm Predictions
All-In PodcastKey Moments
Anthropic's S-1 filing reveals a company valued at nearly $2T, poised to be the largest IPO ever, but its success hinges on unprecedented AI compute demand and navigating regulatory hurdles.
Key Insights
Anthropic's S-1 filing projects annual recurring revenue (ARR) of $100 billion by year-end and a total addressable market (TAM) of $40 trillion.
The company reports $16.3 billion in revenue for H1 2026, a 14x year-over-year increase, with Q2 revenue reaching $11.6 billion.
Anthropic's S-1 filing shows a significant increase in high-spending customers, with nearly 11,900 clients spending $100k or more as of September.
The company has raised approximately $110 billion and has a unique ownership structure with seven founders holding 50.1% of voting power through a single-class F share.
Google's market cap wiped out approximately $500 billion due to concerns that AI advancements could disrupt its core search advertising business.
The midterm election predictions suggest a divided Congress, with Republicans likely retaining the Senate and Democrats securing a narrow House majority, leading to potential political gridlock and investigations.
Anthropic's astronomical IPO and business potential
The All-In Podcast discusses the groundbreaking S-1 filing of Anthropic, projecting it to be the largest IPO in history, targeting a $100 billion raise and a valuation between $1.5 to $2 trillion. Goldman Sachs and Morgan Stanley are set to lead the offering, aiming for a November 16th launch. The company's revenue figures are staggering: $16.3 billion in the first half of 2026, a 14x year-over-year surge, with Q2 alone hitting $11.6 billion. This explosive growth is underpinned by an ambitious TAM of $40 trillion for their AI services. Projections for annual recurring revenue (ARR) are set at $100 billion by year-end. Anthropic also boasts $112 billion in cash, demonstrating exceptional capital efficiency, having spent only about $9-10 billion of the roughly $110 billion raised to date. This financial picture challenges the narrative of massive losses, revealing a company with substantial liquidity and impressive margins, reportedly around 85% gross margin, which are improving due to escalating model efficiency. The company's unique structure includes seven founders holding 50.1% of voting power via a single-class F share.
AI's capital efficiency and improving margins
A key takeaway from Anthropic's S-1 is their remarkable capital efficiency and improving profitability. Despite raising nearly $110 billion, they've spent only about $9-10 billion, leaving them with over $100 billion in cash. This contrasts sharply with the public perception of high-burning AI startups. Their gross margins are reported to be around 85% and are on an upward trend. This improvement is driven by a tenfold increase in cost-efficiency for specific tasks year-over-year, allowing Anthropic to reduce model costs while capturing greater efficiency for their own profit margins. This strategy is termed 'compute multipliers,' achieved through algorithmic improvements, better data curation, and advanced architectures like Mixture-of-Experts (MoE). This suggests that the AI commoditization narrative may be premature, as Anthropic demonstrates a path to profitability and significant competitive advantage.
The AI value chain and massive customer spending
The discussion highlights the burgeoning value chain within AI, illustrating how significant profits are generated at each stage. NVIDIA, for instance, achieves a 4x return on its cost of goods sold for chips, with a 75% gross margin. Companies like SpaceX, using these chips for their AI infrastructure, achieve an 83% gross margin. Anthropic itself reports an 85% gross margin on its AI services. The analysis suggests that every dollar spent on NVIDIA chips ultimately generates $18 in revenue for Anthropic. Crucially, customers are also seeing substantial returns, getting two to ten times the value of their spending. The S-1 reveals a dramatic rise in high-spending customers: from 200 clients spending over $100k at the end of 2024 to 11,900 by September 2025. This indicates a massive, largely untapped market, with Anthropic having penetrated only about 1% of its potential customer base.
Regulatory risks and infrastructure constraints
Anthropic's S-1 filing clearly identifies computing infrastructure as its primary risk factor, with specific concerns around data centers, energy, chips, and the power grid. This is followed closely by regulatory risks, including government policies, sanctions, and export controls. The company's focus on these constraints is seen as a strategic move, particularly in the context of political maneuvering for the 2028 elections. Democrats might aim to slow down AI development by restricting energy and data center expansion to impact the economy before 2028, while Republicans would prioritize accelerating these areas. The intense demand for compute power also creates a competitive dynamic, with AI companies potentially competing with the US Treasury for capital, a scenario unprecedented in market history.
Google's market cap plunge and search advertising challenges
Google experienced a significant market cap drop of around $500 billion following its Q3 earnings report. The primary concern is the disruption of its core search advertising business by AI advancements. While Google's search revenue is still growing, the rate of ad click growth is stagnating, raising questions about the long-term viability of its pay-per-click model as AI provides direct answers. This is compounded by the fact that Google Cloud is performing exceptionally well, growing 25% year-over-year and showing strong operating income. However, the core search business, which funds all other ventures, is showing signs of slowing growth, creating uncertainty about Google's future revenue streams and its ability to transition effectively into the AI-driven era. The company is also grappling with internal strategic direction, with some theorizing that the focus on lucrative cloud services might be diverting resources from AI model development.
Midterm election predictions and political uncertainty
The panel discusses their predictions for the upcoming midterm elections, with a general consensus on the unreliability of polls due to potential bias and outright fabrication. David Sacks points out a pattern of polls consistently overestimating support for Democrats. The prevailing sentiment is that while the exact outcomes are uncertain, a divided Congress is likely, with Republicans potentially holding the Senate and Democrats a slim majority in the House. The discussion touches upon the potential for increased political polarization and investigations, regardless of the outcome. The economic narrative is also debated, with differing views on whether it's a disaster or showing signs of genuine growth. This political uncertainty adds another layer of complexity to the market outlook.
The growing conflict between AI and government control
A significant theme is the escalating tension between the power of AI companies and government control. The immense capital and influence being amassed by AI entities like Anthropic are seen as a potential threat to governmental authority. This could lead to governments seeking to regulate or even co-opt AI capabilities to maintain their own power. The debate centers on whether governments will try to control compute resources, legislate AI development, or compete for capital. The panelists suggest that this power struggle will intensify, potentially leading to increased government intervention in the AI sector to ensure it aligns with national interests and political agendas. The risk of AI development being slowed down for political reasons is also a key concern.
The open-source threat and the race for AI supremacy
The competitive landscape of AI is examined, with a particular focus on the threat posed by open-source models. While Anthropic currently demonstrates strong performance and pricing flexibility, the proliferation of powerful, lower-cost open-source alternatives is seen as a long-term risk, especially in areas like coding. To maintain their edge, companies like Anthropic are increasingly reliant on self-improving AI (RSI) – a concept that also raises concerns among 'doomers' about AI safety and control. The panelists question the sustainability of this race for AI supremacy, as pushing the boundaries of RSI is inherently risky. There's a potential for a rift between companies like Anthropic, who are pushing the limits of AI, and 'doomers' who fear unchecked AI advancement, potentially leading to a situation where open-source models eventually catch up.
Mentioned in This Episode
●Products
●Software & Apps
●Companies
●Organizations
●Concepts
●People Referenced
Common Questions
Anthropic has filed its S-1, the document required for an IPO. This filing reveals significant financial data and business projections, suggesting it could be the greatest business ever founded, aiming to raise up to $100 billion with a valuation near $2 trillion.
Topics
Mentioned in this video
A major focus of the discussion due to its S-1 filing for an IPO. The company is presented as potentially the greatest business ever founded.
Its revenue growth and stable employee count are presented as an example of AI-driven productivity gains.
Its chatbot is mentioned as a direct competitor to Google Search, and its future S-1 filing is anticipated.
Used as a benchmark for large IPOs, with Anthropic's potential IPO expected to surpass it.
Mentioned as one of the lead underwriters for Anthropic's potential IPO.
Its stock performance and business segments (Search, Cloud, YouTube, Android) are analyzed in light of AI disruption.
Mentioned as a company whose ad system copied Google's early model and as a competitor in the AI market for capital.
A platform that provides data and predictions on various events, including Anthropic's IPO and US elections.
Discussed for its role in manufacturing AI chips and its profit margins.
Mentioned as a partner who issued convertible bonds to Anthropic.
Mentioned as a subsidiary of Alphabet that is expected to raise external capital.
Mentioned as one of the lead underwriters for Anthropic's potential IPO.
Used as a benchmark for market share and revenue, with Anthropic's potential revenue being compared to it.
Purchased ChatGPT for all its employees, signaling a trend in AI adoption.
Mentioned as a growing revenue stream for Alphabet.
Mentioned as a potential oligarch who might influence Democratic politics.
His study on recent election polls is cited, suggesting a consistent bias towards Democrats.
Mentioned in the context of midterm election losses during his presidency.
A pollster whose predictions for Trump's performance in Iowa were deemed inaccurate.
Mentioned in the context of Google's cloud business and retaining talent.
Mentioned as a proponent of banning AI development.
Mentioned as having left Google, possibly due to resource allocation issues related to AI models versus cloud computing.
Scheduled to appear on an 'All-In' live election stream, noted for a positive transformation.
Mentioned as a proponent of banning AI development.
Mentioned as a figure associated with the 'America First' movement.
Mentioned as a potential Republican candidate who might be sidelined by a new populist leader.
Authored a widely circulated post detailing alleged changes in Google's ad auction model.
Shared a video suggesting a Republican defeat in the midterms.
A professional football player. Mentioned in a humorous, tangential comment about avoiding 'fast breaks'.
Mentioned as a point of comparison for wealth creation and a political figure.
Mentioned as a potential Republican candidate who might be sidelined by a new populist leader.
Mistakenly transcribed as 'Jeez', but later referred to when discussing political predictions and potential election outcomes.
Mentioned as a potential Democratic presidential candidate in a hypothetical matchup with Trump.
Mentioned as stating that the Democratic party is dominated by the Democratic Socialist wing.
Mentioned in relation to potential Republican candidates.
Scheduled to appear on an 'All-In' live election stream.
His retention at Google is framed as a critical task for the company.
Referred to as the 'favorite organic farmer', he participates in discussions on business, AI, and politics.
Mentioned as a potential new oligarch who might influence Democratic politics.
Mentioned as a proponent of banning AI development.
Praised for Microsoft's debt reduction while increasing revenue.
Scheduled to appear on an 'All-In' live election stream.
Mentioned in the context of midterm election losses during his presidency.
Discusses his performance at 8090, the AI market, and political predictions.
Co-founder of Anthropic, his framework for computing is discussed.
Associated with the 'Contract with America' during the Republican takeover of the House in 1994.
Mentioned as a major cloud computing provider, comparable to Google Cloud.
Noted for its significant growth and profitability, contributing to Alphabet's diversification.
Anthropic's AI model, mentioned as providing significant value and enabling faster business growth.
An AI chatbot that has not yet been released.
A productivity app mentioned in the context of workplace AI adoption.
Mentioned as a growing revenue stream for Alphabet.
A communication platform mentioned in the context of workplace AI adoption.
Mentioned as an AI agent that has not yet been released.
The theory that increased efficiency leads to increased consumption, applied to AI token usage.
Refers to the collective concentration of AI companies and their significant capital needs, impacting financial markets.
A key driver for AI advancement, particularly for Anthropic, but also considered risky by 'doomers'.
More from All-In Podcast
View all 470 summaries
56 minWHOOP CEO, Will Ahmed: 50B Hours of Health Data, a 4 Million Member Data Moat, & the FDA Fight
104 minAI Cyber Panic, Ken Griffin’s Bubble Warning, The Fall of France?
40 minArvind Krishna & Nikesh Arora: Quantum Threat, Open vs Frontier AI, AI Cyber Warfare & SaaS Collapse
41 minDylan Field: Is AI Replacing Designers?, Design is the New Code & Democratizing Creativity
Ask anything from this episode.
Save it, chat with it, and connect it to Claude or ChatGPT. Get cited answers from the actual content — and build your own knowledge base of every podcast and video you care about.
Get Started Free