Anthropic Aiming for $1 Trillion Valuation

Summary of Anthropic Aiming for $1 Trillion Valuation

by Lex Fridman Podcast Fan

14m•May 8, 2026
Listen to Episode

Overview of Anthropic Aiming for $1 Trillion Valuation

This episode is a fast-moving AI industry roundup centered on mega-rounds, layoffs, hardware shifts, and fresh legal disclosures. The biggest themes are Anthropic’s reported path toward a $1 trillion valuation, continued capital spending across the AI stack, Apple’s push to make AI hardware-native, and new lawsuit discovery that reveals how OpenAI’s board drama and Microsoft’s early investment actually unfolded.

Major AI Industry and Funding News

Anthropic’s possible $1 trillion raise

  • The Financial Times reports that Anthropic may try to raise money at a $1 trillion valuation.
  • Major demand is reportedly coming from:
    • Sovereign wealth funds in Asia and the Middle East
    • Google and Amazon
  • Goldman Sachs and JPMorgan are reportedly pitching a more conservative $400B–$500B IPO valuation for late 2026.
  • The episode highlights a major valuation gap:
    • Private secondary markets may be pricing shares much higher
    • This could reflect either frothy secondaries or banks intentionally underpricing for a strong first-day IPO pop

SoftBank trims OpenAI financing ambitions

  • SoftBank is reportedly cutting its OpenAI margin loan target by 40%, from $10B to $6B.
  • This is part of a broader sense that AI financing is still extremely aggressive, but not all capital plans are scaling as originally expected.

Broader AI industry tightening

  • Cloudflare is cutting more than 1,100 jobs in what it calls an “AI-native reset.”
  • The episode compares this to similar restructuring narratives from Coinbase and eBay.
  • Snapchat reportedly canceled a $400M Perplexity deal after weaker-than-expected ad sales.
  • Snapchat’s Q1 ad revenue came in at $1.53B, below expectations.

SpaceX, xAI, and Massive Compute Spend

TerraFab and Texas chip manufacturing

  • SpaceX is reportedly committing $55B to a Texas AI chip plant called TerraFab, with potential expansion up to $119B if additional phases are built.
  • The filing came through a public hearing notice in Grimes County, Texas, tied to a tax-break request.

“Earth compute” and “space compute”

  • The discussion ties this to Elon Musk’s earlier comments about:
    • 200 gigawatts of “Earth compute”
    • Up to 1 terawatt of “space compute”
  • The scale is framed as extraordinary even for a highly cash-generative private company like SpaceX.

xAI infrastructure strategy

  • SpaceX/xAI’s Memphis compute footprint is also discussed.
  • The first Colossus data center reportedly got leased out to Anthropic, while Colossus 2 is being used for xAI’s own needs.
  • The takeaway: this is a very expensive but increasingly strategic race to control compute.

Apple’s AI Hardware Push

Siri may become model-agnostic

  • Apple is reportedly preparing to let users choose which model powers Siri, including:
    • Gemini
    • Claude
    • ChatGPT
  • The host praises this as a practical approach because model quality changes quickly and users may want flexibility.

Camera-equipped AirPods

  • Apple is working on AirPods with built-in cameras.
  • These are expected to:
    • Provide low-resolution visual context
    • Let Siri “see” what the user sees
    • Support AI-assisted questions and awareness of surroundings
  • The episode sees this as a much stronger AI hardware play than standalone novelty gadgets.

Hardware AI: what works and what doesn’t

  • The host argues that AI devices succeed when they build on products people already wear or use:
    • Smart glasses: promising
    • AirPods: promising
    • AI pendants: unlikely to catch on
  • The Meta Ray-Bans are used as proof that camera-enabled wearable AI can feel natural and useful.

Apple’s AI credibility problem

  • Apple is facing criticism for shipping AI promises before the features were ready.
  • The episode references a legal settlement / payout tied to buyers of certain iPhone 15 and 16 models who were promised Apple Intelligence features that did not arrive on time.
  • The tone is that Apple has strong hardware instincts, but still needs to prove it can deliver on AI software.

OpenAI, Microsoft, and the Elon Musk Lawsuit Revelations

New discovery dumps reveal internal drama

  • The Elon Musk vs. Sam Altman lawsuit has produced a trove of internal emails and texts.
  • The episode says the discovery is revealing because it shows the interpersonal and governance chaos behind the scenes.

78 texts between Sam Altman and Mira Murati

  • One major revelation is a chain of 78 text messages between Sam Altman and Mira Murati during Altman’s 2023 ouster.
  • The messages show Altman trying to:
    • Get reinstated
    • Understand who was on which side
    • Bring in Satya Nadella to help stabilize the situation
  • The board had serious governance concerns, and the episode suggests these were not just isolated disagreements.

Ilya Sutskever’s memo

  • Ilya Sutskever allegedly wrote a 52-page memo to the board outlining why Altman should be removed.
  • The memo appears to have been central to the board’s decision-making.

Microsoft’s early OpenAI investment: more pragmatic than mythic

  • Discovery also suggests Microsoft’s first $1B investment in OpenAI was driven less by pure AGI conviction and more by business strategy.
  • Microsoft reportedly worried OpenAI might “storm off to Amazon/AWS” if it didn’t get enough support.
  • Internal documents show Microsoft was also:
    • Skeptical that OpenAI would ever achieve AGI
    • Focused on retaining a valuable customer and avoiding bad PR
  • The episode’s conclusion: Microsoft’s investment turned out to be historic, but the original motivation may have been more defensive and commercial than visionary.

Main Takeaways

  • AI funding is entering a new phase where $1T valuations are being seriously discussed.
  • The AI arms race is increasingly about compute, chips, and infrastructure, not just models.
  • Apple’s best AI strategy may be to become the hardware gateway to multiple models, rather than betting on one in-house model.
  • The OpenAI/Microsoft lawsuit discovery is reshaping the public story of how the company’s early funding and board crisis actually happened.
  • A lot of the current AI boom is still being driven by strategic fear, customer retention, and infrastructure control, not just technical optimism.