All-In with Chamath, Jason, Sacks & Friedberg

Anthropic’s $30B Ramp, Mythos Doomsday, OpenClaw Ankled, Iran War Ceasefire, Israel's Influence

All-In with Chamath, Jason, Sacks & Friedberg·April 17, 2026

INVESTMENT OPPORTUNITIES

  • Anthropic: Brad Gerstner is a long-term investor in Anthropic, having first invested at a $130-150 billion valuation. The company is experiencing an unprecedented revenue ramp, topping a $30 billion run rate by April 2025, which was their initial year-end target. Their product, Mythos, is considered to have near-AGI capabilities, driving significant enterprise demand. It is suggested Anthropic could exit 2025 at $80-100 billion in revenue.
  • OpenAI: Brad Gerstner is also an investor in OpenAI. This company is also on a rapid growth trajectory with highly capable models, though experiencing internal FUD (Fear, Uncertainty, Doubt) and leadership changes. They are expected to release powerful new models soon and are also projected to see massive revenue growth.
  • PolyMarket Bets: The hosts discuss PolyMarket bets on both Anthropic and OpenAI going public (IPO) before 2027. For Anthropic, there's a 51% chance, and for OpenAI, a 44% chance.

OVERVIEW

This episode delves into the rapid and unprecedented rise of AI frontier models, particularly Anthropic's Mythos, and the resulting industry and geopolitical implications. The hosts debate whether the hype around AI's dangers is genuine or a calculated marketing tactic, while also discussing the economic and political forces shaping AI development and deployment.

KEY TOPICS

  • Anthropic's Mythos model and its withholding due to safety concerns
  • The discovery of thousands of vulnerabilities by Mythos in major operating systems and browsers
  • The concept of AGI (Artificial General Intelligence) models and the ethical dilemmas of their release
  • Anthropic's strategic "Project Glasswing" with major tech companies to address vulnerabilities
  • The role of government regulation versus industry self-regulation in AI development
  • The "boy who cried wolf" argument regarding AI safety warnings
  • The unprecedented revenue ramp of Anthropic and its implications for the tech industry
  • The future of coding and the potential for AI-driven automation
  • The impact of AI on existing tech debt and enterprise software development
  • The ongoing competition between frontier AI companies (OpenAI, Anthropic, Google, Meta)
  • Geopolitical implications of AI, including the capabilities of China
  • The influence of US foreign policy, particularly concerning the Israel-Palestine conflict

MAIN TAKEAWAYS

  • AI frontier models are reaching a critical threshold of capability, posing significant challenges and opportunities.
  • While some view AI safety warnings as marketing tactics, the industry is taking proactive steps to address potential vulnerabilities through collaborative initiatives.
  • The economic landscape of AI is rapidly transforming, with companies like Anthropic demonstrating unprecedented revenue growth, driven by enterprise demand for powerful AI solutions.
  • The debate between industry-led development and government regulation in AI is ongoing, with arguments for both approaches in managing risk and fostering innovation.
  • AI's immediate impact on legacy tech debt in large enterprises is questioned, with some arguing that current models are not yet capable of solving complex, deeply embedded software issues.
  • The auto-translate feature on platforms like X (Twitter) is seen as a powerful tool for global communication and truth discovery, potentially transforming international discourse.
  • Geopolitical events, particularly the Israel-Palestine conflict, highlight the complexities of foreign policy and the role of US influence, with diverse opinions on its effectiveness and ethical implications.

NOTABLE QUOTES

"If it prevents your browser history from being released, everybody in the world Chamath, that may be something that you're willing to, you know, let 100 days pass on."
"What do you honestly think is going to get accomplished in 100 days?"
"The ability for any of these models to actually build enterprise-grade software is still shit. S.H.I.T. shit."
"Base Japanese is a whole 'nother level of based."

Summarized with DriftNote — AI-powered podcast summaries

Try it free