S&P 500’s Influence on AI Companies’ IPOs

Summary of S&P 500’s Influence on AI Companies’ IPOs

by Lex Fridman Podcast Fan

12m•June 5, 2026
Listen to Episode

Overview of S&P 500’s Influence on AI Companies’ IPOs

This episode focuses on how index rules, massive AI infrastructure spending, and investor behavior are shaping the next wave of AI-related IPOs. The main throughline is that the S&P 500 is sticking to its inclusion rules, which would block fast-track entry for SpaceX and likely for future IPOs from OpenAI and Anthropic as well. The episode also covers explosive compute demand, enterprise AI budget overruns, Airbnb CEO Brian Chesky’s new AI lab, and the unusual amount of capital overlap between OpenAI and Anthropic investors.

S&P 500 Rules and Why They Matter

  • The S&P 500 reportedly رفض/denied SpaceX’s request to be fast-tracked into the index.
  • The host explains the key eligibility hurdles:
    • A 12-month seasoning period after IPO
    • A 10% float requirement
    • A profitability screen
  • The practical impact is huge: inclusion would have triggered roughly $14 billion in passive buying for SpaceX.
  • The episode argues this also means OpenAI and Anthropic would be unlikely to get immediate S&P 500 inclusion after their IPOs.
  • Other indexes may move faster:
    • Nasdaq 100: about 15 trading days
    • FTSE Russell’s Russell Top 500: about 5 days
  • Main takeaway: the S&P 500 is choosing index discipline over capital flow convenience.

Massive AI Compute Deals and Infrastructure Spending

  • The episode says Google is paying SpaceX $920 million per month for access to 110,000 NVIDIA GPUs and related compute capacity through June 2029.
  • The host frames this as part of a broader trend: AI companies are scrambling for compute, and infrastructure owners can monetize overbuild by leasing capacity to competitors.
  • Another major deal mentioned:
    • Anthropic reportedly signed a $1.25 billion per month agreement for the full Colossus One Memphis facility.
  • Alphabet is said to be facing $180–190 billion in 2026 capex, and it raised $85 billion via stock sale to help fund that spending.
  • Broader point: AI infrastructure is becoming a capital arms race, with demand still outpacing supply.

Enterprise AI Token Spending Is Out of Control

  • Enterprises have reportedly blown past their 2026 AI token budgets by 3x.
  • Examples given:
    • Uber allegedly burned through its annual coding budget in just four months.
    • Another company faced a $500 million Claude bill after failing to set usage limits.
  • The episode highlights how quickly usage can balloon, especially with newer agentic workflows and long-running tasks.
  • Supporting data cited:
    • Per-developer token use rose 18.6x in nine months
    • Top token users were 2x more productive but used 10x more tokens
    • A Cursor renewal at Priceline cost 4–5x more than expected
  • The host recommends:
    • Setting strict usage controls
    • Avoiding open-ended API access where possible
    • Using subsidized or managed plans if available
  • The Linux Foundation launched the Tokenomics Foundation to standardize AI cost metrics, with 180 vendors already participating.

Brian Chesky’s New AI Lab

  • Airbnb CEO Brian Chesky is starting his own AI lab while remaining CEO of Airbnb.
  • The lab will focus on user interaction and design, areas Chesky believes current AI tools handle poorly.
  • He reportedly thinks existing AI systems lack the rich interfaces needed for travel and e-commerce.
  • Airbnb has also reportedly rejected ChatGPT plugin partnerships because the underlying tools weren’t robust enough.
  • The episode frames this as a move toward vertical, design-focused AI products rather than generic chatbots.

Investor Overlap Between OpenAI and Anthropic

  • About 90 investors reportedly hold stakes in both OpenAI and Anthropic.
  • That includes major firms such as:
    • Sequoia
    • Greylock
    • Founders Fund
    • Redpoint Ventures
  • The host notes this is uncommon in Silicon Valley, where investors usually pick a side in direct competitor situations.
  • Key stats cited:
    • 42% of OpenAI backers also hold Anthropic stakes
    • At least 13 of 31 names in Anthropic’s latest raise already own OpenAI shares
  • The overlap is presented as a form of hedging ahead of IPOs and trillion-dollar valuations.
  • Around 30% of the overlapping investors are said to be hedge funds, private equity firms, or wealth managers, which are more accustomed to diversified bets across competitors.

Key Takeaways

  • Index eligibility rules matter: the S&P 500’s refusal to bend for SpaceX could shape capital flows for future AI IPOs.
  • Compute is the new strategic asset: companies are racing to secure GPU capacity and monetize any excess.
  • AI usage is expensive fast: enterprise token consumption is already becoming a serious budget-management problem.
  • UI/UX may be the next AI battleground: Chesky’s lab suggests that design and workflow quality could differentiate future AI products.
  • Investor behavior is shifting: backing both sides of major AI rivals is becoming more common as valuations and capital requirements explode.