OpenAI and SpaceX are Racing to IPO First

Summary of OpenAI and SpaceX are Racing to IPO First

by Lex Fridman Podcast Fan

18m•May 20, 2026
Listen to Episode

Overview of OpenAI and SpaceX are Racing to IPO First

This episode is a fast-moving AI news roundup centered on the accelerating race among major AI companies to go public, while also highlighting how AI is reshaping jobs, search, music generation, and desktop automation. The host argues that AI is forcing companies to reorganize now, even if that ultimately leads to larger teams and more growth later.

IPO Watch: OpenAI vs. Musk-Led Companies

  • OpenAI is reportedly targeting a September IPO, with some reporting it could file as soon as next week.
  • The company is said to have crossed $3.8 billion in annualized revenue, mostly from API usage and ChatGPT subscriptions.
  • The host frames this as a race with Elon Musk’s broader ecosystem, especially SpaceX and related efforts, and cites a LinkedIn comment suggesting Musk’s lawsuit against OpenAI may have been intended more to delay OpenAI’s conversion than to win in court.
  • The key idea: timing matters, because a major IPO window can shape investor attention and market flow.

Layoffs, Reorganizations, and the AI Workforce Shift

Intuit and Meta as examples of “AI restructuring”

  • Intuit is cutting about 17% of its global workforce (roughly 3,000 employees) and closing offices in Reno and Woodland Hills.
  • The CEO, Sasan Goodarzi (name corrected), said the move is part of a strategic reset to simplify operations and improve products.
  • The host notes the irony that Intuit recently reported strong Q3 earnings and 10% year-over-year growth, suggesting these cuts are more about AI-driven restructuring than financial distress.

The host’s broader thesis

  • Companies are using AI to automate tasks that used to require manual labor, especially repetitive administrative work.
  • Rather than assuming layoffs mean permanent shrinkage, the host predicts many companies could become larger over the next 2–3 years, after the current realignment.

Meta’s approach

  • Meta is reportedly laying off 8,000 employees but reallocating 7,000 workers into AI-related roles.
  • The host views this as a better model than pure layoffs: shift people into AI-native work instead of simply cutting them.
  • He argues most employees can learn to use AI effectively in their jobs.

Google’s AI Expansion: Search, Ads, and Usage Growth

  • Google AI Mode has reportedly reached 1 billion monthly users.
  • Usage has reportedly doubled every quarter since launch.
  • The host describes Google’s changes as the biggest overhaul to search in 25 years.
  • Key changes include:
    • a redesigned search experience,
    • more AI Overviews,
    • a chat-like AI interface,
    • and agentic features that may generate custom apps or dashboards instead of simple web links.

AI-generated ads inside search

  • Google is embedding Gemini-generated product descriptions directly into search ads.
  • Sponsored items are being integrated into AI Mode responses, further blending search, answers, and advertising.
  • The host argues this is a logical move for Google:
    • AI responses need to be monetized,
    • and users who want ad-free experiences can pay for premium AI subscriptions.

Stability AI’s Stable Audio 3.0

  • Stability AI launched Stable Audio 3.0, with a major improvement: song generation up to 6 minutes and 20 seconds.
  • The model is trained on licensed music data from Warner Music Group and Universal Music Group, avoiding the licensing battles facing some other AI music startups.
  • Model lineup:
    • 459M, 1.4B, and 2.7B parameter versions.
  • Distribution model:
    • smaller versions are available with open weights for commercial use if a company makes under $1 million in annual revenue,
    • larger use cases require paid API access or enterprise licensing.
  • The host likes this model because it gives smaller creators and startups access while charging larger businesses.

Iris Go: AI Desktop Automation Agent

  • Iris Go raised $2.8 million in seed funding to build a proactive AI desktop automation agent.
  • The product’s core pitch:
    • watch a user do a task once,
    • learn the workflow,
    • then automate it going forward without manual prompting.
  • Notable features:
    • on-device processing for privacy,
    • end-to-end encryption for cloud tasks,
    • beta apps for macOS and Windows,
    • and a pre-install deal with Acer for new devices.
  • The host connects this directly to his own experience using AI agents for bookkeeping and app development, saying this kind of tool could dramatically reduce repetitive work.

Main Takeaways

  • AI is moving from novelty to infrastructure: search, ads, desktop automation, bookkeeping, and music creation are all being transformed.
  • Big tech is reorganizing around AI:
    • some roles are being cut,
    • others are being redirected into AI-focused work.
  • Distribution and licensing matter as much as model quality:
    • Stability AI’s licensing strategy and Iris Go’s Acer deal are highlighted as smart business moves.
  • Monetization is unavoidable:
    • Google and other platforms are embedding ads directly into AI experiences to preserve revenue.

Closing Notes

  • The episode ends with a promotion for the host’s newsletter/site, where listeners can get the stories in written form.
  • There’s also a request for Apple Podcast reviews, with the host trying to catch up to Spotify in review count.