Summary of Is OpenAI Ready To IPO?, The Datacenters in Space Myth, The Kids Boo AI

Overview of Big Technology Podcast — “Is OpenAI Ready To IPO?, The Datacenters in Space Myth, The Kids Boo AI”

Alex Kantrowitz and Ranjan Roy break down a huge AI-news week: OpenAI and Anthropic’s rumored IPO timing, what recent financial leaks reveal about their business models, whether SpaceX’s public-market story is really about space or AI, and why AI is getting increasingly booed at commencements. The episode’s core theme is that the AI industry may be entering a new phase: from hype and rapid growth to scrutiny, public backlash, and pressure to prove real economics.

OpenAI and Anthropic: IPO Timing, Revenue, and the Profitability Narrative

OpenAI’s numbers

• OpenAI reportedly generated about $5.7 billion in revenue in Q1, about $1 billion more than Anthropic.
• But its adjusted operating income margin was -122%, meaning it lost $1.22 for every $1 of revenue even after excluding some items like stock-based compensation.
• The hosts emphasize that this is not just “unprofitable in the usual startup way” — it shows the company’s core operating costs are still far above its revenue.

Why IPO now?

• The conversation centers on why these companies would go public before they are clearly profitable.
• Alex’s theory: OpenAI wants to go public before Anthropic does, so it can control the narrative.
  • Anthropic appears to be growing faster and may soon be able to tell Wall Street a better story.
  • If Anthropic IPOs first, OpenAI could be forced to respond to a narrative where its rival looks like the faster-growing, more disciplined company.

Anthropic’s numbers and strategy

• The Wall Street Journal reported that Anthropic is on track for:
  • $4.8 billion in Q1 sales
  • $10.9 billion in Q2 sales
  • an expected operating profit of $559 million in the quarter ending June
• Ranjan is skeptical that the growth trajectory is as clean as it looks and suggests Anthropic may be benefiting from short-term accounting or commercial arrangements, including a likely discounted deal tied to large infrastructure commitments.
• The hosts speculate that Anthropic may be trying to enter the public markets with a profitability story, because that story will play better with investors than a purely “YOLO growth” narrative.

Main takeaway

• Both companies appear to be moving toward IPOs not because they’re fully “ready” in the traditional sense, but because:
  • they need more capital,
  • public markets offer a much larger funding base,
  • and each wants to define the story before the other does.

Are AI Companies Ready to Go Public?

The new AI business model problem

• The hosts argue that AI companies don’t fit the classic software IPO model.
• In traditional SaaS:
  • you spend heavily on sales/marketing to grow,
  • then you can taper those costs and become highly profitable.
• In AI:
  • compute costs scale with usage, so revenue and expenses can move together.
  • That makes profitability harder to “flip on” later.

Their conclusion

• OpenAI is as ready as it will ever be if it wants to go public.
• Anthropic also appears ready, at least in narrative terms, because it can present growth and profitability together.
• The broader point: these IPOs may be less about maturity and more about capital access, narrative control, and market timing.

SpaceX, AI, and the “Data Centers in Space” Idea

What SpaceX’s filing says

• SpaceX filed to go public with numbers showing:
  • $4.69 billion in revenue
  • $4.28 billion in net loss in Q1
• The filing also frames SpaceX’s total addressable market as a staggering $28.5 trillion.

The AI pivot

• The hosts are struck by how much of SpaceX’s projected TAM is now tied to AI enterprise applications.
• Alex and Ranjan read this as a sign that SpaceX is no longer being presented as just a space company — it’s being positioned as part of the broader AI infrastructure story.

Skepticism about space-based data centers

• Both hosts are deeply skeptical that data centers in space will become a practical near-term business.
• Ranjan’s main objection:
  • If something breaks in a data center on Earth, you send a technician.
  • In space, you need an astronaut.
• They acknowledge Jeff Bezos’ comments and Blue Origin’s reported plans for large numbers of data center satellites, but neither is convinced that the economics or operational risks make sense yet.

Main takeaway

• The “data centers in space” idea may be technically imaginable, but the episode treats it as far more hype than reality for now.

AI Backlash: Why Students Are Booing It

Commencement season turns hostile

• Eric Schmidt was booed at a University of Arizona graduation.
• Other speakers mentioning AI at commencements also got booed.
• The hosts interpret this as a sign that AI’s public reputation is deteriorating.

Why the backlash is building

• The episode suggests the industry has failed to tell a compelling public story about:
  • how AI improves society,
  • how it benefits workers,
  • and what a positive AI future looks like.
• Instead, the dominant public framing has become:
  • job displacement,
  • surveillance,
  • and corporate overreach.

Notable reaction

• A quoted social-media observation in the episode boils the problem down to three possibilities:
  1. Tech leaders can’t tell a convincing story about AI’s benefits.
  2. They haven’t tried hard enough.
  3. They know the downsides and aren’t prioritizing fixing them.

Meta, Keystrokes, and the Messaging Problem

Zuckerberg’s leaked internal remarks

• The hosts discuss leaked audio of Mark Zuckerberg explaining Meta’s use of employee activity to train AI models.
• His point: Meta’s employees are unusually strong, so their work is valuable for teaching models how to code and solve tasks.

Why it lands badly

• While the logic is understandable, the messaging sounds cold and invasive:
  • “We’re monitoring your keystrokes to make our AI better.”
• Alex and Ranjan agree that the comms problem is severe:
  • even if the technical rationale makes sense,
  • the public and employees are unlikely to hear it as anything other than surveillance.

Main takeaway

• Silicon Valley is repeatedly failing to package AI in a way that feels socially beneficial rather than extractive.

Marc Andreessen, PR, and the Limits of Tech Evangelism

A bad example of making the case

• Andreessen is mentioned for saying bots don’t get frustrated, sick, depressed, or file HR complaints.
• The hosts think this kind of rhetoric is exactly why AI messaging is backfiring.

Broader point

• Tech leaders may believe AI is inevitable and therefore don’t need to persuade the public carefully.
• But in the U.S., unlike in China, public opinion still matters.
• That means careless messaging can turn into:
  • political backlash,
  • regulatory pressure,
  • and resistance to infrastructure buildouts.

Key Takeaways

• OpenAI and Anthropic may be heading toward IPOs sooner than expected, but the timing is driven by narrative control and financing needs, not traditional readiness.
• Anthropic’s growth and rumored profitability may be forcing OpenAI’s hand.
• AI economics are fundamentally different from classic software, making future profitability harder to predict.
• SpaceX’s AI story feels overstretched, especially around data centers in space.
• AI’s public image is worsening, with commencement boos and employee-surveillance concerns reflecting broader unease.
• The AI industry’s biggest challenge may now be communication and legitimacy, not just technology.

Final Thought

The episode’s underlying message is that the AI boom is moving from “Can these companies grow fast?” to “Can they justify the scale, the spending, and the social consequences?” The market may still be euphoric, but the public — and perhaps soon Wall Street — is starting to ask harder questions.