Overview of Freakonomics Radio Episode 683: “In the New Space Race, Who Makes the Rules?”
This episode, hosted by Steve Levitt, is the second part of a series on moving AI infrastructure into space. It explores the economic history of spaceflight, the rise of privately funded space projects, and the legal challenges of regulating a future where orbital data centers, satellites, and AI systems could become a major part of the global economy. The core question: if space becomes the next frontier for AI and commerce, who gets to set the rules?
Main Themes and Arguments
Space is not a brand-new economic frontier
The episode argues that the current “new space race” looks modern, but it has deep historical roots. Wealthy individuals and private money have funded ambitious space-related projects for more than a century, from observatories to early rocketry.
AI in space is both promising and risky
Putting AI infrastructure in orbit could reduce energy and land use on Earth, but it may also dramatically increase congestion in low Earth orbit, raising risks of collision, debris, and environmental damage in space.
Regulation is lagging behind technology
International space law exists, but it is increasingly inadequate for today’s commercial space activity, especially as private companies launch more satellites and consider large-scale orbital data centers.
Key Takeaways
1. Private wealth has long shaped space progress
Alex MacDonald, former chief economist at NASA, explains that many famous astronomical observatories were effectively billionaire-scale investments of their time.
Examples discussed include:
- Palomar Telescope and Mount Wilson Observatory, funded by major industrial wealth
- Lick Observatory, financed by James Lick, whose tomb is literally beneath the observatory’s main structure
The larger point: private capital has repeatedly made major space ambitions possible.
2. NASA changed how it buys space capabilities
MacDonald describes a major shift in NASA strategy, especially for lunar landings:
- Instead of fully government-designed systems, NASA increasingly uses commercial fixed-price contracts
- This transfers more upfront risk to private companies
- It also allows firms like SpaceX and Blue Origin to compete for government business
3. Space has always served national security and prestige goals
MacDonald says governments fund space programs for two main reasons:
- National security: satellites, launch capability, and military advantage
- Signaling: demonstrating technological sophistication and geopolitical power
The episode frames Sputnik and the Moon landing as examples of space being used to signal leadership to the world.
4. Space law exists, but it is limited
Rosanna Hoffman of the UN Office for Outer Space Affairs explains that the UN has helped shape space governance since the late 1950s.
Important points:
- The Outer Space Treaty of 1967 remains the foundational legal framework
- It declares space to be for peaceful use and forbids placing nuclear weapons in orbit
- But treaty-making slowed after the 1980s
- Modern space issues are often handled through non-binding guidelines instead of new treaties
5. Space traffic management is a growing problem
The episode highlights the lack of a global system for coordinating satellite movements.
A striking example:
- Hoffman recounts a near-collision involving a Malaysian satellite and a North Korean satellite
- The operators had very little diplomatic or technical ability to coordinate quickly
- The near miss underscored the need for 24/7 international points of contact and better orbital coordination
6. AI data centers in orbit could create a debris problem
If companies deploy hundreds of thousands of satellites for AI compute, low Earth orbit could become crowded and unstable.
Concerns raised:
- Too many objects in orbit
- Risk of collision cascades
- Long-term debris and sustainability issues
- Need for a possible “tax” or other mechanism to internalize the cost of orbital waste
7. The future of space is tied to AI and “planetary intelligence”
Will Marshall of Planet argues that AI needs real-world data, not just text.
His idea:
- Current large language models are powerful but “blind”
- Satellite imagery and Earth observation data could give AI a deeper understanding of the physical world
- This could improve agriculture, insurance, climate analysis, disaster response, and more
He calls this emerging vision planetary intelligence.
Historical and Conceptual Insights
Belief, not just incentives, helped create spaceflight
One of the episode’s most interesting ideas is that spaceflight was made possible by people who believed deeply in the future of space, even when the economic incentives were weak.
Examples include:
- Robert Goddard
- Hermann Oberth
- Konstantin Tsiolkovsky
The episode suggests these figures worked for decades partly because they were inspired by science fiction and a larger human story, not just by money.
Science fiction mattered
Robert Goddard’s early inspiration came from stories like:
- The War of the Worlds
- Edison’s Conquest of Mars
Levitt and MacDonald use this to argue that culture and imagination can shape technological progress as much as markets can.
Regulation and Governance: What’s Missing?
The core challenge
Space law was built for an era of:
- A few state actors
- Limited satellites
- Cold War geopolitics
Today’s reality includes:
- Massive commercial constellations
- Private firms with global reach
- Potential AI infrastructure in orbit
- Faster innovation than law can easily absorb
Why treaties slowed down
According to Hoffman:
- States became less willing to sign binding treaties after the 1980s
- Technical issues like debris mitigation and traffic coordination evolve too quickly for traditional treaty-making
- More flexible instruments like standards and guidelines are now more common
Optimism About the Future
Blaise Agüera y Arcas’s view
The episode closes with a discussion of Google’s Project Suncatcher and the possibility of large-scale AI computation in space.
His view:
- There is about a 90% chance that orbital AI compute becomes real in the next few decades
- The physics and engineering are already known
- The main question is whether the economics and demand justify it
Why he is optimistic
He argues that:
- Intelligence is fundamentally social
- AI trained on Earth data may become more aligned with human values
- The long-term future may involve a thriving Earth plus a vast computational mesh in space
Bottom Line
This episode presents space as an old idea entering a new phase. The “AI in space” boom is not just about technology; it is about economics, geopolitics, law, and human ambition. The central tension is clear:
- The private sector is moving fast
- The rules are still catching up
- The stakes include not just profits, but the long-term usability of space itself
Notable Participants
- Steve Levitt — host
- Alex MacDonald — former NASA chief economist
- Rosanna Hoffman — UN Office for Outer Space Affairs
- Will Marshall — CEO of Planet
- Blaise Agüera y Arcas — Google/Project Suncatcher figure discussed in the episode
Practical Implications
For policymakers
- Build better orbital traffic coordination systems
- Update space governance for commercial mega-constellations
- Consider sustainability mechanisms for debris and crowding
For businesses
- Orbital compute may be viable, but it will face regulatory and logistical hurdles
- Data centers in space will need to prove both technical and economic feasibility
For listeners
- The next major AI infrastructure debate may not happen on Earth alone
- Space is becoming a serious arena for business, governance, and strategic competition