Back to Peter H. Diamandis

The $10B Satellite Empire Putting AI in Orbit, Why Chips Beat Rockets & China's #1 Open Model | #266

Peter H. DiamandisJune 26, 20262h 25m
Topics82
Large Earth Models and Planet's Mission0:00The Launch Tax vs. Compute Tax0:32GLM 5.2: China's Open-Weight Model1:00Guest Introduction1:31Planetary Intelligence: Two Phases4:31Real-World Data Requirements6:31Historical Archive Value7:31Fleet Specifications9:02Why No Competitor Matches the Fleet14:32Space and AI Convergence16:00Predictive Capabilities and Crystal Ball Vision18:30Technical Challenges for Global Prediction20:30Potential Applications23:00Planetary Stewardship Through Satellite Data24:53Planetary Infrastructure and Sovereignty25:31Transparency as a Deterrent26:01Reducing War Through Information27:01Democratizing Satellite Data Access28:01Revenue Breakdown28:31AI Training Data Strategy29:01US Government Oversight and Blacklists29:30International Blacklist Compliance30:31Resolution Limitations and Privacy31:30Space Law and Overflight Rights32:31Democratization of Intelligence Capabilities34:01Onboard Processing and Edge Computing34:30Real-Time Analysis Example35:00Time-Critical Applications36:02Data Volume Specifications37:01Satellite Economics vs. Drones38:30Resolution Improvements39:01Aircraft Detection39:31Starlink Comparison40:31Satellite Mass Comparison41:02Technology Evolution Since 201341:30Generational Improvements42:01Hyperspectral Development43:01AI as the Primary Unlock43:32Natural Language Interface44:01Commercial Use Cases45:30GDP Maximizing and Life Flourishing46:32Dyson Swarm and Orbital Compute47:30Compute in Space vs Terrestrial Data Centers50:41Technical Requirements for Orbital Compute53:30Orbital Visibility and Sun-Synchronous Orbits54:30Orbital Debris and Kessler Syndrome56:31Orbital Altitude and Satellite Lifespan59:00Starlink Orbital Altitude Discussion1:00:00Exponential Growth and Resource Concerns1:00:30Competition and Market Dynamics1:03:00Relativity Space Acquisition and ELIS Mission1:07:00Launch Cost Economics and Alternative Paradigms1:10:30Reusable Rockets and Launch Cost Economics1:16:01Compute Efficiency Over Launch Costs1:17:32Training Versus Inference Location1:21:00Space for Earth Versus Leaving Earth1:22:32AI Brain Drain from Google1:25:30Frontier Lab Competition and Talent Movement1:27:30Singularity Psychology and Anthropic Recruiting1:30:33Agency, Organizational Drag, and Physical AI1:35:32Planetary Intelligence and Space Data1:37:32The Role of Space Data in Advancing AI Beyond Internet Text1:40:51Multimodal Models vs. Real-World Embodiment1:42:31Javier Milei's AI Personhood Proposal for Argentina1:44:30Arguments For and Against AI Personhood1:46:32Resource Allocation for AI Safety and Governance1:48:00Call for Structured AI Governance Deliberation1:51:32Spectrum Approach to AI Legal Status1:52:01Milei's Proposal as Corporate AI Recognition1:54:30Sponsor Message: Blitzy1:58:31GLM 5.2: China's Leading Open-Weight Model1:59:31Chinese Models Challenging the 6-8 Month Lag Thesis2:00:31Distillation Practices Across AI Labs2:03:01Definition of Distillation2:05:30Distillation in Machine Learning2:05:53GLM 5.2 Performance and Chinese Model Efficiency2:07:30AI Alignment, Recursive Self-Improvement, and the Fermi Paradox2:09:01Additional Perspectives on the Fermi Paradox2:12:30Frontier Intelligence Cannot Be Monopolized2:13:32Orin and the OPTI Token Price Index2:15:01Hyperscaler Capex and Cash Flow2:18:01Closing Remarks and Outlook2:22:00
In a Nutshell

Planet operates a 200-satellite constellation generating 25 TB of daily Earth imagery to build "large earth models" that combine planetary sensing with LLMs for searchable, time-stamped physical-world data. The company is moving toward orbital AI compute clusters, arguing that once launch costs hit $200-300/kg, placing GPUs in space becomes cheaper than terrestrial data centers due to continuous solar power and no water/building costs. China's GLM 5.2 open-weight model now matches or exceeds Western frontier models on reasoning benchmarks, showing that distillation plus efficient token use is closing the gap and making frontier intelligence impossible to contain.

AI-Generated Notes

These notes were generated by AI and may contain inaccuracies.

Planet is a $10 billion public company (ticker: PL) with a 450% stock price increase over the past year. The company operates 200 satellites in Earth orbit, generating 25 terabytes of imagery daily. Will Marshall, co-founder and CEO, coined the term large earth models.

Large earth models work by indexing the Earth to make it searchable, similar to how Google indexed the internet. The goal is to enable humanity to become smart stewards of the planet by combining planetary sensing data with large language models.

"Bit like Google index the internet to make it searchable. We're indexing the earth to make it searchable. It will finally enable us to be smart stewards of our planet."

When discussing competition with SpaceX's plans for orbital AI data centers, Marshall noted that everyone except SpaceX pays the SpaceX launch tax, while everyone except Nvidia and Google pays the Nvidia tax. Near-term, launch costs are the dominant constraint, but longer-term, compute becomes the more important factor.

GLM 5.2 is a Chinese model that in some cases matches or exceeds top models from OpenAI and Anthropic. This level of performance in an open-weight model is considered shocking. The Chinese have figured out how to burn tokens to get more intelligence and how to reason more efficiently and cheaply.

Will Marshall is a physicist with a PhD from Oxford who previously worked at NASA. He co-founded Planet with partners after launching PhoneSat, a project that drew Steve Jurvetson's attention and led to major investment. Planet has contracts with every European government and billion-dollar-plus deals across the continent.

Sign in to read the full notes

Get access to AI-generated notes, topic timestamps, and more.