The US Government's AI Intervention and the Future of Lunar Construction
This Week in Startupsgo watch the original →
the gist
Jason Calacanis and Lon Harris discuss the US government's emergency shutdown of Anthropic's Fable 5 and Mythos 5 models, while interviewing GRU Space founder Skyler Chan about his startup's mission to build a lunar hotel using robotic, regolith-based manufacturing.
The AI Regulatory Shock
Jason Calacanis and Lon Harris analyze the recent, abrupt US government intervention that forced Anthropic to pull its Fable 5 and Mythos 5 models. The discussion centers on the fragility of relying on a single, centralized AI provider. Calacanis argues that this event serves as a wake-up call for developers and enterprises, emphasizing that "no single model dependency is safe." He advocates for "multi-model harnesses"—architectures that allow organizations to swap models dynamically to mitigate the risk of sudden service outages or regulatory shutdowns.
The Lunar Construction Frontier
Skyler Chan, founder of Y Combinator-backed GRU Space, explains his vision for the "next SpaceX": building the first robotic hotel on the Moon. Rather than shipping heavy construction materials from Earth, which is cost-prohibitive, GRU Space is developing a payload factory designed to excavate lunar regolith (moon soil) and bind it into structural bricks using a geopolymer approach. Chan frames this as a "civilizational-scale" necessity, drawing parallels to historical expansion where settlers utilized local resources rather than relying on imports from their origin point.
Business Models in Space
Chan outlines a business model that transitions from being a construction contractor for NASA to eventually owning lunar land and infrastructure. He discusses the concept of selling "lunar condos" via refundable deposits, positioning the project as a wedge for broader lunar industrialization. The conversation touches on the inevitable "land-grab" and the regulatory uncertainty surrounding extraterrestrial property rights, comparing the current state of space development to the historical expansion of the Hudson's Bay Company.
The Risks of Centralization
Beyond the technical challenges of space construction, the episode highlights the tension between private innovation and government oversight. The Anthropic shutdown is treated as a case study in how political and regulatory pressures can instantly invalidate a company's core product stack. Calacanis suggests that the future of AI development will be defined by resilience and the ability to operate across a diverse, decentralized ecosystem of models rather than betting the farm on a single provider.