Building large-scale AI data centers in space versus mining Bitcoin in space reveals a fundamental economic divergence. For space-based AI data centers, scaling potential is virtually unlimited. As long as global AI demand continues to surge, Earth’s power grid, land availability, cooling capacity, and regulatory constraints will eventually become binding bottlenecks. Even if space-based computing remains far more expensive upfront, terrestrial costs could rise under persistent undersupply until they match or even exceed that of space-based levels. Bitcoin mining in space, besides the latency concerns previously noted by @peterktodd [https://petertodd.org/2024/keeping-it-cool-mining-bitcoin-in-space], is a zero-sum game constrained by a hard protocol cap. Miners share a fixed daily pie of roughly 450 BTC plus variable fees. At current prices around $60,000–$70,000, total annual revenue tops at ~$10 billion. Revenue does not scale linearly with added capacity in this market. Unless the price of each bitcoin rises to ~$500 billion, it would be economically impossible to operate a Bitcoin mining farm large enough to block the entire planet of Venus. But an AI data center could.(Chun)