Orbital compute farms are coming. Musk’s TERAFAB vision is an important early step, and LEO may be a practical place to begin. But if we think even slightly beyond the immediate engineering horizon, it becomes clear that LEO cannot be the long‑term architecture for civilization‑scale compute.What’s missing from the current conversation is the question: “What comes after LEO?”Because the answer determines whether we build something scalable or something we’ll have to replace later at enormous cost.1. LEO is useful as a starting point — but fundamentally limited. LEO offers fast iteration and easy access, but it also brings inherent constraints: – orbital congestion – debris accumulation – interference with astronomy – long‑term shadowing concerns if scaled – limited energy densityThese aren’t hypothetical future problems. They’re structural properties of LEO.2. Solar‑proximal orbits (Venus/Mercury) are the physically optimal destination. Solar flux increases dramatically as you move inward: – Venus orbit: ~2× Earth – Mercury orbit: ~11× EarthEven for early deployments, this means: – more energy per square meter – fewer satellites for the same output – lower mass per watt – simpler thermal management – better long‑term economicsYou don’t need to wait decades for these advantages to matter. They’re true from day one.3. The Moon is a better long‑term platform than LEO for maintenance and scaling. If we expect lunar manufacturing and launch infrastructure to grow, then the Moon is also a natural place to host compute: – stable surface – easy robotic maintenance – no Kessler risk – nuclear reactors allowed – no interference with Earth’s sky – no cumulative shadowing on Earth’s climateThe Moon is a scalable compute platform in a way LEO simply isn’t.4. LEO as a trampoline is fine — but only if we already know the direction of the jump. The real risk is not starting in LEO. The risk is staying in LEO by inertia, simply because it’s familiar.Humanity tends to fix architectures only after they break. But in this case, waiting until LEO is saturated will make the transition vastly more expensive.We should be discussing the post‑LEO architecture now, not when the bottlenecks appear.5. A civilization aiming at Kardashev I cannot rely on Earth‑adjacent infrastructure. A Type I trajectory naturally pushes us toward: – lunar compute arrays – Venus/Mercury solar compute rings – heliocentric data infrastructure – Dyson‑compatible architecturesThese are not sci‑fi endpoints. They’re the physically optimal places to put large‑scale compute and energy harvesting.Conclusion: LEO may be a practical starting point, but it is not the right long‑term architecture. If we want orbital compute that is scalable, safe, and economically efficient, we need to start planning beyond LEO now — toward solar‑proximal orbits and lunar platforms where energy is abundant and orbital congestion is irrelevant.The conversation about “what comes after LEO” is overdue. It’s time to start it.