---
title: Settling Space
---
Settling space is less a destination than a redundancy strategy: the slow, expensive project of giving a single-planet species a second place to keep its books. For most of history this was pure fiction. What changed is not the dream but the cost curve underneath it — the price of putting a kilogram into orbit has fallen far enough that the question is no longer *whether* but *how*, and the "how" splits the field into two rival architectures that have been arguing past each other for fifty years.
## Why leave at all
Two motives do most of the work, and they are different in kind. The first is **backup** — the actuarial case. Everything humanity values currently sits inside one fragile gravity well, exposed to the same asteroid, the same engineered pathogen, the same bad century. A second self-sustaining home does not make those risks vanish, but it removes the single point of failure. The [Great Filter](great-filter) framing sharpens the stakes: if civilizations routinely die in the narrow window between gaining planet-altering power and spreading beyond their home world, then getting off-planet is not ambition but survival.
The second motive is **growth** — the case from abundance. The energy and material budget of one planet is a ceiling; the solar system's is, for any practical horizon, not. The blog essay this cluster grows from imagines the *center of gravity* of an advanced civilization gradually leaving Earth across the late 21st century — not as an exodus, but because the largest compute, manufacturing, and construction simply find more room out there. Backup is a reason to leave even if life on Earth stays good. Growth is a reason to leave *because* it could get so much better.
## The central debate: habitats or planets
The cluster's spine is one argument with two camps, and almost everything else hangs off it.
On one side: **built orbital habitats**. The lineage runs from Gerard O'Neill through Jeff Bezos and Blue Origin. Don't land anywhere — *manufacture* your world. A spinning cylinder in open space gives you Earth-normal gravity by rotation, sunlight on tap, a sky on the inside, and a design you can tune rather than inherit. The [O'Neill colonies](oneill-colonies) page is the deep dive: habitats sized for tens to hundreds of thousands, assembled in orbit from asteroid and lunar material.
On the other side: **settled planets**, principally Mars, championed by Elon Musk and SpaceX. Here the bet is that a real planet — with its own gravity, raw materials, and room to grow into — is the more durable foundation, even if you start out living in shielded burrows under the regolith. The [Mars](terraforming-mars) page draws the line that most arguments blur: near-term *colonization* (hard but buildable this century) is a different proposition from *terraforming* (which current science does not support). Crucially, the two camps are not mutually exclusive. The essay's far future contains both — O'Neill cylinders in Earth and Mars orbit, lava-tube cities on the Moon and Mars — because by the time either is mature, civilization has the surplus to do both.
## The shared enablers
Whichever architecture wins, the same handful of technologies has to arrive first, which is why progress in any one of them moves the whole cluster:
- **Cheap, high-cadence launch.** Reusable heavy lift — Starship and its competitors — collapses the cost of moving mass off Earth. Nothing downstream is affordable until this is routine.
- **Power.** Compact fission, eventual fusion (lunar helium-3 is one proposed feedstock), and [space-based solar power](space-based-solar-power) — collecting sunlight in orbit where it never sets — together supply the energy abundance the rest of the buildout assumes.
- **In-situ resource utilization (ISRU).** Living off the local rock instead of shipping everything up: asteroid mining, lunar and Martian regolith for shielding and feedstock, and eventually molecular-scale assembly that turns mined material directly into structure. ISRU is what flips space from a place you visit to a place you build.
- **Closed-loop life support.** Sealed worlds — whether cylinders or buried cities — only work if air, water, and food cycle almost perfectly. This is the unglamorous engineering on which every other ambition quietly rests.
A consequence worth naming: once the largest **orbital compute** moves off-planet, it becomes infrastructure worth fighting over. That is the thread the [orbital warfare](orbital-warfare) page picks up — the strategic shadow side of an economy that has left the ground.
## The counterpoint: take care of the world we have
The honest objection is not technical but moral, and it deserves a fair hearing. Pouring trillions into uninhabitable rock and vacuum, the argument runs, is an escape fantasy that lets us defer the harder work of keeping Earth livable — the only place we are actually adapted to. Even the most ambitious timelines keep most of humanity, and nearly all of the unaugmented, here on Earth. A backup that only the wealthy can board is not insurance for the species; it is a lifeboat for a few.
The strongest version of the pro-settlement reply concedes the point and reframes it: the same closed-loop, energy-dense, recycle-everything technologies that make a sealed habitat survivable are exactly the ones a strained Earth needs most. On this telling, settling space and stewarding the home planet are not rivals competing for one budget but two faces of learning to live within hard limits. Whether that synthesis holds, or whether it is a comforting story we tell to justify the rocket, is the question this whole cluster keeps circling back to.
## Related concepts
## Appears in
[The Shape of an Ordinary Day: Life from the Age of Agents to the Far Side of the Singularity](https://mystrangemind.com/p/the-shape-of-an-ordinary-day)