The frame
Dated Claims, and the Tests That Would Break Them
A forecast is only useful if it can be wrong.
The forecast runs on three horizons: the near term to 2050, the primary horizon to 2075 and the long arc to 2100. Each asserts specific, dated things, and each claim is paired with the observation that would break it.
Most of those observations are near term and already scheduled: Starship ship-to-ship propellant transfer and the Block 3 catch, the AI1 compute prototypes, Chang'e-7 and Chang'e-8 at the lunar south pole, Tianwen-3's Mars sample return and China's crewed landing on Long March 10.
| Page | Holds | Rated |
|---|---|---|
| Timeline | 66 milestones, filterable by destination, decade and confidence | Building |
| Experiments Board | 15 ranked experiments, 8 added since, 8 flight tests | Modeled |
| Falsifiers | 40 claims, what breaks each, and when we know | Modeled |
| 2050 - Near Term | Every destination at 2050 | Building |
| 2075 - Primary Horizon | Every destination at 2075 | Projected |
| 2100 - The Long Arc | Every destination at 2100 | Modeled |
| Interplanetary Markets | What trades between bodies, and why bulk does not | Modeled |
| Governance | Light time, the Sovereignty Ladder and who goes | Modeled |
| Discovery Economy | Science off Earth and who pays for it | Modeled |
The three horizons
2050, 2075, 2100
Each horizon asserts a different stage of the space-industrial transition.
| Horizon | What it asserts | Rated |
|---|---|---|
| 2050 - Near Term | Mature Earth-orbit industry with Starmind compute; Artemis and ILRS bases; first Mars landings; the first cislunar propellant economy | Building |
| 2075 - Primary Horizon | Belt industrialization; a Mars settlement past its evacuation line; lunar mass-driver export; the Callisto anchor | Projected |
| 2100 - The Long Arc | Flybys launched toward Alpha Centauri; the Mercury beam array; lens telescopes beyond 548 AU | Modeled |
The Experiments Board
The Tests That Move the Claims
The top of the ranked list, and the flight tests already scheduled.
| Experiment | What it settles | Rated |
|---|---|---|
| Maintenance-record study | Real failure rates and spares depth for a remote site; ranked first | Modeled |
| Partial-gravity ring | Whether Mars can be settled rather than only occupied | Modeled |
| Ship-to-ship propellant transfer · SpaceX | Whether anything beyond low Earth orbit is affordable | Building |
| Chang'e-7 ice survey · CNSA | Second direct measure of lunar polar ice grade | Building |
| Tianwen-3 · CNSA | Mars ascent and return, the hardware no one else has funded | Building |
Full board: Experiments Board.
Falsifiers
What Would Break the Thesis
The conditions that take the forecast down rather than delay it.
| If this holds | What it breaks | Rated |
|---|---|---|
| Propellant transfer never closes | The Moon and Mars stay resupply-fed outposts | Building |
| Partial gravity proves hostile to development | Settlement fails where outposts succeed | Modeled |
| Launch cost floors far above projection | The step reverts to a slow curve | Projected |
| Orbital compute proves uneconomic | The cadence case loses its anchor customer | Building |
All 40 claims: Falsifiers.
China
Chinese Programs on the Horizon
Dates held in public across the Moon, Mars and Jupiter.
| Program | Milestone | Rated |
|---|---|---|
| Chang'e-7 · CNSA | South-pole ice survey; launch postponed 24 August 2026 | Building |
| Chang'e-8 · CNSA | Resource use and construction, 2028 | Building |
| Mengzhou, Lanyue, Long March 10 · CMSA | Crewed lunar landing before 2030 | Building |
| Tianwen-3 · CNSA | Mars sample return, 2028 launch, samples around 2031 | Building |
| Tianwen-4 · CNSA | Callisto orbiter, launch around 2029-2030 | Projected |
| ILRS · CNSA and Roscosmos | Basic station and nuclear power station delivery, 2035 | Projected |
Contested
Deliberately Not Claimed
The load-bearing uncertainties, tagged with the confidence they actually earn.
Horizon years are anchors, not predictions. A slip in propellant transfer or launch cost moves the whole sequence right; a breakthrough moves it left. The order is firmer than the calendar.
The long arc is the least constrained. 2100 rests on compounding that has not started; it is the direction the near-term tests point.
The falsifiers are the content. If propellant transfer, partial-gravity physiology or the launch-cost floor goes the wrong way, the horizons get revised, not defended.
Record
Provenance and Data Sources
- Confidence
- Building for the 2050 near term - Projected for 2075 - Modeled for the 2100 long arc
- Status date
- Q4 2026
- China
- Chang'e-7 and Chang'e-8; Mengzhou, Lanyue and Long March 10; Tianwen-3 and Tianwen-4; ILRS
- Weakest link
- The 2100 long arc, and any falsifier that resolves against the forecast