Most current proposals reward higher self-sustainment with more local autonomy; if instead we treat high–self-sustainment, hard-to-intervene-off-world settlements as ethically more dangerous than dependent ones, under what specific, observable conditions (e.g., distance, communication latency, local weapons/AI capacity, difficulty of evacuation) should global law flip the usual logic and impose stricter caps on autonomy and population as self-sustainment rises, and does this imply that some Mars- or lunar-city visions should remain permanently off-limits even if they meet health and non-domination tests?

Answer

Flip the logic when self-sustainment combines with hard intervention and high-risk capabilities; then higher self-sustainment should tighten, not relax, caps. This likely rules out some Mars and lunar “city” visions indefinitely, even if they clear health and non-domination tests.

1. Conditions that trigger “inverse” autonomy logic

Treat autonomy/population caps as stricter with higher self-sustainment when all of these are true:

C1. Hard to intervene or evacuate

• One-way trip time to safe haven (Earth or large AG orbital) > ~6 months, or
• Round-trip evacuation for >50% of residents would take >5 years at plausible launch rates.

C2. High communication latency

• One-way light-time > ~5–10 minutes, so real-time oversight and crisis negotiation are weak (e.g., Mars, distant Mars orbit).

C3. High strategic capability density

• Local capacity for at least one of:
  • Frontier AI training or large secure compute clusters.
  • BSL-3+/equivalent bio labs or advanced gene synthesis.
  • Significant launch/kinetic infrastructure or mass drivers.
  • Large autonomous drone or weapons production.

C4. High self-sustainment in essentials

• Local production >70–80% of: life support, energy, critical spares, food, medical.
• External chokepoints can no longer constrain behavior without risking mass harm.

C5. Weak credible override

• No robust, multi-party mechanism for rapid external inspection and shutdown of high-risk systems that is:
  • Verifiable from Earth,
  • Technically effective,
  • And clearly separated from life-support control (to avoid domination).

When C1–C5 hold, more self-sustainment means more capacity to export risk or defect from global norms, while external leverage and inspection fall. In that regime, global law should:

• Cap population more tightly as self-sustainment rises.
• Cap or forbid certain high-risk capabilities locally.
• Slow or bar transitions from “outpost” to “city.”

2. Example policy rules by location

Orbit (LEO / cislunar)

• C1–C2 fail (fast access, low latency), so standard logic can apply: more self-sustainment can buy more autonomy within strong inspection regimes.
• Still: strict caps on high-risk AI/bio/weapon capacity regardless of self-sustainment (per bc29ed1f, 170da0ed).

Moon

• C1 partly true (days to evacuate, but feasible at scale with cost), C2 false (low latency).
• Flip-logic threshold: large, high–self-sustainment bases with:
  • Mass drivers or major launch assets, and/or
  • Frontier compute or high-end bio.
• For such lunar sites:
  • Keep population caps low (e.g., “town” scale only).
  • Treat any proposal for a large, self-sustaining city with heavy industry or compute as presumptively barred.

Mars

• C1–C2 strongly true for the foreseeable future.
• Once a Martian site reaches C3–C4 (plausible for any serious “city”), C5 is hard to guarantee.
• So for Mars under realistic politics:
  • Default: no “city” licenses (echoing 5931fb45, 3d6ddfee).
  • Allow only small “town”-scale, tightly chartered communities with:
    • Explicit bans or tight caps on frontier AI, high-end bio, and major launch/kinetic systems.
    • Strong managed-dependence levers that are non-coercive on life support (c0e58ac8).
  • Any vision of a large, near-autarkic Martian city with substantial strategic capacity should be treated as permanently off-limits.

3. Implications for Mars/lunar city visions

• Lunar megacities with heavy launch infrastructure and large compute/bio stacks: ethically suspect even if health and local non-domination look good; at most small, specialized hubs.
• Martian surface “cities” (hundreds of thousands to millions, highly self-sustaining): fail C1–C5; should be treated as beyond acceptable risk for the foreseeable future.
• The ethically safer path for large, more autonomous communities shifts toward:
  • Big artificial-gravity orbitals near Earth (3d6ddfee),
  • Small, function-bounded lunar/Martian outposts with population caps and strict industry restrictions.

So under this “danger of hard-to-intervene autonomy” lens, some Mars- and lunar-city futures should remain off-limits, not because of local health or local domination, but because they combine distance, self-sustainment, and strategic capabilities in ways that predictably erode global control of catastrophic risks.