Three indexes. One separate scale.

Cooperative Index · 48.4

A weighted average of nine supplied Cooperative domain scores. Its editorial weights now assign Medicine 20, Energy and AI 15 each, Space, Robotics, Humanity and Peace 10 each, and Money & Capital and Physics 5 each. A score of 100 represents the Cooperative endpoint.

Galactic Index · 24.9

A weighted average of ten Galactic domain scores, including societal forecasting. A score of 100 represents the speculative galaxy-scale endpoint.

Civilization Index · 33.8

For each paired domain, compute √(Cooperative × Galactic), then round half-up to a whole score. The weight is the mean of the two index weights, also rounded half-up. The overall figure is the weighted average of these combined scores. Societal forecasting keeps its Galactic score and weight.

These scores use the supplied editorial milestones and reference values. They are not probabilities, dates, or a literal percentage of all technology completed.

The scale measures the common level across nine domains plus their progress toward the next level. Levels run from 0 (Pre-industrial), 1 (Planetary-industrial), 2 (Planetary maturity), 3 (Interplanetary), 4 (Interstellar), to 5 (Cooperative).

The current shared level is 1. A weighted harmonic mean combines progress, which emphasizes lagging tracks: F = 1 + 0.389 ≈ 1.39.

Weights are 1 plus the number of outgoing enabling edges, plus a terminal-value bonus of 4 for Medicine & Longevity, Connected Humanity and Peace & Governance. The bonus values those outcomes directly, even when they enable few other tracks. Peace weighs 13, AI 8, Money & Capital 8, and Energy 7. The weighted arithmetic result is 1.48, so the bottleneck penalty is 0.09.

Critical path and gating

The engine tests a 0.2 increase in each track’s progress, capped at 1. Peace & Governance produces the largest increase: +0.102. This is a sensitivity test. Gate notices identify prerequisite tracks that are behind the required level.

Every displayed raw indicator has a source name, source URL and as-of date. Inputs missing those details are omitted. A lower-is-better measure is shown as reduction toward zero.

“Feed-linked” means an input is mapped to a public endpoint. This release uses the supplied 22 Sep 2026 snapshot and has a scheduled daily source review. Verified updates are published after validation; source publication frequencies vary. No historical series is drawn when none is supplied.

Action paths and horizon ranges are editorial scenarios, not promises or reliable forecasts. The What-if control changes only hypothetical next-level progress; it does not update the source data or the three 0–100 indexes. Milestone dates are estimates or context from the handoff. Forecast inputs, including Metaculus medians, retain their labels.