Scoring impact without pretending certainty.
The Civilisation Outlook Event-Impact Model records whether a documented event helps or harms civilisation, how far the impact reaches, how long it may last, how likely it is to spread and how strongly the evidence supports that assessment.
The seven-level Civilisation Scale
The underlying model retains its precise calculation for reproducibility and audit. The main public indicator translates that value into one of seven civilisation conditions so it is not mistaken for a percentage, probability, prediction of collapse or school-style mark.
| Public state | Internal score | Condition |
|---|---|---|
| State 1 | 0 or below | Critical |
| State 2 | Above 0 to 16.67 | Severe Strain |
| State 3 | 16.68 to 33.33 | Deteriorating |
| State 4 | 33.34 to 50.00 | Unstable Balance |
| State 5 | 50.01 to 66.67 | Fragile Progress |
| State 6 | 66.68 to 83.33 | Advancing |
| State 7 | 83.34 and above | Flourishing |
Six dimensions
| Dimension | Outlook weight | What it covers |
|---|---|---|
| Peace & stability | 20% | Conflict, safety, displacement and prevention of violence |
| Shared prosperity | 18% | Livelihoods, affordability, productive capacity and distribution |
| Health & human development | 17% | Health, care, education and human capability |
| Planetary resilience | 15% | Climate, ecosystems, pollution, resources and shocks |
| Responsible technological progress | 15% | Useful capability adjusted for access, safety and misuse |
| Governance & social cohesion | 15% | Rights, trust, participation, competence and cohesion |
Four measures of reach
Every event receives a 1–5 rating for geographic reach, population reach, persistence and propagation potential. This distinguishes a contained local incident from a smaller event capable of cascading across countries or systems.
Rating guide
| Rating | Geography | People | Persistence | Propagation |
|---|---|---|---|---|
| 1 | Local | Up to 10,000 | Days | Contained |
| 2 | Subnational | Up to 1 million | Weeks to months | Limited spillover |
| 3 | National | Up to 50 million | One to three years | Plausible expansion |
| 4 | Regional | Up to 500 million | Three to ten years | Actively spreading |
| 5 | Global | More than 500 million | Structural or 10+ years | Systemic or cascading |
Direction, intensity and evidence
An event can be positive in one dimension and negative in another. Each dimension receives a direction of −1, 0 or +1, an intensity from 0 to 5, a relevance weighting and a written rationale. Evidence confidence from 0 to 1 reduces scores whose consequences remain uncertain.
Related milestones carry an explicit independence factor. This prevents connected records—such as treaty adoption and later entry into force—from being counted as if they were wholly separate impacts.
Category parameters
| Category | Base weight | Half-life | Default emphasis |
|---|---|---|---|
| Conflict | 1.00 | 365 days | Stability and governance |
| Economy | 0.95 | 365 days | Prosperity and governance |
| Technology | 0.95 | 730 days | Responsible technological progress |
| Science | 0.85 | 730 days | Health, technology and planetary knowledge |
| Environment | 0.95 | 1,095 days | Planetary resilience |
| Disaster | 1.00 | 180 days | Health, stability and resilience |
| Anomaly | 0.60 | 180 days | Evidence, natural systems and public reasoning |
| Health | 0.95 | 730 days | Health and human development |
| Governance | 0.95 | 730 days | Governance, cohesion and stability |
| Society | 0.90 | 365 days | Cohesion, prosperity and human development |
Fourteen additional Dataset 002 records have complete v0.1 assessments held in review. Their model version and original inputs will remain preserved when later versions re-score the same evidence.
Source provenance
Every score retains a model-versioned snapshot of the source URL, title, publisher, publication date where known, retrieval time, evidence relationship and a note stating which claim the source supports, contradicts or contextualises. This preserves exactly which evidence v0.1 used even if a later model adds sources. Retrieval time records when Civilisation X consulted the source; it does not imply that a live page can never change.
Initial scored events
These are event-impact scores—not a score for the entire world. Every value is reproducible from the published inputs and can be challenged through the corrections process.
| Event | Reach | Net impact | Label |
|---|---|---|---|
| ChatGPT introduced to the public | 0.82 | +33.65 | Positive |
| WHO ends COVID-19 PHEIC designation | 0.80 | +21.32 | Positive |
| M 7.5 Noto Peninsula earthquake | 0.40 | −23.94 | Negative |
| Severe geomagnetic storm conditions | 0.60 | −6.67 | Limited or mixed |
| EU Artificial Intelligence Act enters into force | 0.84 | +29.57 | Positive |
| Study rejects reliable deep-sea fish precursor link | 0.54 | +16.94 | Positive |
World-score validation gate
The provisional score is published now with its caveats. It cannot be labelled validated or shown without those caveats until the model has at least 120 scored events, 20 events per dimension, four geographic regions, 24 months of observation, historical back-testing, sensitivity analysis and independent methodology review.
Challenge an assessment.
Provide a stronger source, corrected impact input or alternative rationale. Material corrections will remain traceable.