DARK STAR POTENTIAL
Potential asks what a breakthrough could unlock
A Dark Star is a specific civilizational problem that remains unresolved. Its T0–T9 potential tier asks how large an increment a real solution could unlock under stated success conditions. It is not a person score, a realized outcome, or a forecast.
Grand goals belong to Nebulae; tractable bottlenecks belong to Dark Stars. If progress occurs, each person is still assessed separately on the contribution that actually happened. No one inherits a Dark Star’s score or tier.
Magnitude M
M concerns the outcome, not fame or an individual share. Era, cultural origin, reputation, and the amount of surviving documentation are not multipliers. The explanation must say what changed, who or what it reached, how long it lasted, what existed before, what later work added, and where the claim stops. M6 can mark a major capability or framework in one field; M7 requires a foundational framework still usable across generations and many settings; M8 concerns structural shifts in what civilization can do.
Attribution a
Collaboration is allocated once through the shared budget. It is neither a second deduction from M nor a reason to grant one person an entire team result. Sources may support factual actions without determining a precise share; unresolved responsibility stays visible rather than being invented.
Deduplication
One shared outcome generates one Q. Repeated publication, a renamed source work, predecessor input, or a later delivery cannot create another full outcome without an isolable new function. Six value buckets help recall and inspect possible value; they do not multiply the same result six times.
Rank display
Current rankings use unrounded ΣL. Displayed decimals express arithmetic precision, not evidence precise to that place. A current point estimate is contestable and retains stated gaps and sensitivity; a hoped-for total cannot be used to work backward into M or a.
CURRENT COVERAGE
210 people, 1106 assessed outcomes
687 outcomes still have missing evidence. Their current estimates remain visible with sources, reasons, and open questions; they are not presented as final. New evidence revises the same outcome under the same rule.
1 · Survival and health
Reduced death and suffering; health span; disease mechanisms, diagnostics, treatment, prevention, public health, nutrition, and scalable basic health engineering.
2 · Knowledge and truth
Verifiable understanding, explanatory and learning systems: theories, methods, measurement, standards, knowledge organization, and correction of systematic error.
3 · Capability and productivity
Scalable capability and tools: general technology platforms, engineering systems, infrastructure, diffusion, cost reduction, and usable access to productive capacity.
4 · Cooperation and governance
Durable large-scale cooperation: law, public services, accountable institutions, collective action, rights, conflict reduction, and institutional correction.
5 · Sustainability and civilizational safety
Long-term survival under increasing power: externalities, resilience, clean energy with net benefit, and constraints on nuclear, biological, AI, and related catastrophic risks.
6 · Culture, value, and meaning
Ethical frameworks, art, literature, education, faith, aesthetic structure, and narratives that shape how power is used and how meaning is made. Their intrinsic value need not be converted into a technology product to count.
Ideas, education, and art may create knowledge, capability, ethical, aesthetic, or meaning value directly. Scientific and intellectual outcomes face the same test: neither scientific prestige nor an unsupported demand for global net benefit decides M. The six categories are a checklist, not six separate copies of the score. Unsupported claims, pure marketing, rent extraction without a civilizational gain, and safety plans that shift risk onto weaker people or the future do not count.
From nomination to publication
A person or outcome is nominated with sources, then assessed item by item. A second assessor works without seeing the first score and checks identity, facts, translations, what each outcome includes, duplicate counting, and personal credit. If they disagree on an important fact or cause, a red–blue review may be needed before the assessment is updated.
Current scores can still change
A score may be public while sources, objections, and missing evidence remain visible. A fully verified assessment requires completed fact checking, an independent second review, any needed challenge process, and a final decision. Current person scores do not use T-tiers.
Versions remain separate
The current itemized method is the only source of public person scores. People not yet assessed under it retain ordinary introductions without Canonical v1 scores, ranks, or tiers. Canonical v1 and the frozen early v2.2/R1 route remain historical records and are not blended with current scores.
Contribution and harm
The score records independent positive contribution. Direct harm or a major limitation that cannot be separated from an outcome affects its M assessment. Other questions about conduct or harm are disclosed separately instead of being converted into one exchange rate. A serious unresolved dispute can lead to a score cap, a warning, or exclusion.
PUBLIC DOCUMENTS
Full rules, sources, and change history
The library keeps the complete public texts on scoring, evidence, revisions, and civilization-wide research. Each item says whether it is a current rule, a historical document, or an editorial summary.
CIVILIZATION-WIDE RESEARCH
How far has human civilization progressed?
Starlight is also researching how to assess civilization-wide progress. The current method and data do not support a reliable overall grade, so no civilization level is published. The earlier 0.70–0.73 trial readings are no longer used; the standards, data sources, and independent validation must be revisited.
This research does not affect person scores, Dark Star potential, or Nebula values.