Methodology v2.1
From noisy coverage to a relationship signal.
AIEO keeps four questions separate: what was published, what actually happened, whether it was already known, and how the source evidence positions people and the AI side.
The evidence chain
Four steps, one direction
i
Technical prompts, thresholds, and private orchestration code are not published. Public concepts, counts, source links, release scope, revisions, and limitations are published.
Primary relationship lens
Two directions create four complete patterns
i
The four cells use only complete two-sided signals. One-sided, no-clear, ambiguous, and insufficient cases remain visible but are excluded from the four-cell percentages.
The AI side refers to the system, operator, or ecosystem as represented in the source. It does not imply consciousness.
People ↑AI ↑
Both gain
MutualismPeople ↑AI ↓
People gain, AI side is constrained
Human-benefiting parasitismPeople ↓AI ↑
AI side gains, people are constrained
AI-benefiting parasitismPeople ↓AI ↓
Both are constrained
Competition or co-constraintHow coverage and developments are counted
Each AI-related source page dated in the weekly period counts once in the coverage total. Pages about the same concrete occurrence are grouped into one development.
Additional reports = coverage items minus distinct developments. They measure extra attention, not extra reality.
How new and seen-before are decided
A development is new to AIEO when it is not matched to an accepted event in the disclosed historical pool. A later source about an earlier event increases attention but does not create a new development.
Long-gap candidates are compared with the full registry. Human-governed decisions can revise earlier releases, and prior revisions remain archived.
How the secondary human empowerment index works
Expanding effects receive a positive score, contracting effects receive a negative score, and mixed or no-direct-change cases receive zero. Unclear cases are excluded.
The public index runs from -100 to +100. It is accompanied by a plain-language label and the scored versus unclear denominator.
What AIEO does not claim
The five discovery markets are a research sample, not a complete census of global AI news. A search market is not necessarily the location of the reported development.
The relationship lens classifies how source evidence represents the relation. It does not establish objective system performance, private experience, consciousness, intention, or biological fitness.
Read the output