[ PROOFECONOMY ]™GlossaryOriginSpecsProofJack Rating
A ProofJack Rating is the consolidated human versus AI share of a term, specification or work: the average of the human share of its idea and the human share of its name, shown as Human X% and AI Y%, with the record it was taken from.
What it means
Every term in this glossary was made by a human, an AI, or both. The ProofJack Rating says how much of each, in one number. It covers two things only: who had the idea and who chose the name. The wording of the definition is not part of the rating, because almost all of it is drafted with AI and saying so once is enough.
How it is computed
Rating = average of the conception share and the naming share.
| Case | Conception | Naming | Rating |
|---|---|---|---|
| Human had the idea and the name | 100 / 0 | 100 / 0 | Human 100% · AI 0% |
| Human had the idea, AI proposed a name from the human’s own phrase | 100 / 0 | 50 / 50 | Human 75% · AI 25% |
| Human had the idea, AI proposed the name | 100 / 0 | 0 / 100 | Human 50% · AI 50% |
| Human set the requirement, AI proposed the mechanism and the name | 50 / 50 | 0 / 100 | Human 25% · AI 75% |
Each rating names its record: the authorship disclosure of the specification that defines the term, or a dated session record.
Until proven otherwise
Where a specification’s disclosure has no record for a single term, the term takes that specification’s own published shares. Where a term was originated outside any AI-assisted specification and nothing records AI help, it is rated on the author’s attribution, Human 100%, pending measurement. Industry terms and terms whose names others also use are not rated. A rating is never raised without a record, and any record that shows otherwise replaces it.
Self-estimated, then measured
Today’s ratings are self-estimated from the disclosures. ProofJack (PP-SPEC-088) records human and AI actions as they happen, in a hash-chained ledger, and will replace them with measured values as the sessions behind each term are recorded.
What it is not
- Not a quality score. It says how a term was made, not whether it is right.
- Not authorship. The author of record is the human who directed, chose and accepted the work. An AI cannot be an author or an inventor.
- Not AI detection. It is taken from records of who did what, never guessed from the text.