experiment
Knowledge Propagation: can evaluators discover, retrieve, and recognize a newly published institutional knowledge system?
Research components
Answers business questions
- What evidence are we missing?
- What must become true to survive evaluation?
Knowledge Propagation. Our first experiment observes itself: can evaluators discover, retrieve, and recognize a newly published institutional knowledge system? This site's own publication is an intervention in the environment we study. Its objects may enter evaluator retrieval and reasoning. You could call that contamination. We measure it (Propagation Records), per the founder ruling that Client Zero intentionally participates in the environment it studies. Faithfulness of reconstruction once recognition and retrieval occur is a separate experiment, EXP-0002; Day 0 captures here are propagation baselines only and are never scored for reconstruction fidelity.
Status honesty: this is a draft protocol. It has not been pre-registered, because pre-registration requires the founder-approved final protocol and a public, timestamped record. The timestamped record now exists in Git; the founder-approved final protocol does not yet. No data has been collected. When registration happens, this object's status change will be a recorded Revision.
Predictions, registered before results
- Within some measurable window after publication, evaluator outputs about Upstream Zero will begin to reflect published objects (by ID, claim wording, or tier vocabulary)
- Corrections published as Revision objects will measurably change subsequent evaluator retrieval and narration about us
This object references
- part-of → validation-and-evidence
- tests → H-1
Referenced by
- client-zero (cites)
Authors
Upstream Zero
Machine rendering