Machine consensus
Map common models, datasets, evaluation defaults and interface frames before counting outputs as independent.
W-AXIS OBSERVATORY
Machine consensus, weak signals and structural change.
When institutions depend on similar models, data and summaries, individually useful assistance may still produce correlated judgment failure.
RESEARCH INFRASTRUCTURE IN DEVELOPMENT · NO RESULTS CLAIMEDResearch problem
Did the evidence converge—or did the interpretive supply chain narrow?
The Observatory treats machine consensus as a testable institutional hypothesis. It studies dependency concentration, judgment compression, weak-signal retention, counterfactual breadth, authority and reversibility.
Weak signals
Preserve selected anomalies long enough to test them without treating rarity as truth.
Structural change
Look for moments when inherited rules stop mapping the system well enough for action.
Control layer
Record independent priors, provenance, dissent, authority, reversible action and accountable correction.
Working paper · v0.1
Primitive Intelligence Before Machine Consensus
A conceptual paper and research agenda. It has not been peer reviewed and reports no new empirical results.
Read status and abstractProposed instrument
Human First, Model Second
A preregistrable four-condition pilot comparing unaided judgment, default AI, forced counter-case and the W-Axis protocol. It has not yet been run.
Discuss a research collaborationInstitutional interface
Bounded, low-risk collaboration
Start with a private briefing, a synthetic-case working session or a scoped decision-method pilot. No affiliation or adoption is implied.
Request an institutional previewTruth ledger
CONCEPTUAL PAPERMEASURES PROPOSEDPILOT DESIGNEDEMPIRICALLY VALIDATEDPEER REVIEWEDINSTITUTIONALLY ADOPTED
No dataset has been collected, no experiment has been completed, and no causal or institutional-adoption claim is made.