Generated Record Corpus Share (GRCS)

GMI-AGR-033Agriculture and Farm OperationsPilot-ReadyPattern O-40: Generated-Record Recursion

Definition

The share of records in the farm's and packhouse's operational corpus (field events, boundaries, as-applied reconciliations, scouting notes, traceability lot links, sanitation records) that were generated or inferred by an AI system rather than observed or entered, and the share of those that carry no marker distinguishing them from observed records.

The Measurement Gap

There is no metric that tells you what share of the records in the farm's agronomic corpus were generated or inferred by an AI system rather than observed, which means the next model, the insurance claim, and the carbon credit are trained and adjudicated on the system's own output.

Formula

GRCS = N_generated / N_records, by record type; companion: N_generated_unmarked / N_generated

Proposed formula, pending validation pilots.

At a Glance

Unit
Ratio (0 to 1)
Trend Direction
Diagnostic; disclosure share, the companion unmarked share is lower is better
Category
Data Provenance and Record Integrity

Decision Supported

Whether the next model, the insurance claim, and the carbon credit are being trained and adjudicated on observations or on the system's own output, and whether the corpus needs a provenance field.

← Elicitation-Attributable Error Share (EAES-AGR)Platform Dependency Exposure (PDE) →

Assess Your Organization

The AI-GPS Assessment evaluates your AI governance maturity and recommends KPIs like this one for your implementation roadmap.

Take the Free AI-GPS Assessment