Framework paper
VCF: A Canonical Framework for Classifying Realized AI Value
March 2026 · Live PDF · Core framework
VCF (Value Classification Framework) defines the value episode as the atomic unit of AI value measurement and represents each episode through Outcome Primitive, Outcome Intent, Outcome Magnitude, and dual-axis evidence semantics. The framework is designed to make realized outcomes comparable across systems and time while staying honest about uncertainty.
Publication snapshot
Atomic unit
Value episode
Canonical components
OP + OI + OM + evidence
Public projection
3 visible axes
Key claims
Benchmarks measure potential and analytics measure interaction, but neither classifies the realized outcome itself.
The minimum public reporting projection is `OP × OM_goal × claim_strength_tier`, preserving outcome type, structural scale, and evidence strength.
The framework is explicitly source-grounded: outcome theory, validity, quasi-experimental design, and causal language constraints each map to concrete VCF design decisions.
Why this paper exists
The Missing Measurement
VCF begins from a simple gap: modern AI evaluation can report benchmark performance and product activity, but it still struggles to classify what changed for the human. The framework shifts the object of measurement from model response to outcome trajectory.
Unit of analysis
Value Episodes
The canonical unit is a value episode rather than a message, thread, or account. Transport session, attempt, and episode remain distinct so mixed-intent work does not collapse into one noisy label.
Canonical representation
OP, OI, OM, and Evidence Semantics
VCF preserves a coarse comparable top layer, context-specific intent beneath it, structural magnitude in Human Effort Equivalent hours, and a dual evidence model that separates claim strength from source provenance.
Intellectual lineage
Source Lineage as Design Constraint
The paper makes explicit how Donabedian, Campbell and Stanley, Messick, Chollet, and causal-inference literature shaped concrete framework decisions. The references are used as architecture constraints, not decorative signaling.
Why it matters
An Outcome-Grounded Lens on Progress
VCF suggests an operational framing of general intelligence based on the ability to produce high-magnitude outcomes across the full landscape of human work under strong evidence, rather than on test scores alone.