Concept Explainer · Operant Dyad™
Operant Dyad, explained
A concise reference for the concept, aimed at readers who want the definitions and the current evidence status without reading the full paper.
What is an Operant Dyad?
The configured, task-bound working relationship between one human and one functionally bounded adaptive AI system. It forms through repeated correction: the human supplies goals, constraints, and defect reports; retained corrections shape later outputs; the human inspects, accepts, rejects, or revises consequential results.
Why is persistence important?
If nothing were retained between exchanges, each interaction would stand alone and there would be no relationship to examine, only a series of independent outputs. Persistence is what turns a sequence of exchanges into a relationship with a history — and a history is what raises the paper's central concern: whether the person supplying that history still authors what the relationship produces.
What is Relationship State?
The retained record of goals, constraints, defect reports, and corrections that shapes later interaction. The paper distinguishes declared state (what the person explicitly supplied) from inferred state (what the system derived on its own without being told).
How is this different from personalization or User State?
This is the least settled question, and the paper does not claim to have settled it. The wider Operant Dyad research program explicitly has not established that a distinct, irreducible dyadic state exists, separate from things that already have names — a person's own learning and adaptation, an AI system's own persistent state, explicit memory, or shared artifacts (which overlaps with what "personalization" or "user state" usually describes). Current work (Detect → Localize → Identify → Perturb → Probe → Reduce → Map) is testing whether the relationship-level effect Operant Dyad describes is something more than those familiar carriers, or whether it reduces to them. That question is open, not answered.
What is actually claimed?
- That a task-bound working relationship, formed through repeated correction, is a coherent object worth studying on its own terms.
- That such a relationship can be described using Relationship State, Truxion (recurring misalignment cost), and the Portable Operant Packet (a proposed continuity representation).
- That the relationship can fail in at least three named, non-exhaustive ways: Lock-In, Override, and Hidden Scaffolding.
- That preserved human authorship is the governing condition for a healthy Operant Dyad — a normative standard the framework proposes, not a feature it presumes current systems already have.
What remains unproven?
- Whether a distinct, irreducible dyadic state exists at all, as opposed to the effect reducing to simpler, already-named carriers.
- Whether Truxion is measurable in practice.
- Whether a Portable Operant Packet can actually carry continuity across systems.
- Whether Lock-In, Override, and Hidden Scaffolding are exhaustive, or the most consequential failure modes.
Evidence status: Conceptual proposal and research agenda. Not an empirically validated system, a claim of exhaustive novelty, legal advice, or a claim about machine consciousness or personhood.
Where can readers go deeper?
The canonical paper record · Research brief · Research map · The wider Operant Dyad™ program · Public research questions · Persistent Relationships
Preble, Micheal Charles. “Operant Dyad: Defining the Task-Bound Working Relationship in Human-AI Systems.” SSRN 7234238. Revised August 27, 2026. https://doi.org/10.2139/ssrn.7234238.