Research Brief · Operant Dyad™
The task-bound working relationship between a person and an AI system
Micheal Charles Preble · Independent Researcher
Derived from Operant Dyad, Manuscript v2.4, SSRN 7234238 · Conceptual proposal and research agenda
The Question
What is the task-bound working relationship that forms between a human and an AI system through repeated correction, and what has to remain true of it for the human to still be its author?
The Problem
What happens when a human and an AI system form a repeated, task-bound working relationship: what accumulates across that relationship, and who still authors the result?
From the outside, a session with an AI system looks self-contained. In practice, goals, constraints, and defect reports supplied in one exchange can be retained and shape later outputs. Most accounts of AI assistance describe the output; they do not describe the relationship that produced it, or what it leaves behind.
The Contribution
Operant Dyad proposes the configured, task-bound working relationship between one human and one functionally bounded adaptive AI system as an object of study. The relationship forms through repeated correction: the human supplies goals, constraints, and defect reports; retained corrections shape later outputs; and the human inspects, accepts, rejects, or revises consequential results.
The paper introduces four constructs for describing this relationship:
- Relationship State — the retained record of goals, constraints, defect reports, and corrections that shapes later interaction, with declared state (what the person explicitly supplied) distinguished from inferred state (what the system derived on its own).
- Truxion — a proposed construct for the recurring cost of misalignment between what the person intended and what the relationship has retained.
- Portable Operant Packet — a proposed representation for carrying that relationship's continuity across sessions, applications, or systems.
- Lock-In, Override, and Hidden Scaffolding — three non-exhaustive structural failures the paper identifies, where the relationship stops serving the person's authorship.
Preserved human authorship is the governing condition for a healthy Operant Dyad, not a feature presumed to exist in current systems.
How the Framework Works
The paper is a conceptual derivation, not a report of an empirical study. It positions Operant Dyad relative to adjacent work in human-AI interaction, teaming, coordination, authorship, and data portability, and derives its constructs from the structure of the relationship itself: what accumulates, what that accumulation can distort, how it might be carried elsewhere, and where it can fail to serve the person who authored it.
The wider Operant Dyad™ research program then asks where a persistent effect of this kind is actually carried — in human learning and adaptation, in AI-side state, in explicit memory, in shared artifacts, or in some combination — through a research sequence it describes as Detect → Localize → Identify → Perturb → Probe → Reduce → Map. That sequence is under development, not a validated formal model.
Why It Matters
Naming the relationship makes it available for inspection: how it forms, what it retains, what it enables, and where human authority can weaken. If accumulated corrections, context, and unfinished decisions shape later exchanges without the person's continued oversight, the person's practical authorship of the outcome can erode even while formal control appears intact.
The framework matters to the extent that it helps distinguish a relationship a person still authors from one that has drifted into lock-in, override, or hidden scaffolding — and it does that before asking whether the relationship's underlying state is measurable at all.
Evidence status
What kind of support this framework currently has
Evidence status: Conceptual proposal and research agenda. This is not a report of an empirical study.
The wider research program has not established that a distinct, irreducible dyadic state exists. Current experimental work tests whether a persistent effect can instead be sufficiently explained by human-side state, AI-side state, explicit memory, shared artifacts, or some combination of these — before any stronger dyadic interpretation is warranted.
What This Does Not Claim
- Not an empirically validated system.
- Not a claim of exhaustive novelty.
- Not legal advice.
- Not a claim about machine consciousness or personhood.
- Does not claim that a distinct, irreducible dyadic state has been established; that question is the subject of ongoing carrier-localization research, not a settled premise of this paper.
Open Questions
Truxion and the Portable Operant Packet are proposed constructs that have not been empirically tested. Whether a persistent effect in this relationship is dyadic, or reducible to simpler carriers, is the question the active OD-INV-01 pre-pilot was designed to begin addressing. The program's public register of research questions tracks these and related open problems across 14 areas.
Read / Cite the Research
Read the canonical paper record → · Full paper PDF · View on SSRN
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.