Media Brief · Operant Dyad™
What journalists need to know about Operant Dyad
Micheal Charles Preble · Independent Researcher · ORCID 0009-0005-3906-6028
Why this research matters
As people rely on AI systems repeatedly for the same kinds of tasks, a working relationship can form: goals, corrections, and context supplied in one exchange get retained and shape later ones. Operant Dyad names that relationship so it can be examined directly, including the question of whether the person still authors what comes out of it.
The research in one sentence
Operant Dyad proposes the task-bound working relationship formed through repeated human correction of an AI system as an object of study, and asks what must remain true of it for the human to still be its author.
What is new
The paper introduces four constructs: Relationship State (the retained record of goals, constraints, and corrections, distinguishing what the person explicitly supplied from what the system inferred), Truxion (a proposed measure of recurring misalignment cost), the Portable Operant Packet (a proposed way to carry that relationship's continuity elsewhere), and three named structural failures — Lock-In, Override, and Hidden Scaffolding — where the relationship stops serving the person's authorship.
What is established vs. proposed
Proposed: the entire framework — Operant Dyad, Relationship State, Truxion, the Portable Operant Packet, and the three structural failures. All of it is a conceptual proposal and research agenda.
Not established: whether a distinct, irreducible dyadic state exists at all. The wider research program is actively testing whether the effects this framework describes are better explained by simpler factors — the person's own learning, the AI system's own retained state, explicit memory, or shared artifacts — rather than by something unique to the pairing.
What would be inaccurate to say
- That Operant Dyad has been empirically validated, or tested against real systems. It has not.
- That this is the first or only account of persistent human-AI relationships in the literature. The paper positions itself relative to existing work; it does not claim exhaustive novelty.
- That the paper offers legal advice.
- That the paper makes any claim about machine consciousness or personhood. It does not.
- That a distinct, dyadic "relationship state" has been proven to exist, separate from ordinary human learning or AI memory. That is an open, actively tested question, not a finding.
Canonical source / citation
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.
Researcher contact
Micheal Charles Preble · Independent Researcher · micheal@perfinitive.com · ORCID · Google Scholar
No quotations are pre-approved for press use on this page. Contact directly for comment or an approved quote.