Start with the controlling sources. Then follow a concept into the paper that develops it, the tool that applies it, or the field note that tests its boundary.
The shortest path through HAOP: understand the integrated framework, see the complete work-design sequence, then run the entry diagnostic against one bounded workflow.
The controlling public source for the three performers, five-hazard register, five operating modes, Accountability by Control, True Function, verification architecture, grounding, and the applied method.
ARECC applied to AI-enabled work systems as a ten-step sequence from True Function and work-as-done through control engineering and revalidation. Revision 6.0.1 Section VIII is the current statement.
A browser-local entry tool that reads the gap between work-as-imagined and work-as-done in one workflow. It does not certify safety or replace the full HAOP method.
These works control definitions, chronology, and the development boundary of the integrated framework.
Develops friction, choke points, Verification Requirement and Capacity, Verification Gates and Windows, transition states, and the Workflow Profile.
The hazardous condition in which a person is held accountable for a verification that exceeds the capacity to verify under the conditions provided.
Develops the workflow-level mechanism by which recursive transformation can suppress correction signals while the system continues to appear functional.
Framework concepts explained in place, with the deposited definitions quoted and cited. Appendix A of Revision 6.0.1 controls every definition; the pages do not substitute for it.
Responder, Signal Amplifier, Executor, Orchestrator, and Ensemble. Behavior-based and non-hierarchical, with the control surface each mode requires.
A transition-specific condition of contact. Four forms, the anchors that produce them, renewal, false grounding, and Anchor Access.
The functional performer threshold: when AI stops being only a tool, and why the organization remains in the account.
Development status is explicit. These artifacts support learning and work design; they are not presented as validated instruments. Additional practitioner tools are in development.
Captures the deposited nine questions across the three performers, reads the work-as-imagined vs work-as-done gap, points into the five-hazard register, and produces a downloadable design record. Browser-local; sharing the finished record is opt-in only.
The printable worksheet for mapping the nine questions across the three performers on paper, away from the screen.
Practitioner operationalization of the ten-step method and its design record. Not yet deposited or empirically validated.
Companion volume to the Workbook. Not yet deposited; not presented as validated.
A bidirectional hazard-communication instrument between the AI product boundary and the adopting work system. The complete standalone design remains in development.
The university side and the challenge side. Exchanges that sharpen the framework publish, with permission, in the correspondence record.
The research opportunity, the standing challenge, and what exists to study, on their own page.
Public writing that introduces, tests, or illustrates HAOP ideas. These pieces are kept distinct from the deposited framework record.
AI does not choose whether to augment or surveil. The organization sets that direction before the technology reaches the worker.
When AI moves from bounded assistance into operational influence by shaping sequence, priority, visibility, recommendation, routing, approval, or execution.
A systems argument about recursive synthetic data, weak-signal loss, and removing the human system that keeps AI in contact with operational reality.
Challenges, questions, and answers on the framework, published with the sender's permission. Contrary evidence is a contribution.