Answers are grounded in this paper and cite their sources.
Work that crosses programs, APIs, organizations, and people can outlive any one execution context. A timeout may leave the outcome of an external action unknown; a restart may erase the information needed to decide what is admissible next. This article proposes a compact organizational analogue of Alan Perlis's pairing of a loop and a structured variable: the control loop and the state structure. The control loop relates observed consequences to a goal and selects clarification, retry, compensation, or escalation. The state structure preserves participants, obligations, confirmed facts, unknown outcomes, and transition history. The proposal is connected to the Conant-Ashby regulator theorem and developed through actor graphs with stateful mediator actors, persistent interaction history, and accounting consequences. A bank debit example separates safety, liveness, and conformance. The practical criterion is a handoff test: another executor should be able to resume unfinished work from saved state without reconstructing the system's memory by hand.
No takes yet. Share an insight, caveat, or question.
Alexander Vityaz (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: