Money in AI-Operated Production Networks: A Thought Experiment on Future Production Relations

This paper examines monetary coordination in a hypothetical economy in which highly capable artificial-intelligence systems perform nearly all productive labour. The counterfactual assumes broad task versatility while retaining constraints on physical resources, production time, infrastructure, and the authority to dispose of particular inputs. By distinguishing labour performance from access to production conditions and the coordination of claims, the paper argues that extensive automation is compatible with the continued use of monetary or money-like institutions, although scarcity alone does not make such institutions necessary. Drawing on monetary search theory, earlier artificial-agent exchange models, and institutional analysis of resource rights and productive dependence, the analysis separates the functions of account, credit, settlement, and general exchange. It also distinguishes AI as a functional production node from legal and moral subjecthood, and treats prices and payment instruments as manifestations of potentially deeper resource and governance relations. The analysis treats money as a broad functional category and currency as one possible circulating institutional form. It allows locally generated records, claims, or resource objects to acquire some monetary functions through adaptive AI-to-AI interaction, even without a prior common currency or a fully specified normative exchange code. The central contribution is a conditional account of monetary emergence, persistence, and possible institutional substitution under predominantly AI-to-AI production. The paper makes no prediction that money will survive or disappear, and prepares a companion study of distributed monetary dissolution in generative value circuits.