Machine consciousness, behaviorist epistemology, and the economics of replacing human roles
Siti Hurul Aini
Journal of Economic Epistemology and Philosophy September 10, 2026 DOI: 10.55942/jeep.v1i2.2299 (opens in new tab)
Study at a glance
AI-extracted from the abstract| Characteristics | Theoretical or philosophical paper Peer reviewed |
|---|---|
| Key points | Argues that economics and computing made the economic agent consciousness-agnostic by construction, leaving the value of roles occupied by subjects of experience unpriced. Proposes that this 'phenomenal residual' creates a wedge toward excessive automation, that consciousness enters substitution decisions only insofar as it is priced, and that because AI instances can be copied cheaply, even small credences in machine sentience can dominate social evaluation. |
Abstract
Whether artificial intelligence (AI) can be conscious is usually treated as a question for philosophers of mind, and the replacement of human labor by AI as a question for economists. This article argues that the two questions are entangled and that their separation is an artifact of the epistemology of modern economics. In a documented corpus of 100 Scopus-indexed records on AI, labor, and morality (2020–2026), responsibility vocabulary appears in 42.6% of 68 retrievable abstracts and consciousness vocabulary in one; an independent query of 11,454 records confirms that consciousness and sentience are rarely discussed (0.7%). A genealogy traces this silence to twin behaviorist turns—revealed preference in economics and the imitation game in computing—that made the economic agent consciousness-agnostic by construction. The article models a human role as functional, agential, and phenomenal layers and defines the phenomenal residual as the unpriced value of a role that depends on its being occupied by a subject of experience. A schematic welfare account yields three propositions: consciousness enters private substitution decisions only insofar as it is priced; unpriced residuals create a wedge toward excessive automation; and, because AI instances can be copied cheaply, even small credences in machine sentience can dominate social evaluation. A counterpoint from Islamic legal and economic thought supports a principle of non-delegable answerability, and five principles for a consciousness-sensitive economics under Knightian uncertainty conclude the article.