Why Consciousness? Integration as a Cost-Benefit Threshold
Marc Lebel
Zenodo (CERN European Organization for Nuclear Research) March 18, 2026 preprint DOI: 10.5281/zenodo.19099907 (opens in new tab)
Study at a glance
AI-extracted from the abstract| Characteristics | Theoretical or philosophical paper |
|---|---|
| Key points | Argues that consciousness exists because, above a critical threshold of inter-channel conflict, the fitness cost of incoherent responses exceeds the metabolic cost of integration, and that a minimal evolutionary simulation produces a sharp phase transition rather than a gradual continuum. Proposes that artificial consciousness requires both hardware conditions (endogenous energy budget, multi-channel conflict, stakes) and software conditions (rich, high-dimensional arbitration), which current AI architectures lack. |
Abstract
Why does consciousness exist at all? This essay argues that the question, often treated as metaphysical, admits a structural answer. Consciousness — understood as costly global integration across otherwise independent processing channels — exists because, above a critical threshold of inter-channel conflict, the fitness cost of incoherent responses exceeds the metabolic cost of integration. A minimal evolutionary simulation confirms that this logic produces a sharp phase transition rather than a gradual continuum. Three philosophical consequences are drawn: the question "why consciousness?" is structurally analogous to "why flight?"; the hard problem is isolated rather than dissolved; and artificial consciousness requires both hardware conditions (endogenous energy budget, multi-channel conflict, stakes) and software conditions (rich, high-dimensional arbitration). A thermostat satisfies the hardware conditions; a brain satisfies both; current AI architectures satisfy neither.