Are There Conscious AIs? The Quantum Darwinist Theory of Consciousness

This chapter argues that debates about AI consciousness cannot be settled by chatbot self-reports or by surface-level functional similarities; instead, they require a substrate-sensitive, case-by-case approach that asks how a system&s;s information is physically integrated and stabilised over time. The Quantum Darwinist Theory of Consciousness (QDT) connects resonance-based ideas in neuroscience (where unified experience is associated with coordinated, synchronised activity across many components) to a physical picture in which stable patterns become “objective” by being redundantly recorded through interaction with an environment (Quantum Darwinism). The resulting framework identifies a “consciousness grey zone” that includes biological computing platforms (e.g., organoid systems) and some neuromorphic/hybrid architectures, while explaining why standard large language models running on conventional digital hardware are not conscious despite increasingly human-like discourse: their consciousness-like behaviour is best explained by an error theory on which they mirror patterns in human training data without meeting QDT&s;s physical criteria. The chapter closes by drawing out ethical and policy implications of misclassification, arguing that governance should be guided by physically grounded diagnostics rather than anthropomorphic temptation.