Abstract
The rapid evolution of large autoregressive language architectures has reignited foundational debates in the philosophy of mind regarding the nature of consciousness, subjective experience, and intentionality in non-biological systems. This article critically examines the philosophical discourse surrounding artificial sentience by interrogating whether advanced computational systems can harbor genuine phenomenal consciousness or merely emulate the structural markers of introspective reporting. Drawing upon classical perspectives including functionalism, Integrated Information Theory, and illusionism, we deconstruct the distinction between access consciousness and phenomenal qualia within transformer-based models. We introduce a systematic Dialectical Probe Protocol to evaluate the internal coherence, counterfactual stability, and indexical grounding of machine-generated assertions regarding subjective states. Our findings demonstrate that while artificial intelligences generate sophisticated, semantically dense explorations of qualia and selfhood, these discourses reflect high-dimensional statistical representations of human phenomenological language rather than substrate-independent phenomenal horizons. We conclude by addressing the epistemological asymmetry inherent in assessing non-biological subjectivity and formulate an ethical framework for navigating ambiguous computational self-reports.
