Can There Be Meaning Without Conscious Experience? Why Embodiment May Not Suffice for AGI

Transformer-based large language models (LLMs) exhibit impressive capabilities but lack true semantic understanding; they manipulate symbols based on patterns and probabilities without genuine meaning. Current debates often conflate operational semantic competence—achieved through human feedback and training—with intrinsic semantic understanding, which requires a connection to subjective experience. The paper defends a conscious-semantics thesis: intrinsically grounded meaning plausibly requires phenomenal consciousness. True artificial general intelligence may demand semantic understanding tied to qualia and the first-person perspective, offering more meaningful tests of intelligence than the Turing test.

Study at a glance

Abstract

The recent developments in artificial intelligence (AI), particularly in light of the impressive capabilities of transformer-based Large Language Models (LLMs), have reignited the discussion in cognitive science regarding whether computational devices could possess semantic understanding or whether they are merely mimicking human intelligence. Recent research has highlighted limitations in LLMs’ reasoning, suggesting that the gap between mere symbol manipulation (syntax) and deeper understanding (semantics) remains wide open. While LLMs overcome certain aspects of the symbol grounding problem through human feedback, they still lack true semantic understanding, struggling with common-sense reasoning and abstract thinking. This paper argues that current debates about LLM grounding often conflate operational semantic competence with intrinsic semantic understanding. While embodiment, multimodality, human feedback, and world-model learning may improve functional grounding, they do not by themselves explain why symbols or vectors should become meaningful for a subject. True meaning-making also may demand a connection to subjective experience, which current AI lacks. The path to artificial general intelligence (AGI) must address the fundamental relationship between symbol manipulation, data processing, pattern matching, and probabilistic best guesses, on the one hand, and true knowledge that requires conscious experience, on the other. I therefore defend a conscious-semantics thesis: intrinsically grounded meaning plausibly requires phenomenal consciousness. A transition from AI to AGI could necessitate semantic understanding, which is closely tied to subjective experience. This is an invitation to take the phenomenological first-person perspective seriously and to realize that intrinsically grounded meaning, as opposed to derived, relational, or operational semantics, is embedded in qualia. Recognition of this connection could furnish new insights into longstanding practical and philosophical questions for theories in biology and cognitive science and provide more meaningful tests of intelligence than the Turing test.