Do not call your LLMs conscious – they do not like that!¶
Thomas Ludwig (DKRZ Hamburg)
Abstract¶
The talk reviews the current state of the debate on large language models and consciousness. The talk introduces core concepts of consciousness from philosophy, neuroscience, and cognitive science, then contrasts them with how LLMs actually operate. It traces historical attempts to attribute mind to machines, from the Golem to Eliza, before assessing contemporary claims and theories, including global workspace theory, integrated information theory, and computational functionalism. Arguments for and against LLM consciousness are critically examined, alongside ethical and societal implications of anthropomorphizing AI. Ultimately, the talk asks what is at stake — scientifically, philosophically, and politically — when we are tempted to treat statistical pattern recognition as subjective experience.