Communicating with AI

Abstract

This volume, Communicating with AI: Philosophical Perspectives, examines a pressing question of our time: what does it mean to communicate with large language models (LLMs) such as ChatGPT, Claude, Gemini, and DeepSeek? With billions of human–AI interactions on the horizon, understanding whether these systems share our language, possess understanding, or function instead as novel instruments is an urgent theoretical and practical task. The essays collected here bring the tools of philosophy to bear on the distinctive challenges posed by AI. The contributions are organized into four parts. Part I probes the foundations of meaning, asking whether LLM outputs can be genuinely meaningful or whether they should be thought of as mere signals or measurements. Part II explores the possibility of alien semantics and externalist accounts of AI meaning, pushing the boundaries of how different from us an AI mind or language could be while still remaining intelligible. Part III investigates the social and normative dimensions of AI communication, focusing on trust, responsibility, and the risks of projecting human concepts onto machine outputs. Part IV turns to the future, considering the architectures of artificial minds and the unsettling ethical possibilities of superintelligent systems that may reason about values in ways inaccessible, or even hostile, to human interests. A central methodological theme runs throughout: the need to de-anthropomorphize our understanding of communication without emptying it of substance. By comparing AI with animals, aliens, and deities, the essays highlight both the promise and the limits of extending our concepts of mind and meaning beyond the human case. Collectively, the volume offers a rigorous philosophical framework for grappling with the metaphysical, epistemic, and ethical stakes of our emerging conversations with AI.