Communicative AI: A Critical Introduction to Large Language Models

by Mark Coeckelbergh, David J. Gunkel

Large Language Models (LLMs), like OpenAI’s ChatGPT and Google’s LaMDA, are not only the most disruptive and controversial technologies of our time, but also offer an unprecedented opportunity to examine human cognition and philosophically question the very nature of language, communication, and intelligence. What is consciousness? What is language? Are LLMs authors? Are LLMs the end of writing as we know it?

In Communicative AI, Mark Coeckelbergh and David J. Gunkel offer a critical introduction to LLMs, investigating the philosophical significance of this technology and its practical ramifications. Mobilizing resources from contemporary philosophy, history of ideas, linguistics, and communication theory, the book invites us to re-think some long-standing philosophical issues concerning language, consciousness, truth, authorship, and writing. Through a blend of theoretical analysis, accessible explanations, and practical examples, the book provides readers with a comprehensive overview of the role that this powerful new technology is already playing in our lives.

This is a must-read for students and scholars across the humanities and the social sciences, as well as for anyone intrigued by the intersection of technology, language, and human thought.

Genres: Artificial Intelligence

144 pages, Hardcover

Published July 16, 2025

About the author

Mark Coeckelbergh is Professor of Philosophy of Media and Technology at the University of Vienna. He is the author of New Romantic Cyborgs: Romanticism, Information Technology, and the End of the Machine, AI Ethics (both published by the MIT Press), Introduction to Philosophy of Technology, and other books.

Reviews

[Kafkaesk · Rating 3 out of 5]

I certainly don’t know how to rate this, much of the proposed thoughts and conclusions *do* make sense. If we take deconstruction and Derrida to their utmost implications and apply that to LLMs, then yes, LLMs and 'AI‘ accelerate this gesture of de-masking written language, and logocentrism in general.

What I have a hard time grasping and accepting is, doesn’t this whole Spiel rather show the ultimate perverse conclusion of deconstruction? In that case the book is less about LLMs in general, but more about the ultimate implications of derrida‘s 'disciples' shown through their handling of LLMs. In that case it morphs into some kind of accelerationist half-stepping, where the intent to accelerate these processes is hidden, maybe even unconsciously proposed - I don’t think that Coeckelbergh and Gunkel consider themselves as accelerationists, but for me it ultimately falls into that current.

Nonetheless, it opens interesting questions; if we follow this rift, which was opened by the authors, we must rather think about different modes of understanding certain texts, and their intention. Not all texts are created equal, and in case of LLMs, they’ve probably not been created at all. They’ve been calculated. What does this imply then?

[Joel]

A helpful book for thinking about AI in the larger context of what writing and language are and what they mean with or without a human 'author.'

The authors are, I believe, philosophers, so I couldn't quite connect with some of their points about, e.g., Plato. I didn't really agree with their conclusions (basically that 'writing ' can exist without a human writer and that's OK for various reasons), though I did find their premises intriguing and reasonable.

It ticks me off that it took these stupid tech companies to finally force me to get into philosophy of language, but I guess I'm going to have to.