Book Review: Thinking with AI: Machine Learning the Humanities

Edited by Hannes Bajohr, Open Humanities Press, 2025

Thinking with AI: Machine Learning the Humanities, edited by Hannes Bajohr, is a timely and intellectually ambitious collection that examines the rapidly changing relationship between artificial intelligence and the humanities. Published by Open Humanities Press in 2025, the book brings together scholars from fields including philosophy, literary studies, media studies, history, and digital humanities. Unlike many books that approach artificial intelligence primarily as a technological development or a social problem, this volume asks a more challenging question: what can the humanities learn by thinking with AI, rather than simply thinking about AI?

The central argument of the book is that AI should not be treated merely as an external technology that humanities scholars observe and criticize. Instead, concepts associated with machine learning, data processing, pattern recognition, and generative systems can become intellectual tools for reconsidering fundamental humanistic questions. These include the nature of meaning, language, representation, creativity, knowledge, history, writing, and culture. In this sense, the book does not attempt to provide a technical introduction to artificial intelligence. Rather, it uses AI as a lens through which longstanding questions about human thought and cultural production can be reconsidered.

The collection contains essays by scholars including Peli Grietzer, Leif Weatherby, Mercedes Bunz, Hannes Bajohr, Fabian Offert, Lev Manovich, Babette Babich, Markus Krajewski, Orit Halpern, Christina Vagt, and Audrey Borowski. Their contributions approach AI from very different perspectives. For example, the book considers how neural networks challenge conventional ideas about language and symbolic systems, how generative AI changes our understanding of writing, and how multimodal AI complicates the traditional distinction between text and image. Other chapters examine AI's relationship with history, aesthetics, philosophy, and the development of technological knowledge.

A particularly interesting aspect of the book is its treatment of generative AI and writing. The emergence of systems such as ChatGPT raises fundamental questions about authorship and creativity: if a machine can generate coherent text, what does it mean to write? Is generated text simply a new technological form of writing, or does it require us to rethink the concept of writing itself? Mercedes Bunz's chapter, for example, examines the mechanics of large language models alongside established theories of language and writing, arguing that generated writing may represent the beginning of a new form of writing with its own cultural logic.

The book is also valuable because it does not approach AI with either unquestioning optimism or straightforward technological pessimism. Critical AI scholarship has often concentrated on problems such as surveillance, exclusion, inequality, and the concentration of technological power. Thinking with AI retains these critical concerns but attempts to go further. It asks what AI reveals about concepts that existed long before current AI systems, including aesthetic judgment, historical interpretation, language, knowledge, and human creativity. This makes the book particularly useful for readers interested in the philosophical and cultural consequences of AI rather than simply its technical capabilities.

Nevertheless, the book has some limitations. It is an edited academic collection rather than a conventional introductory textbook, and some chapters are conceptually demanding. Readers looking for practical explanations of machine-learning algorithms, programming techniques, neural-network architectures, or step-by-step applications of AI may find the book unsuitable. Its primary purpose is intellectual exploration rather than technical instruction. Some arguments also require familiarity with philosophical and humanities scholarship, which may make sections difficult for readers coming from a purely technological background.

Overall, Thinking with AI: Machine Learning the Humanities is a valuable and contemporary contribution to the growing field of AI and the humanities. Its greatest contribution is its insistence that artificial intelligence should not be understood solely as a technological tool but also as a phenomenon capable of challenging fundamental assumptions about human knowledge, creativity, language, culture, and meaning. For researchers interested in the relationship between AI, humanities, ethics, culture, education, and society, the book provides a strong conceptual foundation. It is particularly relevant to current debates about whether AI will merely automate existing intellectual activities or instead transform the way humans understand and produce knowledge. As AI becomes increasingly embedded in education, research, cultural production, and everyday life, the book's central proposition—that we should learn to think with AI while remaining critically aware of its limitations and consequences—is both timely and significant. The book is also freely available as an open-access PDF from Open Humanities Press, making it especially accessible for researchers and students.