What do you think?
Machine Agency
An engaging exploration of agency that provides students with the critical tools needed to understand and participate in debates about future machines.
The great promise of artificial intelligence’s evolution lives alongside an equally great anxiety. As we develop increasingly autonomous machines that do things in the world, questions about agency—distinguishing machines that can act from those that cannot—are among the thorniest we face. A concise and probing exploration of agency, this accessible textbook provides the critical, technical, and conceptual tools needed to make sense of rapid changes in what machines can do and their role in our lives.
James Mattingly and Beba Cibralic begin with an examination of foundational What is agency? How does it differ from mindedness, consciousness, and intelligence? Can we attribute agency to certain machines, and if so, how and why? They then examine the social and ethical implications of building ever more complex machines, including those concerning moral status and responsibility. Drawing together ideas from philosophy and computer science as well as from information theory, literature, and the history of science, Machine Agency invites students to participate thoughtfully in critical debates about future machines.
• Provides a roadmap for interrogating the concept of machine agency suitable for philosophy majors and non-majors alike
• Investigates the connections between developments in AI and pressing issues in analytic philosophy
• Explores the social and ethical impacts of computational systems with agency
Features robust end-of-chapter exercises
The great promise of artificial intelligence’s evolution lives alongside an equally great anxiety. As we develop increasingly autonomous machines that do things in the world, questions about agency—distinguishing machines that can act from those that cannot—are among the thorniest we face. A concise and probing exploration of agency, this accessible textbook provides the critical, technical, and conceptual tools needed to make sense of rapid changes in what machines can do and their role in our lives.
James Mattingly and Beba Cibralic begin with an examination of foundational What is agency? How does it differ from mindedness, consciousness, and intelligence? Can we attribute agency to certain machines, and if so, how and why? They then examine the social and ethical implications of building ever more complex machines, including those concerning moral status and responsibility. Drawing together ideas from philosophy and computer science as well as from information theory, literature, and the history of science, Machine Agency invites students to participate thoughtfully in critical debates about future machines.
• Provides a roadmap for interrogating the concept of machine agency suitable for philosophy majors and non-majors alike
• Investigates the connections between developments in AI and pressing issues in analytic philosophy
• Explores the social and ethical impacts of computational systems with agency
Features robust end-of-chapter exercises
184 pages, Paperback
Published February 4, 2025
Ratings & Reviews
Friends & Following
Create a free account to discover what your friends think of this book!
Community Reviews
Displaying 1 - 1 of 1 review
July 2, 2025
Interesting & accessible philosophical textbook on the topic of agency in the age of automation— What is agency? Who (or what) possess it? Can we hold machines (or non-anthropomorphic entities) morally accountable for their actions?
In an era of digitization, I think these are the types of ethical and philosophical questions we need to ask ourselves in order to design better systems of AI governance. As the lines between human/machine and animate/inanimate object become blurred, the designation of agenthood holds both important descriptive and normative implications.
The book is written very intuitively for amateur practitioners of philosophy, and so I think it is best reserved for new readers / students seeking a high-level primer about AI, machine agents, and synthetic actors. There are 11 chapters that include follow-up questions, activities, and resources that allow for a more in-depth look at some of these topics.
I'm not 100% sure if I came out of the book with a definitive opinion of what makes a thing "agentic" or not, but I'm also not sure that's the point (any attempt to do so may be imposing a procrustean framework on a protean topic). Instead, I found myself developing a greater appreciation for rich philosophical debates in the vein of autonomous agents including: mind/body dualism (shoutout Plato/Descartes/Socrates), famous thought experiments in the philosophy of language / computing (Turing's imitation game, the decision problem, Searle's Chinese Room argument, the Octopus test), belief-desire vs reductionist models of agency, Complex Adaptive Systems (CAS) frameworks, information theory / cybernetics, and questions of moral status in macroethics.
I think a good book doesn't need to provide all the answers, but instead, encourage the reader to seek out additional questions.
I also read these two articles, and they gave a very fascinating view of what it means to be a moral agent / moral patient (e.g., does A = P or A ⊂ P?)
- Floridi & Sanders. 2004. “On the Morality of Artificial Agents.” Minds and Machines.
- Johnson & Miller. 2008. “Un-Making Artificial Moral Agents,” Ethics and Information Technology
In an era of digitization, I think these are the types of ethical and philosophical questions we need to ask ourselves in order to design better systems of AI governance. As the lines between human/machine and animate/inanimate object become blurred, the designation of agenthood holds both important descriptive and normative implications.
The book is written very intuitively for amateur practitioners of philosophy, and so I think it is best reserved for new readers / students seeking a high-level primer about AI, machine agents, and synthetic actors. There are 11 chapters that include follow-up questions, activities, and resources that allow for a more in-depth look at some of these topics.
I'm not 100% sure if I came out of the book with a definitive opinion of what makes a thing "agentic" or not, but I'm also not sure that's the point (any attempt to do so may be imposing a procrustean framework on a protean topic). Instead, I found myself developing a greater appreciation for rich philosophical debates in the vein of autonomous agents including: mind/body dualism (shoutout Plato/Descartes/Socrates), famous thought experiments in the philosophy of language / computing (Turing's imitation game, the decision problem, Searle's Chinese Room argument, the Octopus test), belief-desire vs reductionist models of agency, Complex Adaptive Systems (CAS) frameworks, information theory / cybernetics, and questions of moral status in macroethics.
I think a good book doesn't need to provide all the answers, but instead, encourage the reader to seek out additional questions.
I also read these two articles, and they gave a very fascinating view of what it means to be a moral agent / moral patient (e.g., does A = P or A ⊂ P?)
- Floridi & Sanders. 2004. “On the Morality of Artificial Agents.” Minds and Machines.
- Johnson & Miller. 2008. “Un-Making Artificial Moral Agents,” Ethics and Information Technology
Displaying 1 - 1 of 1 review