Review of Cameron J. Buckner’s <i>From Deep Learning to Rational Machines: What the History of Philosophy Can Teach Us about the Future of Artificial Intelligence</i> - Cameron J. Buckner, From Deep Learning to Rational Machines: What the History of

From Deep Learning to Rational Machines is an ambitious book that aims to show both that contemporary machine learning (ML) can meaningfully inform philosophy and that the history of philosophy can help guide the development of future ML systems.This is an exciting, timely, and valuable project.Furthermore, in service of these two goals, the book provides an accessible introduction to an extraordinarily successful approach to machine learning, termed deep learning.This will prepare readers to follow the main arguments in the book, as well as other philosophical work on machine learning, which is in itself a valuable contribution to the field.The book is split into seven chapters, with the first two providing a very useful orientation on both contemporary ML and the nuances of the nativist-empiricist debate, and each of the last exploring a different cognitive faculty through the lens of a particular empiricist.Chapter 1 reviews the nativist-empiricist debate.Buckner's central contributions are (i) a recasting of the different positions in the debate with a more manageable continuum of theories, and (ii) a proposal for a modest empiricism of multiple domain general faculties.This proposal, which Buckner calls Domain General Modular Architecture (DoGMA), is inspired by medieval and early modern natural philosophers.