"When I heard that John MacCormick was writing a book on AI, I was immediately eager to read it, and this book does not disappoint. Directly addressing a key question of our age, he brings extensive historical and scientific research and a deep understanding of the internal workings of modern AI models to produce a beautifully calibrated exposition. I shall be recommending this book to basically everyone I know." --Andrew Fitzgibbon, Engineering Fellow, Graphcore "This book is a much-needed addition to the public conversation around AI. MacCormick cuts through the immense media hype to provide an accessible and fascinating exploration of how machines can appear to think like humans, equipping readers to form their own conclusions about the provocative question that Turing posed over fifty years ago."-- Kerry McInerney, University of Auckland and Leverhulme Centre for the Future of Intelligence, University of Cambridge " Thinking AI does a wonderful job of making the foundations of modern artificial intelligence systems accessible to nonexperts. It explains the building blocks of these systems--deep neural networks and reinforcement learning. Using illustrative examples, it examines the connections between AI systems and biological brains and offers nuanced answers to weighty philosophical questions, such as whether machines can fully emulate human intellect and creativity.

" --Marc Najork, Distinguished Research Scientist, Google DeepMind "A worthy and thoughtful contribution to the public's understanding of AI." --Rich Sutton, winner of the 2024 Turing Award "After so much has been written about dangers and threats from AI, it is refreshing to read this celebration of AI from a highly articulate teacher. John MacCormick reminds us of the huge intellectual achievement that modern AI represents and shows us the very high degree to which it can emulate human-like understanding."-- Andrew Blake, University of Cambridge, former Director of Microsoft Europe.