How to Read This Book

Structure

This book is organized into four parts, each covering a distinct era in the history of artificial intelligence:

  • Part I: The Mechanical Mind (Chapters 1–3) traces the origins of computational thinking from Babbage’s engines through Turing’s universal machine to Shannon’s information theory. These chapters establish the mathematical and conceptual foundations upon which everything else is built.

  • Part II: The Birth of AI (Chapters 4–7) covers the founding of artificial intelligence as a discipline, the rise and fall of the perceptron, the golden age of symbolic AI, and the alternative approaches that emerged in its shadow.

  • Part III: The Learning Revolution (Chapters 8–11) follows the resurgence of neural networks through backpropagation, the statistical turn in machine learning, the deep learning breakthrough, and the development of sequence models and memory architectures.

  • Part IV: The Transformer Era (Chapters 12–14) traces the invention of the attention mechanism, the rise of large language models, and the alignment techniques that have made them useful—bringing us to the present moment.

The parts build on each other. Concepts introduced in earlier chapters—Turing completeness, information entropy, gradient-based optimization—reappear in later chapters in new contexts. We recommend reading sequentially at least once before dipping into individual chapters.

Technical Boxes

Throughout the book, you will encounter shaded boxes marked Technical Deep Dive. These boxes contain formal definitions, mathematical derivations, and worked examples that go beyond the narrative. They are designed for readers who want the precise details: the actual equations behind backpropagation, the formal structure of a transformer layer, the mathematics of information entropy.

If you are reading primarily for the narrative, you may skim or skip these boxes on your first pass. The main text is written to be self-contained without them. You can always return to the technical boxes later when you want deeper understanding of a specific topic.

Reading Paths

For the narrative reader: Read straight through, skipping technical boxes. You will get a complete history of AI with enough technical intuition to understand why each development mattered. Return to the technical boxes as curiosity or need arises.

For the technically curious: Read everything, including technical boxes, in order. The mathematical content is cumulative—later boxes assume familiarity with concepts introduced in earlier ones.

For the researcher: Each chapter ends with a Chapter Notes section containing historical context, pointers to primary sources, and suggestions for further reading. The annotated bibliography at the end of the book provides a curated guide to the most important references.

For the browser: Each chapter opens with a Central Question that frames its contents. Reading these fourteen questions in sequence gives a compressed map of the entire book. Start with whichever question most interests you.