Contents

To understand today's revolution, you have to know where it came from. This first part tells the story of the sixty years that preceded ChatGPT, then lays out the fundamental mechanics of machine learning, right up to the threshold of the innovation that changed everything: the Transformer.

With the foundations laid in Part I, it is time for the headliners of the era. This part dissects large language models (Chapter 4), broadens the view to world models and the multimodal (Chapter 5), shows how models become agents capable of taking action (Chapter 6), then maps out the players and the people who make AI (Chapter 7).

We tend to think of AI as software, almost immaterial. That is an illusion. Behind every response from a model lie giant factories, chips that rank among the most complex objects ever manufactured, and an appetite for energy that worries electrical grids the world over. This part descends into the engine room: compute and chips (chapter 8), the alternative of local and open source AI (chapter 9), and the energy question that could bring it all to a halt (chapter 10).

AI does not advance in a vacuum. It meets other disruptive technologies, and from these encounters spring promises (and a great deal of hype). This part examines three major convergences with a clear eye: AI and the blockchain (chapter 11), AI and quantum computing (chapter 12), and the most tangible of the three, AI and robotics (chapter 13), where intelligence finally takes on a body.

After the foundations, the models, the hardware and the convergences, it is time to look at the concrete effects of AI on our lives. This part, the largest in the course, surveys the major fields transformed by AI: science and health (chapter 14), education (chapter 15), creation and culture (chapter 16), work and the economy (chapter 17), finance and other sectors (chapter 18), everyday life and psychology (chapter 19), cybersecurity (chapter 20), law and society (chapter 21), and finally defense (chapter 22). Each time, the same guiding thread: weighing the real promises against the concrete risks, without hype or doom-mongering.

Here we are at the end of the journey, facing the most dizzying questions. After the concrete uses, this final part rises toward the ultimate questions: the philosophical ones that AI revives (chapter 23), such as understanding, consciousness and moral status; the challenge of alignment and safety, that is, how to keep a highly capable AI under control and faithful to our values (chapter 24); and finally governance, regulation and the possible futures of such a powerful technology (chapter 25). This part calls for clear-sightedness, and the humility to acknowledge what no one yet knows.