In Nexus, renowned historian and author Yuval Noah Harari continues the exploration he began in several previous volumes (Sapiens, Homo Deus, 21 Lessons for the 21st Century) of how humankind was able to harness information to its advantage—and remarkable development as a species—and what the continuing evolution of those information networks might hold for our future. In this volume, the author takes a more focused look at how societies from the Stone Age to what he calls the modern Silicon Age have seen their development influenced and determined by the way in which information is created, gathered, utilized, and controlled. While a considerable portion of Nexus is devoted to exploring how information networks developed historically—including an interesting comparison of the way information is used in democratic and totalitarian societies—the real essence of the narrative involves a discussion of how artificial intelligence (AI)-based networks represent a fundamental shift from the past in how information is generated and the purposes for which it is used. Importantly, Harari argues, this shift holds both the promise of vast improvements in our collective well-being as well as the peril of our potential self-destruction. Although I suspect this book will have its critics (e.g., the author’s analysis is too shallow and too lurid; much of the substance is highly repetitive of his earlier work), the fact is that there is a lot to savor here. Harari argues persuasively that advanced societies could not exist without the ability to harness and share information through the stories they tell, the religious beliefs they adopt, and the bureaucratic mechanisms they form. However, while that information is crucial for cooperation, it can also be misused for control. Because of that, it is important to recognize that information itself is neither the same as truth, nor is it infallible. So, for that reason, any well-functioning information system must include effective self-correction tools that can unwind adverse outcomes resulting from data being used to reach a goal that is inconsistent with what was originally imagined and intended. The rise of AI-driven information systems presents a major challenge because their complexity may be beyond human understanding, they are able to learn and create new information on their own, and they typically lack sufficient self-correcting mechanisms. The author is not optimistic about our willingness and ability to manage where this new technology may take us in the future. While Nexus has clearly been crafted for the inquisitive lay reader, it is very much the work of an academic scholar, which has both its good and bad aspects. On the positive side, the discussion in the book is extremely thought-provoking as well as being well-grounded in how the historical development and futuristic speculations tie together. The main frame of the centrality of information networks to our everyday lives and how those systems have evolved to where they may end up beyond our control is essential material to ponder. Conversely, the book is often bloated with what seemed like superfluous side-stories that did not fit the main purpose (e.g., descriptions of centuries-old religious disputes and military strategies) and it is, if anything, over-researched; fully one-quarter of the volume simply contained the reference footnotes for the main text. Overall, given that the role and the use of AI are among the central questions that we are likely to be debating for years to come, it is easy to recommend this book as one that puts so many of the relevant issues on the table, even if it does not really offer any concrete solutions.

In the prologue to Nexus, author Yuval Harari says, “The main argument of this book is that humankind gains enormous power by building large networks of cooperation, but the way these networks are built predisposes them to use power unwisely. Our problem, then, is a network problem.” Networks, though, come in many forms. Harari explores information networks from man’s earliest days forward, beginning with our earliest information networks, stories. He traces the growth of information through biblical times to more modern history like witch hunts and Stalin’s collectivization. He then explores the inorganic networks that have become prominent in our current times and the implications of modern networks on politics. Networks have become more and more powerful, and the author says that in general this power allows them to thwart threats from outside but also increases the danger from within. His final recommendation, which he labels the most important takeaway from his book, is that we abandon naïve and populist views of information such as he describes in the book to build institutions with strong self-correcting mechanisms The Acknowledgments section says that he began working on this book in 2018. From the sheer number of sources he cites, I can easily believe the work took that long! The breadth of the research is impressive, including many works from and about history but then also modern sources like Ray Kurzweil, Robert Sapolsky, PricewaterhouseCoopers, Elon Musk, and Donald Trump! It also made the book fascinating, informative, and added to my already-too-long list of books I want to read! The breadth of the examples described is equally informative and fascinating, such as China’s initial failure to appreciate the potential of the Industrial Revolution and the shock when Google’s AlphaGo AI defeated the human Go champion Lee Sedol. As a devoted catlover, I was especially amused by Google’s cat-image-generating AI called the Meow Generator. The book and its ideas were very interesting, but the best thing about it was that it engaged me and made ME think. Despite the overall credibility of his ideas, I did not always agree with him, and I would have loved to engage him in a discussion. That is a successful book and message, and I was even more convinced of its success when I found myself telling a friend about one of Harari’s ideas before I even finished the book. I will certainly be nominating it for my book group! Despite my high recommendation of the book and its ideas overall, I feel compelled to warn readers that Harari develops his message through well done examples and exploration of situations. He generally does not very explicitly define many of the terms he uses, even when he is using them in a rather specialized way , which can be confusing and annoying. This is a special problem in the beginning of the book. The Prologue left me so unsatisfied that I was tempted not to continue reading but decided to persist based on my enjoyment of his book Sapiens. I am glad I did. So if your initial reaction to the book is similar to mine, do persist. The author gets his point across, and it is an interesting one! I received an advance review copy of this book from NetGalley and Random House.