Nexus by Yuval Noah Harari: The Case for History

Yuval Noah Harari, the secular-Israeli historian and philosopher who shook the world with the publication of his hit book Sapiens a decade ago, and whose talks at Davos made him a top boogeyman amongst conspiracy theorists, has just published a new book Nexus, which claims to offer a comprehensive history of information up to the present day. Though the book is quite vacuous in its arguments, as Harari is appealing to the common reader as opposed to fellow academics, it makes a compelling and urgent case for the study of history, particularly amongst technologists working on advanced artificial intelligence. Unfortunately, no matter how many talks the Hebrew University of Jerusalem polemicist gives, and how many books the already New York Times bestseller disperses, it is unlikely to change the conversation around AI, as our culture increasingly incentivizes us to focus on short-term concerns. The consequences could be fatal not just for all of us, but for consciousness in general. 

Though Nexus would hardly be taken seriously in academic circles, Harari is far from a crank. He is a very serious historian whom both medieval and military specialists hold in high regard. What he is doing in this book, like in his past works Sapiens, Homo Deus, 21 Lessons and even The Ultimate Experience, is zooming out for perspective. AI should be understood in the context of millenia of experience in many realms. In this sense, YNH is a wise old sage amongst a bunch of children with dangerous toys. But this does not mean he can stop the kids from self-destructing.

His consistent worry is that AI could lead to a totalitarian dictatorship the likes of which we have never seen. Kim Jong-Un or Vladimir Putin, for example, will have mechanisms of social control Mao, Hitler and Stalin could not even dream of. Biometrics will make it so that future dictators can know not only everything their citizens do, but also what they feel and think. 

But fear not, for the same technologies could also give us the best healthcare and sex we have ever had. This paradox is at the heart of Harari’s work. He does not believe in technological determinism, for if us animals could not change the future, then what is the point in talking about it? 

Fair enough. But talking and writing about it is not enough. Tech executives x, y and z may read the book and discuss it over drinks with friends, but then go back to focusing on micro concerns like the next update in their preferred Large-Language Model (LLM.) This is how we will destroy ourselves, first gradually, with new updates every few weeks, and then suddenly. It is only through understanding these developments in the long-duree that we can mitigate the worst outcomes. 

What is needed, then, is history departments within business, government and civil society organizations working on these questions. We cannot make the mistake of thinking Harari’s book is enough; it should be one amongst hundreds of historical perspectives. A dozen historians working full-time at OpenAI, or in the federal government, and dispersed throughout NGOs, with direct access to decision makers, could be what saves our species. 

This is not as difficult as it sounds. History PhD stipends are a starvation wage and there are hardly any jobs to be found once you graduate. These entities could pay historians a small fraction of what they pay technologists and they would jump at the opportunity. A relatively tiny donation by Elon Musk or Bill Gates or some other billionaire could send large swathes of the field into an AI research frenzy. 

AI could in turn upend the historical profession, causing a cascade of accelerating returns. If computers are bound to turn the tables on us, it is best we get them on our side. With a little motivation and understanding of history, they may end up aligning themselves. 

Google started, after all, via the digitization of the University of Michigan library. History is a vast resource offering unlimited wisdom about how to proceed. Yet Harari’s books are the only extraction mechanism we are using. Nexus should be just a drop in a far wider web of analysis of the past and what it portends for the future.

Larry Page and Sergey Brin. Courtesy: Pinterest.

Though history is important, we have never been here before. AI is unique in that it can generate new ideas. YNH likes to remind us that while intelligence is mandatory, consciousness is optional. The risk is that training this alien intelligence could in turn make it better at manipulating humans, as it is already doing, arming our enemy with the lessons of history. Worst case scenario, they pull the plug and end all sapience. But if we are all dead then what will it matter? 

I give Harari’s book a B. It should be the next one you read, hopefully this weekend. Yuval is over the target, and focused on the most important questions of our time, most of all due to his mindfulness and meditation practices. 

Nexus should not be the end of the story, however. Let us hope that it sparks a renaissance of human and computer research on the lessons of the past, if not for our own sake, for that of our species. AI poses unprecedented threats and possibilities, and we need all the help we can get. Humans are still in control, “We command immense power and enjoy rare luxuries, but we are easily manipulated by our own creations and by the time we wake up to the danger, it might be too late.” 

Indeed.