Why Nexus Overshadows Sapiens as a Must Read for Tech Executives
I read Yuval Noah Harari’s new book so you don’t have to (but you probably should).
It’s been nearly a decade since Yuval Noah Harari released Sapiens, an ambitious and wide-ranging history of humankind. The book took Silicon Valley by storm, earning the endorsements of tech titans like Bill Gates and Mark Zuckerberg.
It was the intellectual flex of its time—a tech-world rite of passage. Whether you read Sapiens cover-to-cover or just skimmed the dust jacket blurbs to sound smart at cocktail parties, Harari’s latest addition to his canon is likely to pique your interest.
Last week, Harari released Nexus: A Brief History of Information Networks from the Stone Age to AI. As a big fan of his previous work, I bought it and immediately shotgunned it like a madman.
In Nexus, Harari applies the same broad strokes lens that helped him tell the whole of human history in 400 pages to what may be the biggest question of our time… What impact might widespread adoption of AI have on our civilization?
Harari posits that this isn't necessarily uncharted territory for us as a species. Generative AI is simply a technological leap in an information system. This is just the latest chapter in a millennia-long saga of how humanity builds, manipulates, and struggles with information networks. Does ChatGPT present some unique challenges compared to the printing press? Absolutely, but we can learn a lot about how we design this new information system by the mistakes of our past.
Personally, I think this book is an excellent read for anyone who is interested in AI and the evolution of the Internet over the past (and the next) 30 years, but most of the tech executives I work with don't have time for a full anthropology lesson for homework. As such, I'm going to give you a quick overview of the book, highlight key insights relevant for tech leaders, and recommend some reflection questions you can use to implement these.learnings in your organization.
What You Need To Know: A Brief Summary of Nexus
As the title suggests, Nexus doesn’t just cover AI—it spans the entire history of human information networks, from the Stone Age to the present AI explosion. Harari’s thesis is that the AI revolution is not the first time humanity has grappled with the profound implications of a new information network. Spoken language, writing, the printing press, radio, television—each of these represented a seismic shift in how societies functioned. AI is simply the next step in that lineage, but it’s also the most powerful—and dangerous—network we’ve ever built.
Harari warns that while this is familiar territory in many ways, the scale and speed of the changes AI brings are unprecedented. The stakes are higher, and we need to learn from our past mistakes in managing these information networks if we want to avoid catastrophe.
For tech leaders, this message is especially potent. We’re managing complex, interdependent information networks. The way we design these systems will determine not just our success and failure, but possibly the broader shape of the human experience for centuries. In other words…
Let's break it down a bit further….
The Evolution of Information Networks: What We Can Learn from History
Harari argues that human history can be understood as a series of increasingly complex information networks. Each time we developed a new way of storing, transmitting, or controlling information, it led to dramatic shifts in power, economics, and culture.
The Clash of Information Systems: Historically, what we often remember as ideological or political conflicts—capitalism versus communism, democracy versus dictatorship—are actually clashes between different types of information networks. The real struggle was about which system could better harness the flow of information to manage societies and economies.
The Naïve View of Information: The common belief—especially in tech—is that more information equals more power. Silicon Valley has long operated under the assumption that data is inherently good. The more we collect, the more we know, and the more powerful we become. But Harari dismantles this idea. Information, by itself, doesn’t guarantee wisdom or even power. The immediate impact of the invention of the printing press wasn’t the scientific revolution… That would take another century to develop. The immediate impact of that technological revolution was the proliferation of violent conspiracy theories that we remember today as witch hunts.
Mythology and Bureaucracy in Information Networks
Harari's central argument is that large-scale human cooperation—and the power it generates—relies on two competing principles: mythology and bureaucracy. And this tension is as relevant for modern organizations as it was for ancient empires.
Mythology as a Tool for Coordination: Mythology allows humans to create shared realities (what he calls “inter-subjective realities”). Think of money, corporations, or even the concept of “nations”—none of these things exist in the physical world. They’re constructs that only work because we all agree they work. In business, startup founders are often mythmakers. They craft compelling narratives that bind teams together, create investor confidence, and generate early-stage buy-in. But myths alone don’t scale—eventually, you need structure.
Bureaucracy as a Tool for Order: Where mythology creates a sense of purpose, bureaucracy creates stability. It’s the system of lists, records, and processes that allow information to be transferred from one generation to the next. Bureaucracy turns the chaotic myths of a startup into a sustainable enterprise. But Harari points out that bureaucracy often sacrifices truth in favor of order. When you begin to see the world, or your organization, through a bureaucratic lens, you have no choice but to start pushing square pegs into round holes. The world rarely fits neatly into the defined drawers that bureaucracy requires.
For tech leaders, this is a critical lesson: your organization is an information system, and balancing mythology and bureaucracy is one of your most important jobs. Early-stage startups thrive on myth, but as they scale, they need bureaucracy to survive. The trick is not to let either principle dominate.
Balancing Truth and Order: Self-Correcting Mechanisms in the Age of AI
If mythology and bureaucracy are the forces that shape information networks, then self-correcting mechanisms are what keep those networks from collapsing under their own weight.
The Scientific Revolution and Self-Correction: One of the biggest impacts of the scientific revolution was the introduction of self-correcting mechanisms—systems that prioritize truth-seeking even at the expense of order. Peer review, falsifiability, and experimentation allowed scientific knowledge to grow exponentially, but they also created friction. Information systems that embrace conflict, dissent, and iteration tend to outlast information systems that purport to know the “Capital T Truth” at all times.
The Cost of Truth-Seeking: Unfortunately, truth-seeking is messy. It undermines the myths that hold teams together. But without it, information systems become brittle. In most information systems, truth-seeking is not the default state. It’s disruptive. It threatens the shared beliefs and values that hold societies together. But when you start digging for the truth, especially uncomfortable truths, you risk undermining those narratives.
Decision-Making Frameworks: Democracy vs. Totalitarianism in Information Systems
So, how do you balance these competing priorities? How do you decide the right level of myth, order, and truth-seeking in an information system? There are two basic models for managing information networks, and we know them by the names of democracy and totalitarianism.
Totalitarian Systems: These systems are designed to maximize order and minimize dissent. They’re efficient, but they struggle to adapt. In the business world, totalitarian systems look like top-down, rigid hierarchies where decisions flow from the top and are rarely questioned. This might work in a stable environment, but it’s deadly in fast-moving industries like tech.
Democratic Systems: Democratic information networks, by contrast, are chaotic but innovative. They invite multiple perspectives, encourage debate, and tolerate dissent. This makes them less efficient in the short term, but far more adaptable in the long run. This sounds great on paper, but as teams grow and dissent multiplies, managing more democratic information systems requires heavy use of myths and bureaucracy to get teams aligned.
For tech leaders, this means that decision-making frameworks are critical. Do you optimize for order and efficiency, or do you prioritize truth-seeking and innovation? The answer depends on your company’s age, stage, and the industry you're in. But the worst decision is to make no decision at all—to let your information network evolve without intentional design.
Key Takeaways for Tech Leaders Navigating AI and Information Systems
Mythology vs. Bureaucracy
As organizations grow, managing the push and pull between mythology and bureaucracy is one of a CEO’s most important jobs. Your organization is an information system and striking the balance at all ages and stages of your company’s development between mythology and bureaucracy is a big responsibility.
In the early stages of a startup, mythology is essential to getting teams to buy-in, but as you grow, shifting more energy into building the bureaucratic frameworks that perpetuate the myth becomes paramount. While this insight may not be groundbreaking for any leader who has led through the “scale-up” stage of a company, this mental model provides a strong anchor for leaders as they’re navigating inevitably choppy waters.
Truth vs. Order
Defining the balance between truth and order in your information system will be an organizational imperative as you begin to adopt AI into your standard operating procedures.
When AI is merely bolted onto your existing human information systems (e.g. a salesperson leveraging ChatGPT to customize a product demo script for a prospect's needs, a marketer using Jasper to write a press release, etc.) you can get away without giving this critical thought, but a lot of the value unlocked by AI doesn't come from AI being leveraged as a peripheral node to your information network. Most of the value will come from integrating AI into your standard operating procedures and allowing computer-to-human and computer-to-computer networks optimize work.As your organization shifts from AI as a siloed node on your information into a central feature of it, this consideration becomes more important.
This requires intentionality because there's no one-size-fits-all approach. An early-stage startup looking for product market fit or scaling marketing team searching for new revenue drivers may have a strong bias for truth seeking whereas a publicly traded company or operations team may have a strong bias for order. Both come with tradeoffs and negative externalities. Truth-seeking organizations will have a more representative view of their product and market landscape, but will have to contend with a more splintered team that could fracture as company myths are busted. Order-seeking organizations will thrive on operational excellence at the expense of putting blinders on that could help them foresee pitfalls. As with most business decisions, it's on leadership to define which tradeoffs are better for your age and stage of company and what steps you're willing to take to mitigate the risks.
Self-Correcting Mechanisms
To become a more truth-seeking organization, we must incorporate self-correcting mechanisms into our information system.
Management frameworks like Agile do a great job of this by incorporating retrospectives and lessons learned as a key output of the process, but these management structures are typically isolated to specific teams or departments. At the organizational level, your options are to build your own self-correcting mechanisms (e.g. - the CEO who seeks and solicits dissenting opinion and data) or have them built for you (e.g. - The board who fires the CEO).
These mechanisms improve probability of success in many ways, but come with the tradeoff of dissent and discord. The organization whose information systems will scale most effectively in an AI-enabled era will be the ones able to balance self-correcting mechanisms that seek truth from the broad array of data around their company while maintaining order through mythmaking and bureaucracy.
Five Reflection Questions for Leaders
Who on your team is a mythmaker, and who is a bureaucrat? Do you have the right mix of these skillsets for your organization’s current stage?
Is your organization more focused on truth-seeking or maintaining order? Is this the right balance for your current goals?
What self-correcting mechanisms do you have in place? Are they robust enough to overcome the stasis brought about by mythology and bureaucracy?
How are you managing dissent within your organization? Are you actively seeking out conflicting opinions, or are you letting your information network become an echo chamber?
How are you structuring the data within your organization in order to enable AI to become a core part of your decision-making processes moving forward? Are you designing your organization in a way that enables AI to become a centralized force in your information network, or are you bolting onto informational nodes for one-off use cases?
If You Enjoyed This Book, You May Also Like:
The Chaos Machine: A deep dive into how technology is transforming our collective consciousness.
Organizational Physics: A practical guide to scaling organizations using the principles of natural systems.
Thinking in Systems: A primer on systems thinking, essential for understanding complex organizations in the age of AI.
In closing, technology isn’t destiny. How we design our information networks—and the systems that manage them—will determine the future. While there’s a lot that we don’t know about where the next decade of technology is going, there’s one thing we know for certain. We’re the humans and we’re in charge… for now.
I want Yuval Noah Harari to have the final word.
“The decisions we all make in the coming years will determine whether summoning this alien intelligence proves to be a terminal error or the beginning of a hopeful new chapter in the evolution of life.”