Alex Asseily | Founder, Zulu Group
Cumulative knowledge accessible in 1hr or less
Introduction
20th-century media—telephone, radio, fax, television—shrunk distances and amplified reach but largely extended the print model: one-to-many distribution at scale. Their defining feature was concentrated power. A handful of institutions—such as the BBC World Service and The New York Times—shaped global narratives (often with imperial bias) through systems that allowed information to flow outward but rarely back in.
This imbalance breeds propaganda and weakens error-correction: narrow interests override truth. Mass media can be weaponized, with voices dehumanizing populations and conditioning acceptance of atrocities—as seen in Nazi Germany, Rwanda’s genocide, and modern conflicts like Gaza, Myanmar, and Sudan.
Anyone in Gen-X, like myself, has now lived through a series of equally seismic shocks — new technologies that don’t just boost the speed of communication but transform knowledge exchange altogether.
These changes are shaking the world today. It feels chaotic because we are living through a non-linear expansion of the human information network — the impact of the printing press compressed into years rather than centuries. Critical thinking is scaling globally, advancing not only better explanations of nature but also revealing better ways of living and collaborating, from health to politics.
As feedback loops accelerate and obvious truths grow harder to ignore, long-standing institutions strain or collapse, unable to adapt quickly enough. This, I believe, is the most challenging aspect of what lies ahead: the tug-of-war between institutional inertia and the momentum of progress. Authority is no longer anchored in slow hierarchies but is constantly tested by the network itself. If it is wise, it will listen.
This article explores the key technological innovations of the ‘connected era’ (1990-present) across network nodes, links, and cognitive capacity that have in turn been driving new emergent properties from the human information network:
1. Instant, global, bi-directional, network links: Internet & web
The Internet and the World Wide Web (together “the Internet”) created the first digital knowledge substrate for the collective brain, linking nodes (computers and documents) at light speed in both directions and thus enabling low-cost information seeking and publishing. This broad access accelerated error-correction and made information sharing—especially rich formats like video and podcasts—vastly more powerful than they had been previously.
2. Billions of people as network nodes: social media
Social media transformed human communication by turning individuals into networked nodes, breaking down traditional media hierarchies and enabling anyone to publish, challenge narratives, and engage in global conversation instantly. While this democratization increased noise, it also diversified perspectives and accelerated error-correction, and—scaled by smartphones—created a planetary forum.
3. Knowledge density and authenticity: video & podcasts
The explosion of online video and later podcasts massively expanded coherent knowledge sharing by increasing both the density of information flow (knowledge per unit of time) and the level of authenticity.
4. De-fragmentation and cognitive leverage: large language models
Cheap, accurate natural-language translation broke down linguistic silos, expanding and diversifying the nodes of the collective brain. LLMs gave people cognitive leverage to digest and synthesize information in minutes—an enormous shift in the capacity of network nodes.
New technologies don’t automatically yield wiser societies. By 1990, humanity’s collective brain had reached the Moon, discovered the structure of our DNA, and invented personal computers. Cultural consciousness lags far behind - while science, technology, and medicine sprint ahead, most societies remain bound to mythic, tribal, or nationalistic paradigms. Institutions rooted in those levels strain against the faster awakening now unfolding among ordinary people.
The Age of Wisdom represents the tipping point when global information networks become dense, fast and coherent enough to lift cultural consciousness itself and pull our institutions forward into greater alignment with our ideals. This article (and Part 4 in due course) aims to trace the ingredients that make this possible.
Firstly, post-industrial networks lacked density, symmetry, and immediacy despite mass distribution. The Internet delivered both—becoming a force multiplier.
Imagine I’m still sitting in that same Stanford library in 1994. Suddenly, the walls dissolve and the building stretches into infinity in all dimensions, filled with every book, scrap of paper, and article ever written. The librarian transforms into a lightning-fast demon, able to summon exactly what you seek—or anything even tangentially related to your request—in the blink of an eye. Even better, this librarian urges you to contribute your own writings, filing them alongside the rest so others can discover them too. You are asked to pay a modest fee for being in the library but nothing more for any of the rich knowledge you discover while there.
Naturally, the world flocks to the boundless library; creators shelve work where it’s most discoverable. Like this magical library, the Internet invited another great leap forward not only because it worked at lightspeed - so did radios and phones - but because it was architected for both seeking information and publishing it at very low cost.
The Internet didn’t just expand access to information—it transformed our relationship to knowledge itself. Publishing became a high-upside, low-friction, endeavour. And by collapsing the time it takes to find things, the Internet gave us more time to think. Search engines became a superpower because they made discoverable in seconds what once would have been buried beyond practical reach: new data.
Similarly, error-correction (discarding a prevailing idea in favour of something new) depends on accessible evidence. Gregor Mendel’s 1866 paper on genetic inheritance was meticulous and complete, yet published in an obscure journal and ignored by the wider scientific community at that time. His data was correct and complete but sat unused until three separate botanists (de Vries, Correns, von Tschermak) stumbled across it around 1900 - 34 years after Mendel’s paper.
History is full of similar examples: progress stalled not for lack of ideas but for want of more perspectives or accessible evidence. Today, the Internet strengthens feedback loops because the evidence to support them is far easier to access—just ask anyone who’s been fact-checked mid-conversation. What once took decades now happens in days or even hours through peer review and online networks.
Secondly, social networks connected billions of faster nodes that expanded diversity of perspectives and super-charged dialogue.
“The more views we get that we can integrate and make coherent, the deeper our understanding of [something] gets.” (David Bohm, 1917-1992)
Social media didn’t invent conversation or socialization—it simply digitized the dynamics of the campfire. It mapped that topology onto the Internet.
Early Internet nodes were machines and web pages; social media made people nodes again. Individuals were no longer limited to publishing articles through formal media channels or managing personal websites. Publishing became persistent, personal, and instant. Much in the same way the printing press drove a surge in literacy in the masses, social media awakened the masses to their own capacities to publish and react, publicly.
This shift dismantled the hierarchical structures of conventional media that had controlled the societal narratives for centuries. Social platforms democratized publishing and distribution: anyone could report, challenge prevailing views, and engage in public conversation. Relevance moved from page rank to social trust. Asynchronous large-scale interaction between people sped up and broadened conversations, and compressed feedback loops further, speeding up error-correction.
The Arab Spring (2010–2012) vividly demonstrated how a diversity of voices made the movement more resilient and ensured official narratives could not dominate, despite crackdowns. Trans-national ‘friends’ meant that these stories were heard by people around the world who’d otherwise have been cut off from life in Egypt or Tunisia.
Noise rose too, as did distortions due to algorithmic polarisation, but a diversity of views ultimately strengthened the network. Critical thinking—about events, policies, and stories—moved from news editors’ offices to the edges of the network itself, where billions of people could question, test, and correct. Smartphones - far easier to use and more versatile than traditional computers - then scaled this to billions of users. The result: a planetary forum at light speed (albeit with limitations).
Between 2005 and 2015, the phenomenon was still in its infancy. Ideas were shared mainly through text and images, while language silos effectively cut off most of us from the vast majority of our fellow humans, limiting exposure to new ideas, cultures, and shared empathy.
Thirdly, video content and podcasts massively increased knowledge density and authenticity, respectively
A picture tells a thousand words—so perhaps a short video tells a million. A video may not match the full theoretical depth of an academic paper, but it excels in speed and efficiency, distilling the essence of an idea in a fraction of the time.
Video today is a mass-distillation medium. In the early days of YouTube (mid-2000s), most video content skewed toward entertainment. By the early 2010s, we began to see the rise of dedicated ‘explainer’ channels like Vsauce, CrashCourse, and Khan Academy, which showed that video could compress and transmit complex ideas just as effectively as traditional learning formats. By 2020, millions of short, highly visual explainers flooded the Internet, covering science, health, culture, history, and everyday skills. In thirty seconds, you can grasp how your gut biome works; in ten hours, you could absorb the arc of modern Japanese history.
This marks the rise of knowledge density: videos have significantly accelerated humanity’s collective rate of knowledge consumption (coherent information per unit of time) and, by extension, information processing at the nodes.
Podcast Interviews: Baring The Soul
As highlighted in The Age of Wisdom Part 1: Time and Space Shifting, writing down language using abstract symbols (logograms or the alphabet) allowed cultural knowledge to effectively travel through space and time. But in flattening living dialogue into text, it stripped away much of the richness that gives humans their authenticity. 3000 years later, the rise of online video podcasts (notably interviews) has restored some of that lost information. By layering back tone, facial expressions, hesitations, silences, and physical presence, they combined the reach of writing with the intimacy of conversation, reconnecting us to the complex person behind the ideas.
“When you consume somebody…for 40, 50 hours, you do see into their soul. I think there’s almost no way to avoid that. And…if they’re not letting you see into their soul, you’ll notice that [too]” (Dave Smith on the Lex Fridman Podcast).
What Smith really means by soul is authenticity, conceptual depth, even a sense of dignity. It’s worth noting that the US Presidential Election of 2020 was the first time all the leading candidates leaned into this format, recognizing its obvious advantages over canned debates in winning over voters. They barely scratched the surface.
Podcasts are forcing a new kind of transparency. Unlike debates or soundbite-driven media, where rhetoric might mask deeper substance, a three-hour conversation leaves little room to hide. Guests are compelled to reveal the coherence—or incoherence—of their thinking, while good hosts ask the questions listeners might pose, drawing out deeper insight.
Visibility alone no longer confers authority; legitimacy must be earned. In US Presidential debates, real-time fact-checking tickers run beneath the candidates’ words, exposing missteps or blatant ‘untruths’ as they happen. On Newsnight (UK) and similar live debate forums, evasive answers are clipped, circulated, and dissected within hours.
Error-correction happens in real time and in public view, through probing hosts and through subjects self-correcting under pressure and the risk of being shamed. Incoherence or lack of philosophical integrity have become a liability, exposing individuals, organizations, and societies that act against their stated values or ideals (e.g. free speech or pro-life).
Knowledge is no longer policed solely by designated authorities but tested continuously at the network’s edges—millions of listeners checking claims and pressing their leaders to “make it make sense.”
Fourth, we’ve modelled natural language almost perfectly, giving us quantum leaps in language translation and information manipulation.
In 1500 AD there were almost ~40 languages and ~100 distinct dialects spoken across Europe. Despite this, new scientific and philosophical ideas spread relatively fast through the region during the second millennium thanks to Latin (the lingua franca left by the Romans). In Europe, it is estimated that 500k > 1m people could read & write Latin in 1500CE. Vestiges of Latin nomenclature persist to this day in medical and scientific literature. During the Islamic Golden Age half a millennium prior, 100k > 500k could read and write Arabic. It turns out that homogenous protocols of communication really matter in the spread of knowledge – and when they do, knowledge flourishes because it is exchanged & processed by more people. Interestingly, the United States has shared a single language since its founding—likely contributing to the speed of its ascendancy as a global power.
Generative transformers, the foundation of today’s large language models (LLMs), were first deployed in natural language translation. In 2017, the “Attention Is All You Need” paper by Vaswani et al. introduced the transformer architecture, which dramatically improved translation benchmarks like English-to-German and English-to-French.
Today, such translation systems are breaking cultural barriers at unprecedented speed. The challenge of cross-lingual communication isn’t just literal translation—it’s capturing the expressions, tone, and context that carry cultural meaning. A foreign leader’s fiery speech, for instance, may sound harsh to outsiders, yet careful translation can reveal measured arguments about peace and cooperation. Misunderstanding often stems less from words than from the cultural systems shaping our perception. By bridging these divides, translation not only enlarges the collective brain but also creates opportunities for mutual understanding and non-zero-sum collaboration.
Recent examples show how transformative this can be. Palestinian voices translated instantly from Arabic now reach the world via social media, conveying fears, hopes, and common humanity long obscured by headlines or political framing. When U.S. TikTok users briefly migrated to China’s RedBook, they encountered modern Chinese life firsthand—its creativity and cultural vibrancy contradicting familiar domestic narratives. In both cases, propaganda collapsed under the weight of direct human exchange. Translation allows stories to bypass existing filters and dissolve prejudices —making it harder to sustain narrow nationalism and tribal propaganda.
LLMs as cognitive super-power
AI is a giant topic and there are countless angles through which to reflect on the subject matter. Read our article ‘A Better Bicycle For The Mind’ for our view. Through the lens of the human information network, AI may bring two potentially enormous implications: (i) searching for and de-fragmenting accumulated (and published) human knowledge on almost any topic far faster than before ; (ii) extra cognitive capacity for each human to interpret inbound information, to discover errors and to offer refinements.
I made the case earlier that a counter-argument is only as strong as the evidence available to support it. A diligent thinker could use an Internet search engine to gather this evidence and construct a well-reasoned case. In practice, however, most people lack the time and training to sift through endless webpages, evaluate credibility, and piece together a coherent view. We are, in effect, asking average people to become researchers.
Large language models, in contrast, function simultaneously as demon-librarians, editors, and researchers—able to synthesize vast sources into a single, coherent articulation of almost any topic. This synthesis is not yet flawless: outputs may be factually wrong, logically inconsistent, or incomplete. But the leap in accessibility and comprehensiveness is undeniable. What once required hours of searching and sorting can now be surfaced in minutes as a narrative with structure and context.
The risks are equally profound: LLMs could be manipulated to privilege one narrative over another, suppress inconvenient truths, or obscure entire historical records. Critical sources that survive in scattered corners of the web could, in theory, be algorithmically buried, and hidden from practical reach. This makes vigilance essential. We must demand maximum transparency in how models are trained, and insist on philosophical independence from political or commercial forces.
Yet, there is also reason for cautious optimism. Lies and distortions are fragile: to endure, a biased model must constantly resist not just counter-narratives, but also the weight of coherence and logical consistency that naturally emerge when information flows freely. The battle between manipulation and truth is not symmetrical. In the long run, truth has structural advantages.
As a result, billions of Internet users - nodes in this human information network - now have not just theoretical access to the accumulated knowledge of humanity, but the capacity to query it in very sophisticated ways within seconds. This upgrades not only the average quality level of people’s writing, but also the quality of the supporting evidence for any argument they wish to make - whether refuting a factual claim (e.g. who really invented optics?), constructing a legal argument (e.g. what are my civil rights when I’m pulled over?), or building a compelling argument against something that bothers us (e.g. why do politicians take money from industry?).
That, in itself, is a gargantuan leap in the processing and sense-making capability of the human information network. As the error-correction feedback loops occur in more places (nodes), so too does the power (and risks) of the collective brain. The challenge now for us is how to harness these powers while remaining vigilant about how they might be manipulated by narrow interests.
Our success as a species has never rested on the raw intelligence of lone minds, but on practices imitated, refined, and passed across generations. Chimpanzees may excel at memory or cognitive speed, but humans surpass them in language, cultural accumulation, and theory of mind. A person cut off from culture is no different from her primate ancestors.
Each culture has built a body of accumulated wisdom woven into tools, institutions, laws, and stories. The greatest leaps in these came when human information networks received major upgrades: the alphabet gave us writing, the printing press put lifetimes of thought into millions of hands, the Internet created a global library at light speed, and now AI extends our ability to refine knowledge with unprecedented nuance. Each step has not only accelerated the spread of ideas but also strengthened the error-correction that lets better ones rise and outdated ones to dissolve.
Today billions are plugged into this planetary brain—searching, sharing, questioning, and correcting in real time. We imitate not only local habits but also practices from people thousands of miles away. Social media democratized publishing, podcasts restored authentic human dialogue, and LLMs now grant individuals the power once reserved for expert researchers. The result is not just more information, but new forms of creativity and collaboration that no single mind—or culture—could achieve alone.
Every individual can now contribute to the expanding puzzle of reality, which tends toward greater coherence over time. Coherent ideas persist because they align with truth and logic; incoherent ones eventually collapse under their own contradictions. With billions of minds connecting at near-instant speed, knowledge compounds, producing creativity and problem-solving beyond the reach of any traditional institution or individual.
But these powers come with responsibility. The same networks that refine wisdom can also entrench bias; algorithms that illuminate can also obscure; narrow interests are usually reluctant to relinquish their power.
Our task - as innovators and citizens - is to promote systems that privilege coherence, freedom of expression and access, transparency, and truth—strengthening our literacy, the density of our connections, and the feedback loops that let human knowledge leap forward.
As Joseph Henrich notes: “larger and more interconnected groups generate richer repertoires because they preserve and improve innovations more effectively”. Thus, what drives cultural progress is not the raw intelligence of a few disparate individuals, but the sheer capabilities of the information networks we build. Network scale, information density, link speed, and nodes’ cognitive capability, set the ‘thinking capacity’ of the collective brain. The faster and wider ideas travel, the sooner they can be tested, refined, and built upon.
If we succeed, we may witness the dawn of a collective intelligence (human and synthetic) that understands itself and nature more deeply—and acts with greater wisdom. Humanity’s greatest leaps have always emerged from minds working together. This time, we may be building a truly global brain: faster, sharper, more diverse, and, if we choose, more humane than anything before.
In Part 4, we will explore the arc of progress in human information networks, their intersection with consciousness and political philosophy, and the ingredients essential to keeping them healthy.
The Zulu take
Kirk’s assassination.
Beyond the loss of another innocent life, is losing someone seeking open dialogue with others; and despite often divisive political theatre, a willingness to concede his mistakes. His acolytes would do well to take note.
Muskian interference.
Should we stop business leaders from riling up nationalism abroad? No — the answer isn’t to restrict free speech, it’s better arguments and stronger truth-seeking systems.
Robots among us
Finding a humanoid robot in your hallway at 2am may stall adoption - let’s start with dish-washing and laundry instead?
Pakistan’s floods.
After Kashmir’s 2005 earthquake, US military used Chinooks to deliver aid to remote survivors, endearing itself to those it saved. As extreme weather events rise in frequency around the world, would be a good time to make that sort of intervention a habit.
— David Bohm