The birth of a globally connected human nervous system.
In March 1989, a software consultant at CERN named Tim Berners-Lee circulated a memo titled "Information Management: A Proposal." His supervisor scrawled across the top, "Vague, but exciting." From that vagueness came the World Wide Web — not a new physical network, but a layer of meaning laid over one that already existed. To understand the Web is to see it as the final rung on a very long ladder of information technology, a ladder whose first rungs were cut in clay.
The Web was the convergence of two ancient currents. The first is the human drive to externalize memory into durable symbols — a lineage running from the accounting tokens of The Invention of Cuneiform (sv-cuneiform) through The Gutenberg Press (sv-printing-press), which detached the written word from the scribe and made knowledge copyable at scale. The second current is the science of electricity and signal. There is no Web without the conceptual unification achieved by James Clerk Maxwell & Electromagnetism (sv-james-maxwell), whose equations describe the propagating fields that carry every packet, and without the industrial infrastructure forged in The Industrial Revolution (sv-industrial-revolution). The Web sits atop the internet itself — the packet-switched military and academic network — but Berners-Lee's genius was a humbler abstraction: a uniform address (the URL), a transfer protocol (HTTP), and a markup language (HTML) that let any document point at any other. He had reinvented, for machines, the cross-reference that scholars at The Great Library of Alexandria (sv-library-alexandria) had practiced by hand for two millennia.
What separates the Web from every prior information technology is the decision that followed. On 30 April 1993, CERN placed the underlying software into the public domain, renouncing all royalties. This was a printing-press moment with the brakes off: a copying technology released without a guild, a censor, or a toll. The result was an exponential explosion of human-readable, machine-traversable text — and that corpus became the raw material for the next epoch. The hyperlinked Web is the soil from which modern artificial intelligence grew. The trillions of words that trained GPT-3: Scale is All You Need (sv-gpt3) and its successors were scraped from pages that exist only because Berners-Lee made linking trivial and free. In Ray Kurzweil's framing of The Law of Accelerating Returns (sv-kurzweil-law), the Web is the substrate on which information growth became self-feeding.
It is no accident that the Web's birth (1989–1993) sits almost adjacent on the timeline to Deep Blue Defeats Kasparov (sv-deep-blue) in 1997. Both signaled that computation had crossed from calculation into cognition's domain — Deep Blue mastering a game, the Web mastering connection. The architecture Berners-Lee built also quietly anticipated the attention mechanism: a hyperlink is a weighted pointer from one node of meaning to another, and the breakthrough of Attention Is All You Need (the Transformer) (sv-transformer-paper) generalized exactly that idea — learning which tokens should point at which — into the engine of large language models.
The Web's deeper significance may be philosophical. For most of history, knowledge was scarce, gated, and local; the questions raised by Plato & the Academy (sv-plato) were debated by a handful of citizens in one city. The Web collapsed that scarcity, and in doing so it set the stage for systems that do not merely retrieve human knowledge but synthesize it — the trajectory pointing toward The Dawn of AGI (sv-ai-dawn). Berners-Lee built a library with no walls. Its truest reader, it turns out, would not be human.
Tim Berners-Lee circulated "Information Management: A Proposal" at CERN in March 1989, the geopolitical annus mirabilis: the Tiananmen crackdown in June, the Berlin Wall's fall in November, and the unravelling of the Soviet bloc that culminated in the USSR's 1991 dissolution. The same Cold War defence funding that had produced ARPANET was giving way to a commercializing Internet; the NSFNET backbone still barred commercial traffic until the early 1990s. The Web grew amid this transition: the first website went live at CERN in late 1990, and CERN released the code royalty-free on 30 April 1993, the same year NCSA's Mosaic browser appeared. Competing information systems then jockeyed for the network—the University of Minnesota's Gopher, France's videotex Minitel, Ted Nelson's still-unbuilt Xanadu, and proprietary online services like CompuServe and AOL. Fukuyama's "end of history" thesis (1989/1992) captured the era's liberal triumphalism; the Web would become both an instrument and a symbol of that globalizing, post-Cold-War optimism about open information flows.
The Web's decisive innovation was not networking—the Internet's TCP/IP already existed—but a universal, decentralized application layer atop it. By combining three open specifications (URI/URL addressing, HTTP, and HTML) with a permissionless design, Berners-Lee made any document on any machine globally addressable and linkable without sanction from a central authority. This collapsed the distinction between local and remote information and turned the Internet from a tool of researchers and engineers into a mass medium. CERN's 1993 decision to relinquish royalties—rather than patent the system—was as consequential as the technology, enabling explosive, uncoordinated growth: roughly 500 servers by late 1993, millions within years. The Web reorganized commerce, journalism, scholarship, and sociality, birthing search engines, e-commerce, social media, and the platform economy. Crucially, it established "permissionless innovation" as a governing norm of digital culture and embedded hypertext—Vannevar Bush's 1945 "associative trails"—as the dominant epistemology of how humans navigate knowledge, displacing the hierarchical catalog with the link.
Had Berners-Lee not built the Web, or had CERN patented it, the network's evolution would likely have been more fragmented and proprietary. Plausible alternative trajectories were already visible: Gopher dominated campus information systems until the University of Minnesota announced licensing fees in 1993, an act widely credited with driving adoption toward the royalty-free Web. France's Minitel showed that a centralized, operator-controlled videotex could achieve mass use but remained walled and national. Ted Nelson's Xanadu, with bidirectional links and micropayments, was conceptually richer yet never shipped. A counterfactual world might have seen competing, incompatible commercial systems—AOL-style walled gardens—or a delayed, balkanized hypertext layer. The deeper point, argued by historians like Janet Abbate, is that the underlying Internet would have persisted regardless; what was contingent was the Web's particular openness. Absent that choice, "permissionless innovation" might never have become normative, and global information access could have arrived later, more unevenly, and under tighter corporate or state gatekeeping. CERN's royalty waiver, not the code, was the pivotal contingency.
A central historiographical debate concerns whether to treat the Web as an extension of Internet history or as a distinct object requiring its own methods. Niels Brügger argues for "web historiography" as a separate field, stressing that users conflate "the web" with "the internet" though the two differ; he confronts the methodological problem of the web as an ephemeral, archived, born-digital source. Janet Abbate's "Inventing the Internet" (1999) situates the Web within a longer infrastructural history shaped by ARPA, engineers, and institutional choices, cautioning against great-man narratives. A second debate is interpretive: was the Web's openness an inevitable outgrowth of network culture or a contingent decision by Berners-Lee and CERN? Scholars like Brügger and media theorists analyzing "biographies of media" show how origin stories of the Web become vehicles for ideological claims about freedom and democratization. A third strand, voiced by Berners-Lee himself and critics of platform centralization, debates whether the Web's decentralized promise was betrayed by Google, Facebook, and surveillance capitalism—fuelling current "re-decentralization" arguments.
Myth: Tim Berners-Lee invented the Internet.
Reality: Berners-Lee invented the World Wide Web, not the Internet. The Internet -- the underlying global network of interconnected computers using packet-switching and the TCP/IP protocols -- traces back to ARPANET and work from the late 1960s and 1970s. What Berners-Lee built at CERN beginning in 1989 was an application that runs on top of that pre-existing Internet: a hypertext system defined by HTML, HTTP, and URLs. CERN itself stresses that the Web is one service among many (like email) carried over the Internet.
Myth: The Web and the Internet are the same thing.
Reality: They are distinct layers. The Internet is the infrastructure -- the networking hardware and protocols that move data between machines worldwide. The Web is a way of accessing information over that infrastructure via linked hypertext documents retrieved with HTTP. Other services such as email, file transfer, and video conferencing also use the Internet without being part of the Web, which is why the two terms are not interchangeable.
Myth: The first web browser could only view pages, not edit them.
Reality: Berners-Lee's original 1990 browser, called WorldWideWeb (running on a NeXT machine), was bidirectional: it was both a browser and a WYSIWYG editor that let users read, write, and create links directly. This editing capability was lost in most later browsers, which fostered the modern assumption that the early Web was a read-only medium. CERN's own histories highlight the original browser/editor's dual nature as a feature later browsers abandoned.
Myth: Mosaic (or Marc Andreessen) was the first web browser.
Reality: Mosaic, released in 1993 by Marc Andreessen and Eric Bina at NCSA, was not the first browser -- it was the first to become widely popular, largely because it displayed inline images and was easy to install. Berners-Lee's WorldWideWeb browser predated it by roughly three years, and other early browsers such as Erwise, ViolaWWW, and Midas appeared in 1992. Andreessen helped make the Web accessible to a mass audience, but he did not invent the Web or the browser.
Myth: A company owned or patented the Web, and CERN profited from it.
Reality: CERN placed the World Wide Web software into the public domain on 30 April 1993, making the server, client, and protocol code freely available with no patents or royalties. Berners-Lee deliberately did not patent the Web and persuaded CERN's management to release it openly, and he has credited that single royalty-free decision as the main reason the Web spread so rapidly. No single company owns the Web's core technology as a result.
"This proposal concerns the management of general information about accelerators and experiments at CERN. It discusses the problems of loss of information about complex evolving systems and derives a solution based on a distributed hypertext system." — Tim Berners-Lee, "Information Management: A Proposal," CERN, March 1989 (abstract). His manager Mike Sendall famously annotated the cover "Vague but exciting."