There is a particular kind of reading experience that feels less like consuming a book and more like walking alongside someone who helped shape the world you grew up in. Tim Berners-Lee’s memoir This Is for Everyone: The Unfinished Story of the World Wide Web, written with journalist Stephen Witt and published by Macmillan in September 2025, is that kind of book. It is part autobiography, part technical history, part manifesto and for those of us who came of age alongside the web itself, it reads as something far more personal.
I read it with a particular kind of recognition: not the recognition of a reader encountering a familiar argument, but the recognition of someone who was there, at the edges of these same moments, fumbling with the same technologies, lit up by the same ideas. Tim’s story is, of course, far larger than mine. But the web was never the story of one person. It was always a shared journey, and part of what makes This Is for Everyone so quietly remarkable is that Berners-Lee seems to understand that, even as he writes about his own extraordinary life.
Before the web: connections already being made.
Tim Berners-Lee opens his memoir with a charming portrait of his origins, born in 1955 into a household where computing was already in the air. His parents, both mathematicians who worked on the Ferranti Mark 1, Britain’s first commercial computer, gave him what he describes as a childhood soaked in logical thinking, gentle eccentricity, and what his mother called “watchful negligence”. The detail that his mother could hold a five-hole paper punch tape up to the light and read the binary code is not just a lovely anecdote; it is the foundation of a book that argues, over and over, that the people who build technology matter as much as the technology itself.
I think about my own entry point into this world, which predates the web by a few years and is considerably less glamorous than Tim’s. In 1989, working in my first proper job for Autodesk Australia, I was ordering CAD books from Ventura Press in California via something we called email, though calling it that feels generous. It was a dial-up system, a command-line interface, a terminal editor, plain-text composed and sent in batches when the modem connected. There was no browser, no interface, no graphic of any kind, just a blinking cursor and the peculiar thrill of knowing that a message was travelling somewhere impossibly far away where I had never been. Tim hadn’t invented the web yet. That written proposal was still sitting on his desk at CERN, where his boss Mike Sendall famously scrawled “Vague but exciting” across the cover page.
What strikes me now, reading Tim’s account of those years at CERN in the late 1980s, is how ordinary the context was for something so extraordinary. He wasn’t at Stanford or MIT. He wasn’t flush with venture capital. He was a programmer helping the world’s largest particle physics laboratory keep track of its own complexity, and the solution he dreamed up was to combine two existing technologies, the internet and hypertext, into a single navigable system. The internet was already there. Hypertext was already there. The leap was the combination. And in 1989, while I was typing purchase orders into a command line in Melbourne, he was sketching the architecture of something that would change everything.
I was at Autodesk’s UK office in 1990 when a group of Russian visitors came through the building. It didn’t feel random. Autodesk had already been pushing into that market, having run what was promoted as the first software-focused trade show in the Soviet Union, organised by a Western firm in Moscow in 1988. When the Russians came, it felt like business mixed with history, doors opening, connections being made in the slipstream of the Cold War’s dissolution. The year before, the Berlin Wall had come down. Berners-Lee submitted his first web proposal in March 1989. These were not coincidences of timing so much as expressions of the same underlying impulse: the conviction that information, like people, should be free to move.
Gopher, Mosaic, and the first moments of the web
Tim’s account of the web’s early spread from a single NeXT computer in a small room at CERN to a rapidly expanding network of servers is the most technically detailed part of This Is for Everyone, and, in some ways, the most interesting. He describes building the first web page, the first browser, and the first web server, all on the same machine, in what he clearly remembers as a period of almost feverish creative energy. The decision to release the source code freely, without patent or licence, is treated here not as altruism but as a logical extension of the web’s design philosophy: if the point is universal connectivity, then restricting access at the point of creation would be absurd.
The protocol that perhaps best captures the pre-web internet, and which Berners-Lee mentions as part of the landscape the web eventually overtook, is Gopher. Developed in September 1991 at the University of Minnesota by Mark McCahill and his team, Gopher was a menu-driven system for navigating and retrieving documents across the Internet. It was structured, hierarchical, and genuinely useful, a real internet service that predated the web’s public availability.
In 1992 and 1993, when I was on an exchange from La Trobe University in Melbourne to UC Santa Cruz in Northern California, the University of California’s Gopher system was the primary way I navigated the internet. I also received my first email address, something like craig@ucsc.edu, though memory is an unreliable archivist. These were the months, I now realise, when Berners-Lee’s invention was just beginning to move out of CERN and into the world. Marc Andreessen, then a student at the National Centre for Supercomputing Applications in Illinois, was working on what would become Mosaic, the first widely used graphical web browser, released in 1993. I was using Gopher menus to look up academic resources in the UC library system, and within two years, like many others, I would be building my own website.
This is one of the things that This Is for Everyone does well: it helps you triangulate your own experience against the broader history. The web wasn’t announced and adopted; it spread in the way that good ideas sometimes do, unevenly and through enthusiasts and accidents, until one day it was simply everywhere. Tim is honest about this, and his account of the early browser wars, Mosaic giving way to Netscape, Netscape being devoured by Microsoft’s Internet Explorer bundled free with Windows, is told without triumphalism. It’s a history full of choices that seemed reasonable at the time and later revealed their consequences, many unintended.
One important figure who appears in Berners-Lee’s book was a key part of my earlier education and later Master’s and PhD work. Ted Nelson, the visionary who coined the term “hypertext” in 1965, had been developing Project Xanadu, an ambitious global networked publishing system in which documents could be linked, quoted, and reused with full attribution, for decades before the web existed.
In April 1988, my employer, Autodesk, announced that it would acquire an 80% interest in the Xanadu Operating Company and named Nelson a “Distinguished Fellow”. Nelson’s own account describes his role as largely evangelising the project rather than running its engineering effort day to day, and Autodesk eventually relinquished its interest after the hoped-for product never materialised. That Autodesk, the company I was working for and ordering CAD books into the ether, was, for a brief time, a key chapter in the Xanadu and hypertext story, a coincidence I didn’t fully appreciate until much later. My friend Sam, who was also at Autodesk, remembers going to Ted Nelson’s birthday party on his houseboat in Sausalito around 1992 or 1993, though, as Sam would be the first to admit, recollections from that particular beery era should be treated with appropriate scepticism.
Nelson’s paper, published in 1965 as “Complex Information Processing: A File Structure for the Complex, the Changing and the Indeterminate,” argued that computers should allow people to connect ideas across documents rather than keep information trapped in separate files, a vision that was technically ahead of its time and perhaps ahead of any time! https://www.craigbellamy.net/2007/03/16/ted-nelson-1965-complex-information-processing-a-file-structure-for-the-complex-the-changing-and-the-indeterminate/. What Berners-Lee built was simpler, more pragmatic, and less philosophically complete than what Nelson imagined. The web’s one-way links, which Nelson considered a compromise, turned out to be the very feature that allowed it to scale globally. And Nelson’s insistence that technology should serve human authorship, not corporate convenience, echoes loudly through the later chapters of This Is for Everyone.
Building on the web: hypertext, history, and first experiments
By 1995, back in Australia after returning from Santa Cruz, I was living in Sydney and studying at UNSW for my BA Honours year in History. The thesis was a traditional narrative history, “The Question of Hajime,” about a Japanese fighter-bomber pilot who was shot down during the 1942 bombing of Darwin and captured on Melville Island by local Aboriginal Australians. It was analogue scholarship in every formal sense: footnotes, archival research, and a bound copy submitted through the usual channels. But it was also during this year that I built my first website, a vivid, exuberant thing with four galleries, bright pastel colours, and animated banners, viewable in the Internet Archive. And the first thing I published on it was, of course, my Honours thesis.
This is almost a precise parallel to what Berners-Lee describes as the web’s foundational move: taking documents that existed in one form and making them accessible, navigable, and shareable. The web’s earliest uses, as he notes, were academic, researchers sharing preprints, physicists linking technical documents, and my own first act of publication was entirely consistent with that impulse. We were all, in our different ways, trying to solve the same problem: how to get information out of the locked box of a single document and into the open.
It was also during this period that I encountered hypertext theory as a formal discipline. When I enrolled in a master’s by research at the University of Melbourne in 1996, the thesis was titled The Author of History in the Age of Electronic Reproduction: Hypertext and the Historian. It was very early days indeed, both for the web and for critical discussions about it https://www.craigbellamy.net/2012/03/16/ma/. The thesis asked a question that feels like deja vu in the present AI-saturated education landscape: “This thesis seeks to define hypertext history authorship, discuss how it differs from a book, and, in doing so, hopefully reveal some best practices. Will hypertext produce simplistic catalogues of empirical facts or uninterpreted primary sources, or will hypertext, with its combination of image, text and sound, offer the historian fresh scope for authorship?” In 1996, this was genuinely uncharted territory. There was almost no existing scholarship on what it meant to write history as hypertext, rather than about hypertext. The thesis forced a confrontation between the humanist tradition, narrative, interpretive, author-driven, and a technology whose architecture seemed to imply that the reader, not the writer, should determine the path through a document (there was no Wikipedia at this stage, perhaps the greatest example of hypertext).
Berners-Lee is too modest, or perhaps too technical, to spend much time in This Is for Everyone on the philosophical dimensions of what hypertext does to the act of reading and writing. But his discussion of the web’s design as fundamentally non-hierarchical, anyone can link to anything, there is no central authority determining what connects to what, carries these implications. The web wasn’t just a technology. It was an epistemological proposition. My thesis, grinding through the awkward gap between history as a discipline and the web as a medium, was trying to work out what that proposition meant in practice. The answer took years and didn’t really answer it definitively.
After graduating from Melbourne in 1998, I did what seemed logical given the circling tension between writing about hypertext and actually making it: I went to art school at RMIT, to the wonderful Animation and Interactive Media programme, a centre then at the genuine forefront of multimedia practice. These were the heady dot-com years, and the focus was messy. My PhD by project, eventually completed in 2002, was www.mikbar.com.au: The Everyday City and Globalisation https://www.craigbellamy.net/2012/11/24/milkbar/. It took a long time to find a topic that worked hypertextually, a structure that justified and exploited the form rather than simply transferring a linear argument into a networked skin. I learnt hard lessons about scope and other things, but I got through it, and it fed directly into later work in the digital humanities that, in the spirit of Ted Nelson, if not always his method, tries to keep theory and practice in conversation with each other.
Berners-Lee was, by this point, at MIT, steering the W3C, the World Wide Web Consortium, through the treacherous politics of making international technical standards in an area moving faster than any standards body had previously had to manage. He describes the browser wars, the HTML debates, and the fracturing of W3C with the clarity of someone who found the process both fascinating and exhausting. The dot-com boom was peaking, then collapsing. The web’s architectural idealism was already in tension with the commercial imperatives of the companies skimming the top of it.
Growing pains: when the web turned
The section of This Is for Everyone that I found most compelling and most troubling is the extended account of how the web’s original vision was eroded. Berners-Lee is not naive about this, and he is not letting himself off the hook either. He describes with genuine candour the moment he realised that social media platforms were not simply building on the web but actively deforming it: harvesting user data without meaningful consent, designing for compulsive engagement rather than genuine utility, and concentrating power in the hands of a handful of corporations in ways that directly contradicted the decentralised architecture he had designed in 1989.
His account of meeting Mark Zuckerberg at his home because the Facebook founder had the flu, and of being impressed by his intelligence and ambition, is honest in its ambivalence. He found Facebook useful, then watched it become something else. The Arab Spring appeared to demonstrate the emancipatory potential of social media. The years since have demonstrated the opposite with equal force. Berners-Lee is clear that the problem is not the technology itself, but the incentive structures built on top of it: the attention economy, in which the product being sold is not a service but the user’s own cognitive resources, perhaps a lesson for AI in education.
This is for Everyone is more engaging than typical technical histories because Berners-Lee highlights how social conditions shape the use of technology. He shows how the web’s open access model resulted from significant political negotiation, in which he has been actively involved for thirty years. The book’s distinctiveness lies in its ‘socio-technical solutions’, which continue with Berners-Lee’s Solid project and the Social Linked Data protocol, aiming to return data ownership to individuals through user-controlled “pods” rather than corporate silos. The goal is to create an ‘intention economy’ rather than an ‘attention economy’ that prioritises user goals over advertisers’ goals.
Beyond boomers and doomers
I am now, as Tim Berners-Lee is, up to my ears in artificial intelligence. The specific context is education, a sector whose relationship with AI is currently somewhere between bewildered and evangelical, depending on who you ask. I work at a university that has positioned itself explicitly as an “AI first” institution, and there is something in the rhetoric that feels, to anyone who was present at the birth of earlier technological revolutions, uncomfortably familiar: the transcendent narratives, the language of inevitability, the confident assertion that transformation is coming and that the only question is whether you will be on the right side of it.
Berners-Lee is characteristically rigorous about this in This Is for Everyone, dismissing the boomer/doomer binary as unhelpful and media-driven. His argument is that the question “Is AI good or bad?” is no more answerable than “Is regulation good or bad?” The answer depends entirely on what you are regulating, and for whose benefit. What matters is not the technology, but the design choices embedded in it: choices about who controls the data AI is trained on, whose goals the AI is optimised toward, and what accountability structures exist when those systems cause harm.
AI is harder to navigate than hypertext was, and hypertext took a generation. When I was writing a master’s thesis in 1996 on what it meant to author history in hypertext rather than about it, the questions felt urgent but also manageable, bounded by the relatively slow pace at which the technology was entering institutions and the quality research about it. AI is entering at a far less optimistic historical moment than the post-Cold War opening that provided the web’s original cultural context. The sense of expanding possibility that accompanied the fall of the Berlin Wall, that same sense of doors opening that I felt watching Russian sales engineers walk through the Autodesk offices in 1990, has been replaced by something more anxious and more contested.
And yet. Berners-Lee’s own closing note in This Is for Everyone is genuine optimism, not the techno-utopian kind, but the kind that comes from having been right before, under difficult circumstances. He invented the web when no one was asking for it, in an institution whose mission was particle physics, with no venture capital and no track record. He gave it away, not because he was indifferent to its value, but because he understood that its value depended on its universality. Thirty-five years later, he is doing something similar with Solid: building an alternative architecture patiently, collaboratively, in the conviction that getting the design right now, at the beginning of the AI era, matters more than any short-term commercial advantage. “We can still build the future we want,” he writes. “There’s still time to build machines that serve the human, rather than the other way around”.
That is not naivety. It is the position of someone who has seen what determined people can do when they believe in the principles they are working from. For those of us who grew up alongside this story and remember the dial-up squeal, the Gopher menus, and the thrill of publishing something, anything, to the open web for the first time, This Is for Everyone is a historical partner. It tells you where you were, and helps you understand, with more clarity than you probably had at the time, why it mattered. The hard work, as Tim knew in 1989 and as I am relearning now in the context of AI, is never in the dreaming. It is always in the application. But the dreaming matters too. The dream must be right before the work can begin. Tim Berners-Lee got the dream right. This book explains how, with data, passion, and the authority of someone who has been thinking about this for longer than most of us, we can still build our way back to it.
This Is for Everyone: The Unfinished Story of the World Wide Web by Tim Berners-Lee, with Stephen Witt. Macmillan, September 2025.
This review was prepared with the help of Claude Sonnet 4.6
Leave a Reply