The 1989 CERN proposal that accidentally built the internet age

In 1989, Tim Berners-Lee wrote a proposal at CERN to solve a problem that sounded painfully ordinary: scientists could not easily find and share information across different computer systems. The lab was full of brilliant people, powerful machines, research papers, project notes, software manuals and technical records, but much of it was scattered across incompatible systems. Finding the right document could be harder than understanding the physics it described. From that frustration came the seed of the World Wide Web — the system that now carries news, banking, shopping, streaming, education, social media, memes, conspiracy theories, family photo albums and the occasional desperate search for “why is my printer making that noise?” This is the story of how a practical workplace fix became one of the most important inventions in modern computing.

A problem only CERN could make this big

CERN in the late 1980s was not short of intelligence. It had physicists, engineers, programmers and visiting researchers from around the world. What it lacked was a simple way for all those people to share information. Documents lived on different machines. Systems did not always talk to one another. Teams changed constantly. Researchers arrived, left and took their working knowledge with them. A useful file might exist somewhere, but only if you knew the right person, the right computer, the right command or the right corridor to walk down.

In other words, CERN had invented a very advanced version of “does anyone know where that document is?” Tim Berners-Lee, then a British software engineer at the lab, saw the problem clearly. CERN did not need one giant database controlled from the centre. It needed a way to connect information across different systems, so that one piece of knowledge could lead naturally to another.

“Vague, but exciting”

In March 1989, Berners-Lee submitted a document called Information Management: A Proposal. It suggested using hypertext to link information together, allowing users to jump from one document to another through clickable connections. His manager, Mike Sendall, wrote a now-famous note on the proposal: “Vague, but exciting.”

It remains one of the greatest understatements in technology history. Most of us would be thrilled if our annual review accidentally predicted the future of civilisation. The idea was simple but powerful: information should not sit alone in isolated files. A document about a project could link to the people involved, the software used, the technical notes, the background research and anything else relevant. Today, that feels obvious. In 1989, it was quietly revolutionary.

The internet before the web

It is important to separate the internet from the web. The internet already existed before Berners-Lee’s proposal. It connected computers and networks. But for most ordinary users, it was technical, fragmented and not especially friendly. It was powerful, certainly, but so is a jet engine, and most people do not want one sitting on their desk.

The web gave the internet a usable face. Instead of typing complex commands or knowing exactly where something was stored, users could click links. Pages could point to other pages. Information became something you could browse. That single change helped move networked computing from the world of specialists into everyday life. The web did not make the internet possible. It made the internet understandable.

The first web server had one job: stay on

By the end of 1990, Berners-Lee had built the first web browser and the first web server on a NeXT computer at CERN. That machine reportedly carried a warning label telling people not to turn it off. This may be the most relatable detail in the entire origin story. The future of global communication depended, at least briefly, on nobody unplugging the wrong box.

From that small start came the basic building blocks of the web: addresses, browsers, servers, pages and links. These were not glamorous ideas on their own, but together they formed a system that could grow almost without limit. And grow it did.

The decision that changed everything

The web might have remained a useful CERN tool, or a specialist academic system, if one crucial decision had gone differently. In 1993, CERN made the World Wide Web technology freely available. That openness mattered enormously. It meant developers, universities, companies, hobbyists and institutions could build on it without paying licence fees or asking permission.

This was the web’s superpower. No single company owned it. No central authority had to approve every page. Anyone with the right tools could publish. Anyone could link. Anyone could build. Early websites were often clumsy, slow and visually alarming, with grey backgrounds, blue links, blinking text and design choices that probably should have required adult supervision. But they were alive. They were open. They were connected.

From research tool to daily life

The World Wide Web began as a way to help scientists share information. It became the platform for almost everything. News moved online. Shops moved online. Banks moved online. Music, video, maps, education, software, job hunting, dating, gaming, government services and personal publishing all found a home on the web.

The result was one of the fastest cultural and technological transformations in modern history. The web changed not just how people found information, but how they behaved. We learned to search before asking. We learned to compare prices instantly. We learned to publish our opinions to the world, sometimes before checking whether we had any. This was the web’s great promise and its great problem: everyone could speak.

The web gets complicated

The early web was open and messy. The modern web is powerful, polished and far more controlled. Much of today’s online life happens inside large platforms. Search engines, social networks, app stores, video sites and AI tools shape what people see and how they find it. The humble hyperlink still exists, but it often competes with feeds, rankings, recommendations and automated summaries.

A web page used to be a destination. Now it can be content for an algorithm, training material for an AI model, or a preview inside someone else’s platform. That shift has created real tension. Publishers worry about losing readers. Users worry about privacy. Developers worry about compatibility. Regulators worry about monopoly power. Everyone worries about misinformation, except the people producing it, who seem oddly energetic.

AI and the next web

Artificial intelligence is now one of the biggest forces reshaping the web. AI tools can summarize articles, write code, generate images, translate pages, answer questions and help people navigate huge amounts of information. Used well, they can make the web more accessible and useful. Used badly, they can flood it with low-quality content, blur the line between human and machine writing, and make it harder for original sources to survive.

This creates a difficult question: if AI systems read, summarize and repackage the web, what happens to the websites, journalists, researchers, artists and developers who created that information in the first place? The web was built on linking. AI often prefers answering. That may sound like a small difference, but for publishers and creators, it could be the difference between being visited and being invisible.

The return of an old question: who owns the web?

Berners-Lee has spent much of his later career warning that the web has drifted away from its original ideals. The early web was decentralized. The modern web is often concentrated in the hands of a few powerful companies. This is why projects focused on data ownership and decentralization matter. Berners-Lee’s Solid project, for example, imagines a system where users control their own personal data and decide which apps can access it.

It is an attempt to return power to the individual user. That may sound idealistic, but then again, so did the web in 1989. The difference is that this time, the problem is not connecting information. It is controlling what happens to it after it is connected.

The hidden weight of the web

The web can feel invisible, but it is not weightless. Modern websites depend on servers, data centres, networks, scripts, videos, ads, tracking systems and cloud infrastructure. Every page load uses energy. Every autoplaying video, oversized image and unnecessary script adds to the load. The early web was basic because it had to be. The modern web is often bloated because it can be.

Some websites now arrive with so much code that your browser appears to be assembling flat-pack furniture in the background. This is not always progress. As sustainability becomes a bigger issue in technology, web designers and developers are being forced to ask whether faster, lighter, simpler websites are not just better for users, but better for the planet too.

Why the 1989 proposal still matters

The importance of Berners-Lee’s proposal is not only historical. It remains relevant because it captured a principle the technology industry keeps forgetting: the best systems help people do something useful without trapping them. The web worked because it was open. It worked because it was flexible. It worked because it allowed different computers, documents, institutions and people to connect without needing one central owner.

That idea is still worth defending. The web of 2026 is richer, faster and more capable than the web of the 1990s. But it is also more commercial, more surveilled and more dependent on powerful intermediaries. The question now is whether the next version of the web will serve users, or simply observe them very carefully while selling them trainers.

Still vague, still exciting

Tim Berners-Lee did not set out to create the modern digital world. He set out to fix a knowledge-sharing problem at CERN. That is what makes the story so appealing. The World Wide Web was not born from a grand corporate strategy or a flashy product launch. It came from a practical need, a clever idea and a willingness to make the result open for others to build upon.

The 1989 proposal was described as “vague, but exciting.” Nearly four decades later, that phrase still fits. The web remains unfinished, unstable, brilliant, frustrating, useful, dangerous, funny and deeply human. It gave us access to the world’s knowledge. It also gave us pop-up cookie banners. No invention is perfect.