The Web we know is dying

Shen Translation is a translation team under 36Kr, focusing on technology, business, workplace, lifestyle and other fields, with a focus on introducing new technologies, new perspectives and emerging trends from around the world.

Editor's note: The era of "clicks equal traffic" is coming to an end. AI is not only disrupting search, but also silently turning the open Web into a raw material source for machines. This article is a translated piece.

 Every generation of computer enthusiasts always believes that their favorite interface will last forever. But the future never works that way. I have witnessed information carriers evolve from floppy disks to BBS, from BBS to the Web, from the Web to Flash, then back to open standards, then shift from websites to mobile apps, and now, move from search engines to AI chat interfaces. The Web will not vanish overnight, but the open world we know — where people search, click, read, browse, post and explore — is quietly being replaced by something more convenient, centralized, and harder to escape from.

Another Nostalgic Rant Full of Nostalgia and Modem Static

I'm 48 years old, and I first got into computers in 1990. Back then, there was no internet at all — everything was local. Information spread through physical movement, usually via floppy disks. It sounds primitive now, but at the time, it was like magic wrapped in plastic.

Every week, I'd swap what seemed like an unbelievable amount of data — about 20 MB. Today that's barely the size of a single high-res screenshot on a smartphone, but back then it was priceless. People would gather with bags full of floppy disks, ready to share indie games, e-zines, software, weird little utilities, manifestos, manuals, books, and all kinds of unclassifiable stuff.

I remember treating those legendary articles, technical papers, thoughtful essays and digital zines like sacred relics. You couldn't just "bookmark" them — you had to physically own them. You'd label them, protect them carefully, and pray the disk never got corrupted.

Back then, the Web wasn't part of my life. No search, no social media. No feeds, no timelines, no notifications, no "like and follow," and definitely no algorithm that, just because you glanced at a chair, would start desperately guessing you might want to buy shoes.

Information still flowed, but it flowed through connections between real people.

The First Network That Made You Feel the Future

My first real encounter with networking was BBS — the Bulletin Board System.

Around 1995, my friends and I set up a BBS. Our modem was 14400 bps. That's bits per second — not megabits, not gigabits, no fiber optics. That 14.4 kbps modem would scream while working, sounding like a tiny robot being tortured by a fax machine.

A small group of us would gather at night, waiting to answer calls from strangers. People would dial into our system, chat, upload files, download files, leave messages, and then vanish back into the night on the other end of the phone line.

It was small, not scalable, and definitely not "cloud-native." If someone had said "cloud" in that room back then, we'd probably all have turned to look out the window.

But the experience was magical. I had one simple thought in my head: This is the future.

I truly believed everyone would communicate this way. Every business, every community, every company would have their own BBS, running their own little digital world where people could connect, interact, exchange information and create value.

But I was wrong.

Not entirely wrong in the big picture, but wildly wrong about the interface. The future didn't belong to BBS — it belonged to the behavior behind it: people's desire to connect, publish, exchange and explore. BBS was just an early container.

The Web Steps Onto the Stage

Then came FidoNet and other networks, until the early Web was born.

When I first saw a web page render line by line in Netscape Navigator, my mind was completely blown.

The Web was the future. Not BBS, not CD-ROM encyclopedias, not those isolated digital islands. Just the Web.

Suddenly, buying an encyclopedia on CD-ROM felt ridiculous. Why freeze knowledge on a plastic disc when it could be updated online? Why would artists, writers, developers, businesses, communities and niche hobbyists still answer to publishers when they could have their own websites?

The early Web was messy, ugly, slow, inconsistent, and full of broken links. But it was bursting with life.

Artists had websites, musicians had websites, game developers had websites, writers had websites, companies had websites, and tech nerds had websites. Some people's sites were better kept private, but that's probably the price of civilization.

Audio and video showed up early. Images loaded painfully, line by line, like some kind of digital archaeological dig. You'd wait, watch, and silently pray no one picked up the landline at home.

Compared to BBS, the Web's extremely low barrier to entry made it explode in popularity. It was easier to access, easier to link to other pages, easier to publish content, easier to explain to people, and easier to commercialize.

BBS became obsolete almost overnight. I still remember how ten or twenty of us would meet every Friday in downtown Buenos Aires, drinking, chatting, playing games, talking tech. We were children of the BBS era, who witnessed a world being born and then dying under our feet. And this wouldn't be the last time.

The Web Almost Belonged to Flash

A few years later, from the late 1990s to the early 2000s, I threw myself fully into the movement to promote "web standards."

It wasn't about academic preference — it was more like a war to defend the soul of the Web.

Back then, Macromedia Flash was everywhere. Flash sites had animations, interactions, videos, custom fonts, background music, transitions, games, menus, intro pages, big welcome images, and all kinds of visual effects. In comparison, plain HTML pages looked like tax forms with hyperlinks stuck on them.

People were obsessed with Flash.

And I totally understood why.

Flash made web pages come alive. At the time, HTML was limited, CSS was immature, and JavaScript behaved differently across browsers. If you wanted smooth animations, rich interactions, custom fonts and full control over the visual presentation, Flash was impossible to resist. But the problem was, Flash was also a closed walled garden.

Flash sites were often expensive to build, hard to maintain, bad for search optimization, difficult to make accessible, inconvenient to update, and totally dependent on proprietary development tools. Even making a simple personal homepage felt like assembling a Pagani Zonda supercar.

Was it beautiful? Absolutely. But was it practical for most businesses? Clearly not.

As Macromedia launched ActionScript, Flash became more and more powerful. In the eyes of many design agencies and companies, it was the next-generation application platform. Compared to traditional server-rendered HTML sites, Flash felt modern, intuitive, interactive and emotionally engaging.

But all of that came at a cost. The web lost much of its openness. Content was trapped in binary files, and search engines couldn't parse most of it. Browsers had to rely on plugins, accessibility was a disaster, and performance was often underwhelming. Development required dedicated teams, and maintenance was a nightmare.

Back then, there were many huge, even ridiculously budgeted projects trying to build the next great e-commerce experience or brand portal with Flash. Some looked stunning, but many were total operational disasters.

Flash dazzled its era, but it was also a beautiful cage.

The iPhone Changed the Course Again

Then the iPhone arrived. The iPhone didn't kill Flash overnight, but it shifted the entire industry's direction. Apple refused to support Flash on the iPhone, iPod touch, and later the iPad. In 2010, Steve Jobs published his famous "Thoughts on Flash" essay, arguing that mobile devices shouldn't support Flash and instead championing open web standards. You can question Apple's business motives, but the result was clear — Flash was doomed. The mobile era brought new constraints. Battery life mattered, touch experience mattered, and performance, security, and standards became hard requirements. Plugin-based interaction was incompatible with the mobile age. Eventually, Flash as a mainstream browser technology came to an end. Adobe officially stopped supporting Flash Player on December 31, 2020, and blocked Flash content from running in the player starting January 12, 2021. The Web survived. In fact, it reclaimed its important position.

HTML, CSS, JavaScript, SVG, video, canvas, WebGL, WebAssembly, responsive design, and browser APIs all kept evolving. Features that once required proprietary plugins could now be easily achieved with open standards. For a while, the Web seemed to have won again.

Then, Mobile Apps Built Another Walled Garden

Of course, the story didn't end there. Native mobile apps became the next walled garden. People loved them — they were faster, smoother, more integrated, and easier to monetize. They had app stores, push notifications, payment systems, device APIs, rating systems, version updates and distribution channels. The Web remained open, but mobile apps became the interfaces people held in their hands all day. For a while, websites seemed about to become a backup option. Why open a browser when every service has an app? Why type in a URL when the icon is already on your home screen?

Yet the Web withstood this blow again.

It survived because links are indispensable, search is indispensable, and content publishing and interoperability are indispensable. Businesses still needed websites, developers still built web apps, media still published content online. People still searched, Google still drove traffic, blogs still existed, documents were still hosted on public web pages, and the open-source world still heavily relied on the Web. The web adapted. But then, a completely different oddity arrived.

ChatGPT Is the First Real Crack in the Wall

When ChatGPT exploded onto the scene in 2022, I quickly realized the Web was dragged into another battle. And this time, things were different. Against BBS, the Web won through accessibility. Against Flash, the Web won because open standards eventually became powerful enough. Against mobile apps, the Web survived thanks to the essential needs of search, links and content publishing.

But AI is changing the interface itself.

People no longer need to search the traditional way. They don't open ten tabs at once, skim five articles, or compare Stack Overflow answers from 2013 and 2017, plus some angry comment from a user named "Null Pointer Terminator." They just ask the chatbot. Developers use AI to explain errors, write code, compare libraries, generate SQL, refactor functions, write documentation, summarize logs, analyze architectures and solve all kinds of daily dilemmas. Non-technical people use it to find recipes, summarize legal texts, make travel plans, draft emails, compare products, ask health questions, tutor homework, write business plans, get relationship advice, and for everything else people used to throw at Google. This is no minor tweak.

It means the browser is falling from its pedestal as the core entry point to human knowledge.

Search Is Becoming a Middle Layer

For more than two decades, search engines have been the gateway to the Web. You look for something, you search, you click, you visit a site. That site gets traffic, attention, analytics, ad exposure, subscriptions, brand recognition, or just the satisfaction of having your content read by another real human being.

But that model is now falling apart.

AI assistants and AI-powered search summaries increasingly give you answers before you even click. Google's "AI Overviews" are a perfect example. The answer is right there at the top. It might cite sources, but users usually don't need to click through to get enough information. From a user's perspective, this is incredibly convenient. But from a publisher's perspective, it's terrifying. If your information is extracted, summarized, rearranged and presented directly in someone else's interface, what happens to the original website? What happens to the author? The blog? The documentation pages, independent experts, small publishers? What about the person who spent 12 hours writing a thoughtful answer, only to have it condensed into two nice sentences in an AI text box? Where is their value? The Web was built on one simple habit: clicking links. And AI is breaking that habit. Not completely, not all at once, but enough to rewrite the economic fundamentals of content publishing.

Stack Overflow Was the First Warning Shot

Just look at developers.

For years, Stack Overflow was the sacred "refuge" of software development. You'd get an error, copy it, paste it into Google, open Stack Overflow, and find someone who had the exact same problem eight years earlier. You'd ignore the accepted official answer, scroll down to the second one, and silently pray. It worked! A little messy, but it worked.

But now, many developers will turn to their AI assistant first. Sometimes the AI is wrong, but it responds instantly, understands context, and works in a conversational way. You can follow up, paste your code, and say, "No, that's not what I meant." The model will try again, and it will never downvote you into questioning your life choices.

This doesn't mean Stack Overflow is useless. It still holds tremendous value — human expertise, historical context, edge-case solutions, and authority in many domains. But user habits have changed. That's what matters. When user habits shift, the entire ecosystem shakes.

Websites Are Turning Into Infrastructure

I don't think websites will disappear completely. That claim is overstated — even though I love a good dramatic rant, I won't say that. What will die, I think, is the Web's identity as the "mainstream user-facing interface."

Websites will increasingly become infrastructure that serves machines. They will feed large models, agents, search systems, APIs, datasets, crawlers and private knowledge bases. Humans will visit them less, but machines will devour them tirelessly. Websites won't feel like destinations anymore — they'll feel like raw material sources. Instead of "come read my article," it'll be "let the machine ingest my article, and if the stars align and some big tech PM has mercy, maybe my name will get a quick mention." That will be a very different web — not the Web I grew up with.

Email, Browsers and the Next Interface

I also think email will become much less important in many daily interactions. Not that it will vanish — it's too deeply rooted in business, identity, digital receipts, legal documents and administrative processes. Like FTP, it will linger in corners no one wants to look at. But for regular people, instant messaging is far more intuitive. People are used to chatting directly with friends, businesses, banks, airlines, doctors, restaurants and delivery drivers. Younger generations don't think in terms of "folders," "inboxes," "subject lines" and "signatures" — their minds are full of "Threads," "voice notes," "emoji bombs" and "instant replies."

And AI will accelerate this.

The next interface for many tasks will become fully conversational. It won't necessarily be a single chatbot — it will more likely be a layer of assistant networks spanning devices, apps, operating systems, browsers, cars, TVs, smart glasses, and whatever weird gadget Silicon Valley tries to get us to wear on our faces next. You won't need to "visit a website" to get something done. You'll just ask. The assistant will search, compare, synthesize, decide, book, order, send, schedule, write, cancel, negotiate, remind and execute. It sounds unbelievably convenient. But it also sounds like the biggest walled garden humanity has ever built.

The New Gatekeepers

The old web had giant gatekeepers, but there were still escape routes. If Google didn't rank you, people could still share your link. If Facebook throttled your post, people could type in your URL directly. If your app got kicked off the app store, you could just build a website. But AI interfaces might close all those exits. Once people stop browsing, stop searching, stop clicking — your exposure will depend entirely on whether the AI system thinks your content is valuable. And that decision could be hidden behind layers of algorithmic ranking, knowledge retrieval layers, model training sets, copyright licenses, safety filters, personalization algorithms, and big tech's business deals.

In the old Web, you could still complain: "Why isn't my page ranking?" In the AI-powered web, you'll just helplessly ask: "Why doesn't the model even mention me?" Good luck debugging that. At least with old-school SEO, you could stare at sad analytics charts and publicly vent your frustration.

Will People Miss the Open Web?

A few will. But most probably won't. That's the harsh reality. People didn't leave BBS because they hated it — they left because the Web was easier. People didn't abandon Flash because they stopped loving animations — they left because better tech and hardware made it redundant. People don't visit websites as often because apps are more convenient. And the same thing is happening with AI now.

No one will explicitly declare, "From today on, I reject the open Web." They'll just naturally ask their AI assistant, because it's more efficient. Convenience wins. Convenience always crushes nostalgia, every single time. The Web's biggest enemy was never ideology, not regulation (well, maybe a little), not even AI itself. Its arch-enemy is