The internet isn’t dead. But if you’ve spent any time online lately, you might understand why some people wish it would hurry up and get there.

In late October 2026, WIRED published an essay with a deliberately provocative title: Please Let the Internet Die. Written by Miles Klee for the magazine’s Ghost Week series, the piece makes a familiar complaint with a more radical conclusion. As oligarchs capture the internet, AI-generated garbage floods it, and it has changed into an environment that seems designed to unleash humanity’s worst instincts. Maybe it’s best to stop trying to save it.

It’s a dramatic position, but not an entirely unreasonable one.

As WIRED’s promotional post puts it, being online increasingly means being manipulated and exploited. It’s becoming increasingly difficult to distinguish genuine human conversation from manufactured engagement as algorithms compete for attention, platforms monetize outrage, and automated content makes it harder to distinguish between the two.

Many people who remember the web as an endless collection of interesting people, obscure websites, and independent blogs may be surprised at how much has changed.

Still, there’s a difference between acknowledging that the internet has serious problems and declaring it ready for the digital graveyard. Klee makes some compelling observations, but his argument also raises a question worth considering: Are we witnessing the death of the internet, or simply the consequences of allowing a handful of companies to decide what it becomes?

We’ve Heard Predictions of the Web’s Demise Before

This isn’t the first time WIRED has announced that the internet as we know it is finished.

Back in 2010, Chris Anderson and Michael Wolff published The Web Is Dead. Long Live the Internet. Their argument was that people were abandoning the open web in favor of applications, closed platforms, and services that offered more convenient ways to consume content.

The prediction generated plenty of debate, and for good reason.

As The History of the Web explains, one of the central problems with the original argument was how it interpreted web traffic. Rob Beschizza pointed out that the much-discussed chart showed the web’s share of overall internet traffic rather than its total usage. The distinction matters. The web’s percentage could decline while the actual amount of activity continued growing.

Scientific American also challenged the obituary, arguing that the web remained an important and expanding part of the digital world.

With this history in mind, the latest declaration might seem like another premature obituary. Klee, however, makes a somewhat different argument.

In 2010, open web platforms were losing ground to closed ecosystems and applications. As of 2026, the internet itself has become an increasingly unpleasant place to be.

The infrastructure is still there. It’s what we’ve built on top of it that’s the problem.

The Internet Has an Engagement Problem

The central argument in Klee’s essay is that today’s internet rewards the wrong things.

According to one discussion of the article, the problems include algorithmically manufactured outrage, AI-generated content sludge, disappearing links, and platforms that prioritize monetization over meaningful experiences.

These aren’t isolated annoyances. They’re symptoms of an ecosystem where attention is the primary commodity.

Social platforms don’t necessarily need you to enjoy your time online. They need you to stay online. Anger, controversy, fear, and arguments can accomplish that goal.

As a result, the internet can appear very active but offer surprisingly little in terms of value.

Every day, millions of posts, videos, comments, and articles are published. Although the web is much bigger now, finding genuinely interesting content or having a worthwhile conversation can be challenging.

Klee also draws on technologist and journalist John West and his book The Internet Will Die, and So Will You: On Learning to Live in Our Digital Age. As discussed in this reader’s commentary, West explores how fragile the supposedly permanent digital world actually is.

An example is the 2001 rail accident in a Baltimore tunnel that damaged cabling and disrupted internet connections across the nation. A secondary account of the WIRED essay highlights the incident.

I think it’s a useful reminder that the internet isn’t some magical cloud of indestructibility. In order for it to work, it depends on physical infrastructure, companies, cables, servers, and the people in charge of maintaining them.

But its fragility isn’t necessarily evidence that it should die. It’s evidence that we shouldn’t take it for granted.

Dead Links and AI Bots: Signs of a Bigger Problem

Klee’s argument is strengthened by two developments: the declining availability of online information and the growing presence of automated processes.

Start with link rot.

Research has found that approximately one-quarter of the links in New York Times articles published between 1996 and mid-2019 no longer work. A Pew Research report also found that more than half of Wikipedia entries contain at least one dead link citation. Both figures are referenced in coverage of the WIRED essay.

That’s more than a minor inconvenience.

The internet has become one of the largest repositories of cultural, historical, and journalistic information. Many of those materials, however, are dependent on websites that can disappear, change ownership, or cease to exist.

Suppose you rely on a library where one-quarter of the references lead to empty shelves. Essentially, that’s the problem we’re creating online.

Then there’s automation.

According to secondary coverage reports, a cybersecurity firm has found that AI agents now outnumber human users, since their traffic has increased by approximately 80 times in 2025. Considering that the secondary report doesn’t identify the company, we should treat that statistic with caution rather than treating it as fact.

Nevertheless, the broader concern is legitimate.

According to the Dead Internet Theory, automated conversations and bots masquerading as human activity were alleged to be at the centre of a massive internet conspiracy. Today, AI-generated content is more sophisticated, making this distinction more difficult.

As an AdExchanger newsletter observed, the theory doesn’t sound quite as far-fetched as it once did.

In a world where machines increasingly create, distribute, and interact with content aimed at attracting other machines and algorithms, what exactly are humans getting out of this arrangement?

Where the Argument Goes Too Far

Here’s where the internet’s latest obituary falls short.

Despite dead links being a serious archiving issue, the internet is not dying. Additionally, an increase in automated traffic does not necessarily indicate a decline in human participation.

In other words, these are problems related to maintaining, measuring, and monetizing the internet. Despite this, they do not mean that the underlying technology has reached the end of its useful life.

By declaring the internet dead, one risks confusing a broken business model with a broken technology.

Other commentators have questioned whether Klee’s headline accurately represents his actual proposal. A Halloween essay on Substack suggests that the piece isn’t really advocating for shutting down the internet. It’s more of an argument for abandoning the current version of it.

Meanwhile, Mathew Ingram’s roundup of reactions highlights several competing perspectives.

According to cryptographer Matthew Green, it’s like throwing whatever you can in front of zombies. Ben Thompson, meanwhile, suggests that the collapse of the advertising- and click-driven internet economy could create opportunities for subscription-based publishing.

That second possibility is particularly interesting.

Maybe changing the economics could improve the experience if advertising incentives have made online publishing a race for clicks. If people feel like their attention isn’t unlimited, they will probably be more willing to pay for useful information, independent journalism, and communities that don’t abuse their attention.

Rather than declaring the entire system irreparable, it would address one of the underlying problems.

Maybe We Need a Smaller, More Human Internet

Klee’s essay doesn’t necessarily suggest that everyone should abandon the internet. It’s about reconsidering where we spend our time online.

There are still alternatives to the major platforms.

The experiences offered by independent blogs, personal websites, niche forums, email newsletters, private group chats, and smaller online communities are not necessarily dependent on maximizing engagement. It’s often more difficult to find them and less polished, but that’s not necessarily a disadvantage.

Think of it as the difference between fast food and something prepared in your own kitchen. One is optimized for scale and distribution. The other might be slower, less convenient, and more personal.

People may eventually prefer the digital equivalent of home cooking, according to Ingram’s roundup. The challenge is making that alternative sustainable.

The marketing budgets, recommendation engines, and distribution networks of major platforms aren’t available to smaller websites. The independent creator still needs an audience, and the community needs a place to gather. Unless alternative internets are easier to find or maintain, people will continue to use the same platforms that they complain about.

That’s why simply telling people to log off isn’t useful.

Rather than turning every interaction into an advertisement funnel, we need better ways to discover independent creators, preserve online information, and support worthwhile publishing.

The Internet Doesn’t Need to Die. Our Relationship With It Needs to Change.

WIRED has been wrong about the web’s death before, and there’s no reason to assume this latest prediction is entirely accurate.

However, Klee’s essay is important because it points to something deeper than technological nostalgia. Many of the incentives driving the internet’s change have made it worse for people who use it.

Servers will continue to run. Websites will continue to appear. AI-generated content will become more sophisticated, and social platforms will find new ways to monetize our attention.

None of that means we have to accept the current arrangement.

Instead of treating the biggest platforms as synonymous with the internet, perhaps we should stop referring to them as such. If we spend less time in spaces designed to provoke us, we can devote more time to people, publications, and communities that make being online worthwhile.

I don’t think the internet needs another obituary. We need a better alternative to the version we’ve permitted to take over.