We Used to Follow Links. Now We Follow Feeds. Here's What That Actually Broke.

Somewhere around 2010, something quietly died on the internet and most people didn't notice because the thing that replaced it was more convenient.

What died was the linker. Not a person, exactly — more like a role. The blogger who ended every post with a "worth reading" section. The forum regular whose recommendations you'd actually click on. The personal homepage with a lovingly maintained "links" page that served as a window into someone's entire intellectual life. These people were the connective tissue of early web culture, and they're mostly gone now — replaced by algorithmic feeds that do the pointing for us, faster and at scale.

The trade-off felt reasonable at the time. Who has the bandwidth to maintain a blogroll in 2024? But spend some time thinking about what actually got lost in that exchange, and the accounting starts to look a lot less favorable.

What a Link Used to Mean

In the early web — and we're talking late 90s through the mid-2000s here, before social platforms swallowed everything — a hyperlink carried weight. When someone linked to something, they were making a claim. This is worth your time. I vouch for this. My reputation, such as it is on this corner of the internet, is attached to this recommendation.

That's not a small thing. It meant that the act of linking was curated, intentional, editorial. You followed a link because a specific human being whose taste you had come to trust had put it there. The path through information was shaped by judgment — fallible, idiosyncratic, sometimes wrong, but always authored.

The discovery experience that came out of this was genuinely strange and genuinely good. You'd land on an essay about urban planning and find a link to a musician you'd never heard of, because the person who wrote it happened to be obsessed with both things. That lateral movement — across topics, across disciplines, guided by a single person's curiosity — built a kind of knowledge that algorithmic feeds are structurally incapable of producing.

How the Feed Killed the Linker

The shift happened in stages. First came RSS aggregators, which were actually fine — they centralized content but preserved the human curation layer. Then came Twitter, which seemed like a linker's paradise at first. Sharing links was the whole point. Except Twitter also introduced the retweet, the like, the trending tab, and eventually an algorithmic timeline. The curation layer started to erode.

Facebook accelerated the process. The News Feed, optimized for engagement from the jump, turned link-sharing into a performance — what you shared said something about you to your social network, which changed the calculus entirely. You weren't curating for someone who trusted your taste. You were broadcasting to an audience that included your coworkers, your family, and that one person from high school.

And then the platforms got smarter. YouTube's autoplay. Spotify's Discover Weekly. TikTok's For You page. Instagram's Explore tab. Each one made the case that you didn't need a human intermediary anymore — the system could figure out what you'd want next, faster than any individual curator could, with no friction, no personality, no editorial voice getting in the way.

They were right about the speed. They were wrong about everything else.

The Skills That Atrophied

Here's what happens when you stop navigating the web through human-curated links and start relying entirely on algorithmic feeds: you lose the ability to browse with intention.

This sounds like a small thing until you try to do it again. Go ahead — open a new tab right now and try to find something genuinely interesting without using a search engine or a social platform. Just... navigate. Follow your curiosity across the web the way people used to. Most people find this surprisingly difficult. The infrastructure for it has largely collapsed, and the habit has atrophied from disuse.

There's also a deeper skill that gets lost: the ability to evaluate a source based on its curatorial track record. When you followed a linker, you were building a model of their taste over time. You learned when to trust them and when to be skeptical. You understood their biases and blind spots. That kind of calibrated trust is different from — and better than — the undifferentiated stream of content that a feed produces, where everything arrives with the same implicit endorsement of the algorithm.

The People Rebuilding It

Something interesting has been happening in the margins of web culture over the last few years. A small but genuinely committed group of internet users has started rebuilding the personal web ecosystem that the platforms ate.

Some of it is happening through newsletters, which have become one of the few spaces where individual editorial voice still shapes what you read. Substack, Buttondown, and similar platforms have created a new generation of linkers — people whose entire value proposition is I read a lot of things and I'll tell you which ones matter.

Some of it is happening through personal websites. The IndieWeb movement, which has been quietly chugging along for over a decade, advocates for owning your own content and your own links. People are building personal homepages again, complete with blogrolls and hand-maintained link collections. It's niche. It's labor-intensive. It's also a pretty direct rebuttal to the idea that algorithmic curation is just better.

And some of it is happening through Discord servers, group chats, and private forums where the link-sharing culture of the early web has retreated — small enough that the social dynamics work, intimate enough that the curatorial layer is still human.

Why This Matters More Than It Seems

You could look at all of this and say: okay, some people are nostalgic for the old internet. That's fine. But the implications go beyond personal preference.

The replacement of human curation with algorithmic feeds has had real effects on how information moves, how culture forms, and how people develop taste. When the path through information is shaped by engagement optimization rather than editorial judgment, the stuff that spreads is the stuff that triggers a reaction — not the stuff that rewards sustained attention or builds genuine understanding.

The linker, for all their flaws, was trying to do something different. They were trying to share what they found valuable, which is a fundamentally different goal than maximizing your time-on-platform. That difference in intent shaped everything downstream.

Rebuilding that layer — even partially, even imperfectly, even in small corners of the web that most people will never find — is one of the more interesting forms of cultural resistance happening right now. It's quiet, it's slow, and it requires actually developing a point of view worth sharing.

Which, come to think of it, is exactly the kind of thing that doesn't show up in a feed.