Centralized technology is shrinking the internet | Opinion

The internet maintains the illusion of an infinite scroll, with more pages of text and hours of video than we could ever consume in 100 lifetimes. The reality is that this “infinite scroll” occupies a small sliver of the actual internet itself, though. Over 90% of the internet is hidden in the Deep Web, making it largely inaccessible to the public, and it certainly is not being indexed by mainstream search engines.

Meanwhile, the free and open web we can access has given way in the last few decades to the kind of large corporate control we’ve seen in other media. In the U.S. alone, six major media companies (i.e., AT&T, CBS, Comcast, Disney, NewsCorp, and Viacom) collectively control 90% of the media.

Unlike dominant forms of media in the 20th century, like television, though, the internet offers an increasingly invasive feedback loop of desire meant to sell you something. Every networked device, from earbud to smartwatch to smartphone to laptop to Smart TV, is a recorder that can capture your consumption down to the millisecond.

Not only that, equipped with microphones and cameras, these are devices that can capture audio and video recordings of you, whether you are volunteering that data in the form of a selfie or not. With a network controlled only by a few Powers That Be, the priorities of that network and how it was turned on us began to shift, with an emphasis on user “engagement,” regardless of the quality of that “engagement,” paired with hyper-targeted advertising that could take advantage of these highly distractive and highly invasive devices.

The problem of the closed, limited, and exploitative web will only grow with the maximization of shareholder value through the direct promotion of AI and the indirect promotion of the computer chips required to power that AI. Get used to a closed web, less in touch with its own humanity, increasingly populated by junk content and junk values.

Choking off the last breaths of the “dead internet”

How did we get here?

Even before the rise of AI, there has been the terrible problem of content quality control on the internet, a problem that has gotten worse with the prioritization of better advertising through user targeting. The end goal of the internet and the algorithms that now drive it has become selling you something you already bought yesterday. Anything that does not align with that sales pitch either becomes deprioritized or discarded.

A tiny number of players not only control what becomes visible, trending, or remembered, but also sideline divergent or fringe content, not for merit, but simply because it doesn’t fit their algorithmic or commercial priorities. Independent creators, critical voices, and niche information struggle for visibility, often buried beneath sponsored or mainstream content.

The surge of AI-generated content and synthetic media is more than a philosophical dilemma; it’s a dangerous shift that threatens to warp our shared perception of reality. Discerning truth from fabrication was already difficult; now, it’s becoming exponentially harder.

And an internet that is mostly AI might get here much more quickly than we think. A 2024 “Bad Bot” report from cybersecurity firm Imperva estimated that nearly half of all traffic on the internet at the time was automated, with bots making up 42.3 percent of internet traffic in 2021, a number that jumped to 49.6 percent in 2023. Only a few years later, a large-scale AI detection analysis of 900,000 newly created web pages (April 2025) found that 74.2% of them contained AI-generated content.

Other projections suggest that by 2026, up to 90% of online content may be synthetically generated, although this remains speculative.

It may not be very long before we see the “theory of the dead internet,” when there is more content churned out by bots than actual humans, become the “fact of the dead internet.” The internet as we know it, arguably the most radical tool for mass human expression since the Gutenberg printing press, will rest on a junk heap created by decades of Silicon Valley penny-wisdom and pound-foolishness.

Taking back the web, one human at a time

By championing human-centric content and advocating for creators with balanced perspectives, not just the perspectives that gin up the most engagement at the least cost, whether they come from actual humans or not, the web can still become a place people want to share, driving a self-reinforcing flywheel where every new tool that unearths the unfound fuels greater discoverability, engagement, and growth.

Decentralization offers one possible way out. It is an incentive structure that is distributed, not hoarded by the very, very few, and it restores agency to users and communities. It offers resilience to the very fabric of the web that extends beyond the benevolence and motivations of the handful of the richest people on the planet.

It is clear that extrapolating the status quo of the last decade will not lead to an internet that remains a free, fair, and interesting place. Content will disappear without recourse, and the boundaries of discovery will be set not by what exists, but by what corporations allow. We must continue to prioritize privacy, permanence, and independence.

Back in the earliest days of the internet, DirecTV had a popular commercial where a web surfer was told, “You’ve reached the end of the internet.” If these large companies have their way in continuing to needlessly shrink the internet for their own limited and profit-based motivations, it won’t be long before we reach the end of the internet.