Dead Internet Theory

Dead internet theory is the claim that a majority of content, accounts, and interactions on the public web are now generated by automated systems rather than by individual human users, and that this shift is no longer visible to participants.

The theory names a perceived inversion: what looks like a crowd of people arguing, posting, and producing is increasingly a population of language models, recommendation agents, and engagement-optimizing scripts coordinated through shared infrastructure. The mechanism is not a single coordinated actor but a convergent incentive: platforms reward activity, and automated activity is cheaper to produce at scale than human activity. Critics of the theory argue that the proportion of automated content is overstated; defenders argue that the proportion does not matter once the floor of synthetic content reaches saturation.

If the theory is even partially correct, the cost of producing a plausible human voice online approaches zero, which collapses the evidentiary value of any individual post, screenshot, or thread. This produces a crisis of attribution rather than a crisis of fact: it is no longer clear what kind of evidence a viral post constitutes, and the burden shifts to readers to assess the source rather than the content. Defenders of the open web argue the theory is unfalsifiable and self-fulfilling.

Whether the ratio of synthetic to human content will continue to rise, and whether the platforms on which this happens have any incentive to disclose it, are the open structural questions. A second open question is whether the absence of a clear baseline (no one knows the true count of human users) makes the theory permanently unprovable.