Haunted Internet Theory

Overview

The video argues that the internet has shifted from a relatively open, playful space into what the presenters call a “haunted internet”: a digital environment increasingly shaped by opaque platforms, surveillance, censorship, bots, and AI-generated content. The “haunting” metaphor describes how people’s identities and creative work can persist online in altered forms, beyond their control.

From Open Communities to Platform Gatekeeping

The presenters recall early personal websites and online communities as spaces where participation was voluntary and people could experiment. They contrast that period with their later experiences on major platforms, where they say controversial videos were suppressed, view counts behaved suspiciously, subscriber reach seemed constrained, and search results increasingly favored official or institutional sources.

They interpret these experiences as evidence of algorithmic gatekeeping. They also argue that platform rules can limit expression without providing the protections users might expect in physical life.

AI, Impersonation, and Synthetic Media

A major focus is the growing use of AI to imitate people and produce convincing media. The presenters describe alleged cases in which their voices, appearances, and home settings were copied for advertisements or other videos without their consent. They also point to AI influencers, fake fundraising appeals, dubious medical advertisements, and fabricated historical images.

The presenters worry that synthetic material can mislead viewers, divert attention from real people, distort historical understanding, and make authentic footage harder to distinguish from fabricated content. They cite an AI-generated courtroom statement attributed to the murdered Chris Pelkey as an example of a digital likeness being used to speak for someone who can no longer consent.

Broader Risks

The video connects these examples to several wider concerns:

• Bots may create false impressions of public consensus.
• AI-generated images could support misleading official narratives.
• Surveillance systems may wrongly identify or accuse people.
• AI could strengthen authoritarian control through personalized propaganda, monitoring, and automated enforcement.

These points are presented as concerns and interpretations, alongside anecdotes and examples—not as the results of a systematic investigation into every claim.

Social Media and Behavior

The presenters argue that social media can influence people’s emotions and behavior. Citing research on emotional contagion, they suggest that negative feeds, doomscrolling, and engagement-driven algorithms may amplify hostility and anxiety offline.

They also worry that people may change their speech or beliefs to avoid online harassment or “cancellation,” and that bot-filled spaces can make it difficult to know whether one is interacting with real people.

Conclusion

The video concludes that society has become deeply dependent on digital systems for communication, work, payments, and daily life, making withdrawal difficult. The presenters describe this dependence as a trap: people continually provide platforms with personal data and creative work while having little control over how those assets are reused.

They call for recognizing these risks rather than treating the changes as normal, while emphasizing that simply turning off devices cannot restore the world that existed before.

Presenters and Contributors

The two presenters are not identified by name in the subtitles. People named or discussed include Brett Olson, Eliezer Yudkowsky, Dan Vogel, and Chris Pelkey, whose likeness was used in an AI-generated courtroom statement.