Combatting the Dead Internet Theory in an AI World

(Source: sdecoret/stock.adobe.com)

The Dead Internet Theory may have started out as an exaggeration bordering on conspiracy, claiming that most of the internet is run by bots. However, given the mammoth advances in artificial intelligence (AI) over the past few years, what was once speculation is starting to look like a plausible future. On top of that, the algorithms already in place prior to the boom in generative AI often favored homogenous styles of media, further contributing to the empty, flat feeling of the internet.

The dangers of a dead internet are that reality becomes warped as fake material becomes indistinguishable from real, trust is lost, and people no longer turn to the internet as a useful place of knowledge or connection. There’s a secondary impact, too, in that models trained on this less genuine, bot-generated content produce more of the same and become a hindrance rather than a help to people. Still, companies can combat this situation by implementing measures that impact the way they present their brand to the world, as well as how they interact with customers and build meaningful communities around their products and published information.

Theory Origins

Initially popular in 2021, the Dead Internet Theory proposed that the majority of traffic and interactions on the internet was attributable to bots. When the theory came out, it was difficult to believe that it might be the case because automation capabilities were not as powerful, and it was more straightforward to verify whether information came from a genuine entity. However, nowadays, looking at social media like LinkedIn, posts often appear to have the same kind of format with the same pattern of content. Replies on forums and comments under posts are starting to feel all the same—bland, generic congratulations, agreements, or appreciation. Even profile photos come across as overly polished to the point of being fabricated or highly edited. Why is that the case?

Why the Resurgence Now?

Large Language Models (LLMs) and other generative AI are increasingly capable of producing extremely convincing text, images, and video. Add to that the sophisticated workflows producible by AI agents without supervision, and the reality of an internet driven by bots for bots seems more within reach. With each new release of a video or image generation model, the outputs produced become more realistic and then flood the various platforms people frequent, making the prevalence of bots or humans posting like bots feel more acute.

Generative AI, however, is not solely to blame for the “dead” feeling of the modern internet. Social media, search engines, and most large content curating sites often participate in the gamification of attention. Interactions are then optimized for metrics rather than true engagement, and that comes off as shallow—being just unique and interesting enough to spark the transfer of attention but not providing much else beyond that. Couple this with the style of generative AI’s outputs, and the outcome is generic engagement slop.

This landscape of algorithmic curation also encourages and rewards behavior from humans that mimics AI-generated content, further contributing to the feeling that the internet is “dead.” Cultivating a meaningful following or a fruitful community can be costly in terms of time and effort and yields mixed results where some content is well-received while the rest falls flat. Without some kind of automation, it might feel like a waste to try and build a community, producing engaging content only to find it doesn’t work because of a black box algorithm at the helm that changes frequently with limited feedback on how to improve. The instinct then is to outsource that portion, so at least if it fails, it fails at a low cost. The result is people using AI to hedge their bets alongside bots in the form of agentic AI using generative models. So, what does this mean for the internet going forward?

Implications for Knowledge, Trust, and Community

Whenever a customer reads a blog post from a company only to find out that it was completely AI-generated without human oversight and contains incorrect information or has a nonsensical image in the headline, their trust is eroded. Any time spam comments are left to sit on posts or forums, the impression is that the site is not free from unverified users, including bots. If people using the internet cannot be certain that what they are reading, consuming, or learning is legitimate or that the person they are communicating with is not an automated process, it jeopardizes their ongoing use of the internet as a source of information or as a place to organize a community around particular topics. This leads to several key impacts, such as epistemic collapse, where misinformation is indistinguishable from truth, and relationship breakdown, where the substrates used to form online connections disappear. Both of these could mean the internet becomes unusable and people retreat offline. But it’s not all doom and gloom. There are ways to fight back.

Reviving the Internet

At the enterprise level, companies with an online presence have several options against the advance of the Dead Internet Theory. Firstly, they should avoid AI where it matters to customers, so there is no opportunity for loss of trust. If AI automation is required, transparency around its use is imperative, as is verifying information that is exposed to the public. Similar advice applies if a company produces white papers or blog articles to solidify its reputation as a subject matter expert. The “dead internet” feeling can be avoided by investing in experts to deliver genuinely helpful information and by keeping up to date on what actual consumers in the niche require instead of trend chasing.

Additionally, building closed communities and forums where some verification is required can help keep the bots at bay and foster genuine exchanges. Appropriately moderating content is also key. This goes hand in hand with educating employees and community members on the signs of AI-generated artifacts and focusing on critical reasoning skills to make informed decisions about content online.

Conclusion

What started out as a borderline conspiracy theory is now finding roots in reality as generative and agentic AI, along with curation algorithms, saturate many spaces on the internet. As people become increasingly unable to distinguish genuine content from fake, the internet risks becoming completely untrustworthy as a source of information and unreliable as a facilitator of building online communities and brand presence. Fortunately, companies can fight back by ensuring limited AI interactions with customers, upholding a commitment to verifying community members or content slated for release, and advancing education on the signs of AI-generated content to limit its impact. By doing this, the future of the internet can be reclaimed as a place for connection and learning rather than empty bot traffic.