Dead Internet Theory and the rise of Intranets?

Recently Rosie covered the Digg shut down. Digg placed the blame on the prevalence of AI Agents flooding the internet with an endless stream of derivative content while at the same time boosting that content on various platforms. That combination makes it difficult for our old content curation algorithms to perform well, as most relied on basic view count or votes that can be easily faked by AI Agents.  

If this sounds familiar, it's because it's an extension of a timeless battle on the internet. Back in my day, we'd call 'AI Agents' 'Bots'. Bots could have been programs, they could also just be a bad actor with multiple accounts/logins that allowed them to flood the system. The process was harder before AI, but still extremely prevalent. 

There's a theory that dates back to the 2010s called 'Dead Internet Theory.' It states that eventually - if not already - the internet is mostly comprised of bots. That there are more bots online then actual people, and that statistically you are more likely to be interacting with a bot then a real person at any given time. AI has poured fuel on this fire, making it more likely then ever. Eventually, the theory claims, this will lead to the death of the internet, as it will simply become a bot net created by and for bots. 

Interestingly, this has long been pontificated in science fiction. Cyberpunk dystopias frequently feature 'intra-nets' or internal, private, and carefully curated information networks. This is due to the wider public 'internet' being unsecured and generally full of garbage, misinformation, or broadly too vast to be useful (to put things mildly). 

We may be at the threshold of crossing into an era where private 'intranets' become the norm while the wider internet dies off. A Chrome extension already made headlines for blocking post-2022 content, freezing your internet browsing in a pre-AI world. The rise of private member organizations shows people are interested in more carefully managed and intentional content. 

The tradeoff in my view is information silos. With an intranet model, information can become easily fragmented with any one intranet not containing all the required information or having large gaps in the knowledge base. 

What does the Motaverse think? Where is the internet headed, are intranets the future, and how is AI shaping it? 

Further reading in the Observatory: