Dead Internet Theory vs Unreadable Internet

There are two theories that we often hear talked about on Imgur, and they are similar, but not the same. They are the Dead Internet Theory and the Unreadable Internet Theory. For anyone that is unfamiliar with these two theories, I thought I would do a little explaining.

First the Dead Internet Theory (DIT). This theory is generally considered a conspiracy theory, and it states that the internet that we all use every day has been intentionally converted, over time, to consist mostly of bots and algorithms designed to capture your attention. The theory states that user generated content has been subverted by AI generated content, and that the internet is “dead” in the sense that it is no longer controlled by users.

There is some truth to this conspiracy, as 50% of all traffic on the internet is now bot traffic. Bots are software applications that run specific tasks automatically, and without requiring human control. Examples of bots could be Scrapers, designed to extract user information from the internet, and Social Bots designed to falsely increase numbers of social media engagement for a user. Those celebrities that have 20 million followers? Yeah, maybe not.

In contrast, the Unreadable Internet Theory (UIT) is the theory that over time the internet has become far less user friendly. One reason is that much content is created by bots. So very similar to the DIT. However there are other factors at play here as well. For instance the sheer number of popup ads that appear on websites, which is a real pain point for me. Also that there is so much AI generated content that is incorrect. Content creators making an article require multiple clicks to get through to make their hit numbers increase. Signal to noise ratio rapidly shifting to more noise than signal. So many companies require you to download an app on your device to view their content. The list goes on and on.

There is a difference between good bots and bad bots. Bots can be very useful. For instance bots that monitor websites for content that violates community standards, or bots that scour the internet looking for new or updated websites to make our search engine work better are both highly valuable and provide a great service to users.

In contrast, bots that impersonate people and businesses, or scrap content to feed into an AI are dangerous or fraudulent. Many of these bots directly assist in scams that target vulnerable populations such as the elderly. These types of bots are very bad and shouldn’t be in use.

Where is all this going?

Every year internet bot traffic increases by about 2%. This year was a watershed event, where the total of bot traffic now equals the total of human traffic on the internet. That number is only going to increase as companies try to squeeze every penny out of the internet to train large language model artificial intelligence.

Bots making the internet less desirable is one problem, but the rampant use of AI is an entirely different one. Tech companies are leaning hard into AI being the future of profitability. It allows them to employ fewer humans, and still have the same work output, or much greater output.

There is a very real future where humans will become unemployable in larger and larger numbers because AI controlled equipment is taking their jobs away from them. Have you noticed all the fast food restaurants that have an automated order screen now? That's a job. What about the chat bubbles that pop up on websites to assist you? Every one of those is a job.

Several major superpowers, the United States included, are installing AI into their weapon systems for faster than human responses to threats. Do you remember all those stories that came out after the cold war about how we mistook a flock of geese, or a rising moon, for an incoming missile strike and almost had global nuclear war, but one lone man stood in the way and prevented the apocalypse? That’s a human job too, and AI will not be able to make the moral decisions that humans can. That gut feeling that something is wrong and they shouldn’t annihilate the world is something computers don’t have.

What prompted me to write this down today was an article in the morning paper. I live in Louisiana and police announced on Monday that they had solved an eleven year old murder through the use of AI. Do you really want AI to determine your guilt or innocence? Solving murders is good, but it seems to me there is an awful lot of risk to our society if we start letting AI control these things.

So what can you do about it? I honestly have no idea. “You can cut down all the flowers, but you can’t keep spring from coming.”