A Human's Guide To Detecting AI Generated Content

The Dead Internet Theory

I believe google is one of those that makes bots, after all they work like a search engine, where they get the most accepted content first, Is the same as doing an ad.

Heres a relevant image I found on image board bots:

Conclusive Notes:

Really, bots are most likely a much smaller problem compared to algorithmic filter bubbles. "if you liked that, you'll love this!". Think about it logically, what's more effective:

1) Bot spamming to derail conversation / divide & conquer on subject you don't want discussed

2) Algorithmically shadowbanning conversation on that subject so that almost no one ever actually sees it, and those who do are essentially trapped in their own bubble without access to the outside.

Of course both are used and both can be effective in different situations, but I believe option 2 is far more widespread. For a metaphor, certain online games to a similar thing, like csgo and titanfall. Those players who get reported a lot for toxicity or cheating, instead of just being banned, are simply placed in a "low trust factor" que, so that they only meet other cheaters.

There's no way to check your own trust factor, so as far as you know you might think things are normal and the game has just become worse, meanwhile you're separated from the regular players. In a similar way, meta-meatspace, algorithmically driven social media platforms lock you in a closed of version of the site when you participate in discussion they don't want. We think of websites like twitter as a single space, and they were that way once, but with modern AI dictating what you see, these places are more like independent patchworks of separate closed rooms.

Rooms which you can't escape from, rooms which you don't even realize you are in. This makes you easier to surveil and control. Thanks to the GDPR, you can actually see what room you're in by requesting your data, but this only gives you a vague idea. You might think the goal of these rooms is to lock you in with only people similar to you, but in fact, the goal is to generate high intensity emotional responses like outrage or humour.

These emotional states make you more likely to stay on the site for longer, or interact with the site, allowing them to collect data.

Therefore, in these rooms you are actually more likely to see things you disagree with, but only the very surface of them. You will not have to actually face a detailed counterpoint to your argument, only a brief and incomplete summary maximizing for high intensity emotion and minimizing for coherant logic.

Twitter does this by imposing a strict character limit, it is physically impossible to discus complex ideas in such a short space, so conversations naturally devolve into insults and shock value.

>reddit enforces this via pseudo-democratic upvote downvote system, which is a little more subtle than twitter's heavy handed approach.

The posts with the most upboats go to the top, and thus get seen the most. site-wide upvoats even contribute to an rpg like xp system linked to your single-identity account on the site. It is clear that the goal of this game is to make the number go up. This voating system discourages controversial posts. imagine two posts.

One gets 100 total votes, 50/50 upvote downvote, this cancels out and is equal to 0. Then another post gets 1 upvoat, it is now above the first one, even though the original post had far more interaction and discussion. So, in order to maximise upvoats, you have to say the most commonly agreeable things, appeal to the lowest common denominator as it were. In this way, controversial or challenging discussion is avoided.

Neither of these examples even account for those sites algorithms, selecting which retweets actually show up on your feed for example. The reasult could be called a type of "dead internet", because really, you never even get to the internet, you are trapped in your room. If you liked that, you'll love this. The internet may as well be empty.

Is The Dead Internet Theory Real?

There are bots out there, sure, but the theory does not describe the internet of today, let alone in 2021. Social media sites have always taken measures to block spam bots and still do, even as the bots are evolving, aided by generative AI.

At the moment, generative AI is not capable of creating good content by itself, simply because AI cannot understand context. The vast majority of posts that go viral—unhinged opinions, witticisms, astute observations, reframing of the familiar in a new context—are not AI-generated.

The internet might feel boring, broken, spammy and algorithmic, but we are not drifting alone in a sea of electronic NPCs. Other than reposting content made by people, bots don't lead the internet in the way the theory suggests—influencers do, and the bots follow their lead.

The weird, witty commentary, willful misinterpretations, personal attacks and unhinged opinions that fuel online discourse are still flowing from human users. But the AI-generated garbage that surrounds it appears to be increasing.

There are points in the "ur-text" that have truth to them, and have only become more relevant in the years since. For example, algorithms do dictate our browsing experience, and can make (or break) viral posts.

The internet of today is much more sterile than the wild, unpredictable internet of the past, as the diverse ecosystem of small, user-created sites was replaced by a handful of huge platforms built by large corporations who seek to monetize our browsing and sharing, often to the detriment of user experience.

The internet of today feels far more restricted and corporate than it ever has. Even Tim Berners-Lee, the inventor of the World Wide Web, is disappointed with the state of his creation, stating: "The Web is not the Web we wanted in every respect."

There are still interesting, funny things happening online all the time, but the good stuff is becoming increasingly harder to find, and trends are blurring into marketing campaigns—like the Stanley cup, and even the Grimace Shake.

The Dead Internet Theory might not reflect the reality of the average browsing experience, but it does describe the feeling of boredom and alienation that can accompany it.

Is there an escape?

Not unless we stopped using technologies altogether, but we sure need it. Where do we draw the line?

I don't have solid answers, but we can discuss in comments.