Mark's digital junkyard: how Facebook became a factory of AI nightmares

If you haven't visited Facebook in a long time, you may have preserved some remnants of mental peace. What was once a platform for sharing photos of grandchildren and news from former classmates has transformed into an endless feverish dream created by neural networks. Users are increasingly encountering content that the internet aptly dubbed "AI slop" — low-grade garbage generated by artificial intelligence, ranging from ridiculous to genuinely disturbing. And the problem isn't that AI draws poorly, but that Meta's algorithms literally force you to look at it.

It all started with relatively harmless, albeit strange images like "shrimp Jesus" or giant watermelon sculptures. However, over recent months, the absurdity has escalated. Now the feed is filled with images of mutilated soldiers, children with extra limbs in the mud, or surreal monsters being passed off as "miracles of nature." Ironically, Meta, which invests billions in developing Llama and advanced visual models, has become the largest distributor of the cheapest and dirtiest digital content. This happens because the system rewards engagement at any cost. The more shocking or strange an image is, the more likely someone will leave a comment — whether it's outrage from a real person or an automatic bot response.

This phenomenon vividly demonstrates the "dead internet" theory in action. We're observing a closed ecosystem: neural networks generate horrifying pictures, bot farms like and comment enthusiastically with "Amen!", and Facebook's algorithm, seeing such activity, pushes the post even higher. Real users in this scheme become mere random witnesses, on whom Meta tests its ability to hold attention. For the company, it's business: while you're scrolling through the feed trying to figure out why this child has eight fingers, you're shown advertisements. The quality of the content itself is secondary to the advertising machine.

The situation becomes dangerous when this AI garbage begins to exploit human empathy. Scammers use images of "sick children" or "destroyed cities" to collect donations or drive traffic to dubious websites. Meta takes the position of a neutral observer. Despite all statements about "safe AI" and labeling of generated content, the moderation system is clearly unable to cope with the scale of the problem. In pursuit of engagement numbers, the company has created a monster that devours the remnants of common sense on the platform.

What does this mean for the industry as a whole? We're witnessing how the world's largest social network degrades into a state of content landfill. If we once feared that AI would replace artists, now we should fear that it will finally bury the ability to find something genuine on the web. Facebook is becoming a mirror in which not our friends are reflected, but the darkest and most absurd hallucinations of neural networks, approved by a soulless algorithm for an extra penny of profit.

Bottom line: Meta has created the perfect machine for producing digital noise, and there seems to be no way back. Are we ready for the entire internet to soon look like a bad neural network trip?