Mark's Digital Dumping Ground: How Facebook Became a Factory for AI Nightmares

If you haven't logged into Facebook in a while, you may have kept the remains of your peace of mind. What was once a platform for sharing photos of grandchildren and news from old classmates has become an endless fever dream produced by neural networks. Users increasingly run into content that the internet has aptly dubbed "AI slop": low-grade junk generated by artificial intelligence, ranging from ridiculous to downright frightening. And the problem is not that AI draws badly. It is that Meta's algorithms practically force you to look at it.

It all began with relatively harmless, if strange, images such as "shrimp Jesus" or giant sculptures made of watermelons. Over the past few months, however, the level of absurdity has risen. Now the feed is packed with images of mutilated soldiers, children with extra limbs sitting in the mud, and surreal monsters passed off as "wonders of nature."

Ironically, Meta, which is investing billions in Llama and cutting-edge 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 is to leave a comment, whether it's the outrage of a real person or an automated bot reply.

This phenomenon is a vivid demonstration of the "dead internet" theory in action. It is a closed ecosystem: neural networks generate creepy pictures, bot farms like them and post rapturous "Amen!" comments, and Facebook's algorithm, seeing that activity, pushes the post even higher. Real users in this scheme are merely accidental witnesses, on whom Meta tests its ability to hold attention. For the company, this is business: while you scroll, trying to work out why the child in the photo has eight fingers, you are shown ads. The quality of the content itself is secondary for the ad machine.

The situation becomes dangerous when this AI junk starts exploiting human empathy. Scammers use images of "sick children" or "destroyed cities" to collect donations or to funnel traffic to dubious websites. Meta, meanwhile, plays the role of bystander. Despite all its statements about "safe AI" and labeling of generated content, the moderation system clearly cannot cope with the scale of the problem. In pursuit of engagement numbers, the company has created a monster that is devouring what remains of common sense on the platform.

What does this mean for the industry as a whole? The world's largest social network is degrading into a content dump. We used to fear that AI would replace artists. Now the bigger fear is that it will finally bury any chance of finding something genuine online. Facebook is becoming a mirror that reflects not our friends but the darkest and most absurd hallucinations of neural networks, approved by a soulless algorithm for an extra cent of profit.

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

*Meta has been designated an extremist organization and is banned in Russia.