One has to admit, that went quick

I am also concerned of AI-slop, so thanks for sharing. However, I do not think the headline here does justice to the article. There is a more interesting message in there. First of all, the study concerned new content, not all of the internet. Second, and more important, the researchers seemed to have found some kind of a saturation point for AI slop, where it’s share of new content was not rising anymore. I think it would be crucial to understand the mechanisms behind this current limitation of AI slop, and whether they could be utilized to combat AI slop in the future.

Not surprising since over 50 percent of internet traffic is bots. Dead internet theory is REAL.

Also, how long until we see model collapse due to ingesting more AI generated content than real content?

Luckily we are already used to the bots.

Can we at least get a progress bar or something?