A small reflection on the Epstein files, tied to artificial intelligence, otherwise you’ll yell at me when I go off topic. I’m following the story, I’m very interested. It’s a big, ugly thing with a lot of implications, but I keep asking myself one question: now that the American side of TikTok has passed to American investors, why is all this deep-dive Epstein content still circulating on TikTok? Why don’t they block it?
Third-millennium censorship is probabilistic
One thing that’s clear about TikTok and other mass social platforms like it is that, for them, the recommendation engine is something controllable, steerable. They can selectively suppress content that touches even remotely on a given topic.
You’ve surely seen plenty of creators using words like “unalived” instead of “killed,” or saying “didn’t make it” instead of saying someone died. Why? Because anything to do with violence and death gets pushed less by the algorithm. And I want you to notice: they’re not so stupid as to censor content outright. They simply make it less probable. It’s a bit more nuanced than you’d expect. The censorship of the third millennium is a probabilistic censorship, not a clean-cut one.
So why Epstein and not ICE?
Anyway, I’m digressing. The Epstein files: why is all this content circulating even on the American side of TikTok, while the content about ICE has already been completely sedated?
I’d think hard about this. Because yes, it sounds paranoid, but no, it isn’t paranoid. The Western side of TikTok selectively chose to let the Epstein content circulate and stopped circulating the ICE material. As for why, I honestly don’t know. But it’s worth sitting with the question.
