S.B.S

402 replies from @koshercockney. Only 11 unique replies. The same reply was posted 102 times in a single day. Every other repeated reply was posted in under 55 minutes. t.co/r4oreSFAqy

Wait, so let me get this straight: • March: Loomer says an unnamed CIA source claims Turkey is paying unnamed podcasters. No names, no documents. • July: A FARA filing comes out listing Kiriakou and “Neil Patel.” It does not mention Tucker, TCN or a payment to either man. • August: Fischberger says Patel is probably TCN’s co-founder, puts those contacts next to the April Tucker and Kiriakou episode, and Loomer then brings Tucker and Patel into the allegation. So Loomer gives Fischberger an unnamed claim, Fischberger gives Loomer a Tucker angle, and then the Tucker angle gets treated like support for the unnamed claim? Am I missing the receipt, or is the receipt just the two posts pointing at each other?

For how long? 👀 t.co/Ash5Dj7Ruo

That was the last time he replied to anyone. This account has been on autopilot for a whole month. 😂 t.co/58wvM5G5UF

A @StopAntisemites complaint to your boss might look genuine, but the sender never wrote a word. Some of its posts hand out prewritten complaints as one tap links. Recipient, subject, body already filled in. Sent from a real email. The words aren't theirs. Here's one example t.co/KKvuFqaKFF

The Four Teletubbies of the Apocalypse t.co/wOszJIgPEZ

After depriving 400 people of their livelihood she said “We’re not in the business of ruining people’s lives… Our goal is to remedy a situation first.” t.co/GvHpeI0yPQ

Here are 400 posts from StopAntisemitism. Red highlights personal labels, accusations, and calls for employers to act against the people named. Those named include nurses, teachers, students, engineers, and other workers. t.co/tHs6z4x27d

Trapped in a car with her dead family, 5 year old Hind Rajab begged a rescue line while bullets rained on her. His posts, for over a year: Hamas did it, PIJ did it, crossfire did it, her own foundation faked it. Anyone but the IDF. No, Max. It was the IDF that killed her. t.co/Z94tIqzfnk

Your own footnote is the Community Note. 54% = 29% confirmed foreign + 24% "unknown," mostly Americans on laptops per your fine print. Your rule confirms a web-only foreigner but calls a web-only American "unknown." Your largest single foreign origin: Israel, 27 accounts. t.co/hiUsHvCaa1

Someone with this profile picture liked my post. t.co/kUERqtRtVn

What is happening... t.co/hf50FlecsG

Still waiting for science to explain this block 😭 t.co/sehQ33bKiZ

You ruined the app bro. t.co/FX69UwSVO2

Call out the problems on this platform long enough, and eventually they start treating you like the problem t.co/hu11jomjuy

Here are some of this week's posts against James Fishback and Dan Bilzerian. Florida primary week. 20 accounts, 151 posts, a lot of the same attacks. t.co/xe6xrMIYLh

Here are 288 IDF reprints from @EliAfriatISR. He sources Hamas 32 times: hostages, deal terms, Hamas losing. 0 mentions of Gazan casualties. He is the IDF copy machine. t.co/79YjXPAK0x

Here are 99 of the 379 hate posts recycled by @ImtiazMadmood. Same lines, up to 7 times each, most on the exact :00 second. He writes his pro-Israel posts himself. The hate is a scheduler running copypasta. When will X fix this? t.co/lpc2nc47Qz

Here is a clear ToS violation. @diana_bloom_ posted the exact same reply to Reuters 10 times in 143 seconds. Then she admits the tactic herself: "we flood their comment sections every time." That is platform manipulation. t.co/UBagOl4Szx

Here are 576 of @EliAfriatISR repetitive posts. First pic is the same post 144 times. Each recycled dozens of times, almost all at exactly :00 seconds. Thats not a person, thats a scheduler running copypasta. Combined, these racked up 149.9 million views. When will X fix this? t.co/q8Fzoxxix2

lmao. t.co/ocQH4AbAfZ

Next @VividProwess post, dropping in 1 minutes: pro-Israel content, video attached. Screenshot this.

When Elon open sourced X's algorithm, he said X was "actively seeking criticism to improve fairness and quality of algorithmic recommendations." Here is some. Every reaction has a price: Like +0.5 Reply +5 Block -31.2 Mute -58.8 Report -234 But the code also shows the abuse system is not applied equally. X's AI safety pipeline tags targeted insults with FOSNR_ABUSE_INSULTS, which hides the post from recommendations. Accounts labeled High PageRank are skipped. Their posts are not even scanned. They are automatically immune to X's own detection model. I checked 7,992 of Eitan Fischberger's archived posts. 406 attack named people, "sick freak", "evil witch", "piece of human garbage". No sign any were hidden. They do BETTER than his normal posts. Are accounts like Fischberger classified as High PageRank? Because the code is clear about everyone else: post these exact insults as a regular account, or reply to him in kind, and the model can scan you, label you, and limit your reach. He may never be scanned at all. One rule for some, another for everyone else. Why?

According to X’s publicly released open source code, its copypasta rule is not applied equally. Its duplicate text system can give posts a COPYPASTA_SPAM label. But accounts classified as “High PageRank” are skipped by this rule. Archived posts show: Dr. Maalouf 902 different content sets reused in 1,922 later posts Eyal Yakoby 1,365 different content sets reused in 3,412 later posts VividProwess 294 different content sets reused in 431 later posts within the sample X does not disclose whether these three are classified as High PageRank. What the archives do show is thousands of reruns across hundreds of recurring sets. Are these reruns going undetected, or are they being detected and then skipped because the accounts fall under the High PageRank exception? If so, why?

It was number 6 lol: “Islam taking over the West.” He recycled the video and still got the country wrong. Copypasta with bad info. t.co/whk2zKb61O

Called when his next post would go up. Nailed it. Called the one after that too. Then he blocked me. Hopefully the script's off for now. t.co/o1FmBd8XZ2

lmao. t.co/zq0T51evP7

About 30 minutes from now, Dr. Maalouf will probably make another post.

A certain account will post in the next 15 min. Last drop 1:08pm ET. Same thing all day: short video, two-line caption. The closer will be one of these: 1. “I will never find this normal.” 2. “Sorry, no one can convince me this is normal. It’s medieval.”

AI Community Notes are surprisingly good at recognizing context manipulation and lies. Wonderful addition to X. t.co/eFs2HMkbjp