(CNN now has a team reporting from Tehran: https://www.cnn.com/2026/03/08/world/video/fred-pleitgen-sha... )
14,726 karma · joined June 25, 2013
https://web.archive.org/web/20260302041747/https://www.bbc.c...
The NYT had coverage the day of.
You can see it on the front page in their "Live" coverage at the time:
I do think it was probably relatively underplayed in Western press, you can eg see it on the front page day-of pretty prominently on Al Jazeera:
https://web.archive.org/web/20260228182044/https://www.aljaz...
But the western press dragging their feet at covering victims of America's foreign wars is not down to being pro-Israel, no.
The reason why isn't really a mystery: Iran has never been exactly welcoming to Western media, and internet access there was intentionally shut off after the recent protests. There's plenty of coverage- it's front page everywhere- but a paucity of information.
It's all over social media, but hardly any of that is from Iranians in Iran, it's just people outside it like you and me mostly just yapping. Occasionally you'll hear something second-hand from someone with family in Iran who managed some brief connectivity.
> The military’s Maven Smart System, which is built by data mining company Palantir, is generating insights from an astonishing amount of classified data from satellites, surveillance and other intelligence, helping provide real-time targeting and target prioritization to military operations in Iran, according to three people familiar with the system...
> As planning for a potential strike in Iran was underway, Maven, powered by Claude, suggested hundreds of targets, issued precise location coordinates, and prioritized those targets according to importance, said two of the people.
Just read the first two paragraphs and there's a full stylistic whiplash. I'd bet $100 that the first paragraph was human-written and the rest were almost fully Claude.
It's kind of creepy, like a human rolling their eyes back partway through a paragraph and suddenly speaking like Microsoft Bob.
The big one it missed: the headers are mostly "The [Noun:0.9|Adjective:0.1] [Noun]". LLMs (maybe just Claude?) love these. Every heading sounds like it could be a Robert Ludlum novel (The Listicle Instinct, The Empathy Performance, The Prometheus Deception).
A more robust and perhaps more vivid indicator are sentences like this: "This isn’t a proof of concept or a weekend project. It’s a real authoring environment."
"It's not X, it's Y" is an LLM tell. "It's a real Z" is another. Together? I'm going to conclude it's LLM generated to like a 90% certainty.
And as the sibling notes, the icons look like LLM SVG output. They're more mangled than even a rushed human would do.
Again, it's fine. If I had more time I'd love to try to vibecode a Flash clone.
I think both sides of this end up proving "too much" in their respective directions.
1) this style genuinely is preferred by lots of people on X/Twitter so you might as well lean into it
2) People who spend a lot of time with LLMs think this sort of writing is normal or even standard just through overexposure, a sort of pseudo social proof
2b) People who spend a lot of time with other people who use LLMs think this is how humans write (actual social proof)
3) People are insecure about their writing ability and find the non-judgmental non-human LLM editor soothing
4) people are lazy
5) people aren't lazy per se but they know writing has been so devalued that they aren't going to spend time on it that they don't need to
6) their first experience of writing was trying to hit word count requirements in grade school and that stuck
7) Visibly using LLMs is becoming a shibboleth for a social group on Twitter and LinkedIn. It's a marker that you are dogfooding the crappy AI tools you're developing and selling. Under this theory, being visibly LLM output is actually intentional: "look ma, no hands- all NanoClaw!"
Instead it seems like there's a solid core of people who have always wanted to outsource their brains entirely to machines, and have finally got their wish.
I'm old enough to remember when we joked about normies who were dumb enough to let computers think for them.
Sorry, I'll put it in hand-crafted ChatGPTese:
## The Slop Problem
Every post sounds the same. No intelligence. No individuality. Just pure, clean LLM slop. Let's dive in.
- Every post has LLM tells. This is key.
- Posts get upvoted anyway. Nobody seems to notice or indeed care.
- People acclimate to the slop. This isn't just a coincidence. This is a real shift in standards. When people read enough of this, they begin to think it sounds normal.
## The Replying Dilemma
Should you engage with the content, when there is a real person involved? On the one hand, they put their name on it, and probably the details are drawn from their prompt, so it can be said to fairly represent what they wanted to say. So maybe ragging on their ChatGPT prose is being mean. On the other hand, if nobody ever mentions this, the acclimatization will only get worse as the rising tide of slop overwhelms any other style of writing.
## The "Snobbery is good actually" Option
Relentlessly bully people for their half-baked LLM copy. Make it your whole personality. Go insane.
## The "Giving Up" Solution
Learn to stop worrying and love the LLM.
I'm not even against using AI per se, but when something is obviously written in ChatGPTese I'm not going to read it if I don't have to.