Posted by Vincent_Yan404 1 day ago
I grew up in "Factory 404," a secret nuclear industrial city in the Gobi Desert that officially didn't exist on public maps. This is a memoir about my childhood there.
It was a surreal place: we had elite scientists living next to laborers, a zoo in the middle of the desert, and distinct "communist" welfare, all hidden behind a classified code.
This is Part 1 of the story. I'm happy to answer any questions about life in a Chinese nuclear base!
You are right—the generation that built '404' is aging, and many of their stories are fading into silence. One of my primary motivations for writing this was the realization that if I didn't document these memories now, they might be lost forever.
I hope my first-hand account can provide a more nuanced, human layer to the historical data you've gathered. There is so much more to tell beyond the official records.
It obvious that something was used to do the translation, but it doesn't feel worse than any other machine-translated texts, as long as I get the gist and the overall idea of what the author is trying to communicate, I feel like it's good enough.
What "dominant Western narratives" apply here? I'm not going to bicker. I'm just curious.
And like the person said, there is nothing inherently wrong with such a narrative. Like them I'm also curious about non-western narratives.
If most groups, cultures, religions, countries were more curious about "non-native" stories, maybe we'd all be a bit more open-minded and understanding.
Your initial assertion that people online criticizing the West are "often" criticizing "absurd" things is simultaneously wrong and condescending, some sort of thought-terminating cliché.
I was bracing myself for something edgier.
I don't know what I expected, "China has freer speech than America because Facebook censored antivax content" or something :)
So I'm not sure that specific comment would be considered to be a "dominant western narrative" unless you're going to tell me that older (and so who have lived through it) educated people in China who don't speak a word of English have a western mindset because they're educated.
Accounts from well-off diaspora of any country will always be negative. It’s a self-selecting group with specific interests.
EDIT: By the way, it's not that hard either to find books written by Chinese writers not part of the diaspora that are critical of the cultural revolution (Serve the people by Yan Lianke, 3 body problem by Liu Cixin) or the great leap forward (4 books by Yan Lianke). Obviously, writers living in China that have to deal with censorship tend to be less directly critical of it compared to writers from the diaspora but that doesn't stop some criticism to shine through.a
I was curious about the 'narratives' it mentioned.
They might be wrongheaded; they might be valid.
Either way, it piques my interest.
This is extremely manipulative. The only reasons to say something like this are to shame the person you're respond to and/or attack and discredit them and force them to respond defensively. Don't do this.
(it also immediately outs you as not having any valid points to make, because someone with a reasonable response doesn't need to stoop to emotional attacks)
I have some very good friends which are Chinese but are not able to read English, do you mind if I do a AI translation, and if you can check it to see if it translate what you're trying to convey ? (I propose that as I think it would be too much to ask to ask to redo the text in Chinese)
Edit: haha I see you actually did the reverse ! Do you mind sharing also the original CHinese script ? That would also help me with my own mandarin learning !
I did write and publish this story in Chinese first. You don't need an AI translation for them; the original text exists and has been quite popular in the Chinese corner of the internet.You can search for it using the title:《我在404长大》
I definitely appreciate the style of the HN English article, but I think the browser-translated version possibly gives a bit more context to some of the story.
e.g. This is the English version "We would clutch candy wrappers in our hands, giggling endlessly. The teacher would scold us for disturbing the nap, but we Hid behind our parents, still laughing."
This is the browser-translated version: "I kept giggling when I saw her, and she giggled too, and we kept laughing with small sugar paper during our lunch break. When my parents came to pick us up, the teacher criticized us for being undisciplined, and we still hid behind our parents and giggled."
We had secrecy education (保密教育) starting as early as primary school. We were taught from a very young age that our city didn't exist to the outside world, and we simply didn't talk about it. But when I was a kid ,I didn't know anything about 404.
The term that has no connotations of fiction is probably "narrative".
I think many languages have closely related words for fictional narratives and nonfictional narratives.
E.g. if you said someone was good at "storytelling" as a skill, then I would expect it to be most likely fictional. I agree that "tell you the story of..." could easily be nonfictional.
https://chaiwanbenpost.net/article/%25E4%25B8%25AD%25E5%259C...
Did you write the exact same paragraphs by chance, were the two blog posts a collaborative effort, did you get together and pooled your recollections, or what?
The author you mentioned is Li Yang. We know each other, and our parents know each other as well. He published his piece before I did. Since the person involved was his classmate, he was able to provide more first-hand details, such as the part about riding a bicycle to see that boy.
When we had safety education at school, the teachers would still use examples from twenty years ago—like someone getting hit by a car. This is how it was in the plant: once something happened, people would keep talking about it for twenty years.
Soviet Architecture: Many of our residential and administrative buildings were designed and built by Soviet experts, giving the city a distinct 'Stalinist empire' aesthetic that felt very grand compared to the surrounding desert.
Elite Salaries: The wage levels in our factory were on par with those in Beijing, which was extraordinary given our remote location.
The 'Post-Scarcity' Bubble: For many families, daily expenses were minimal because the 'unit' (Danwei) provided almost everything. We regularly received rations of high-quality rice, flour, and oil as part of our work benefits, so we rarely had to spend money on basic survival.
In a country that was still struggling with scarcity, living in 404 felt like living in a futuristic, well-provisioned fortress. Stay tuned for Part 2, where I'll talk more about this 'gilded' lifestyle.
I'd be very interested to hear any thoughts you might have about Jung Chang's book "Wild Swans".
I read this book a year or two ago and learned a lot from it, but I also learned that many people who grew up in China take issue with the author's account. I'd be grateful for any remarks you may be able to share.
It’s true that many people who grew up in China have a complicated relationship with narratives that focus on negative historical periods. There is often a defensive reaction, a feeling that such stories are 'smearing' the country's image.
However, as a writer, I believe that truth is always more important than a curated image. Authentic memories are often scarce, precisely because they are difficult to tell. My goal with the '404' series is to provide a piece of that missing truth—not to judge, but to document a reality that actually existed. In the long run, I believe a society is better served by facing its complex past than by forgetting it.
"Facing a complex past" is a big theme in Germany, too, of course, and I think it's the only proper way to deal with it. Direct witness accounts and retelling are important and add something that a dry history book can't provide. Keep up the good work!
Can you expand? A code under what system? What were some other code numbers and what (unclassified) things did they refer to? Did each code refer to a specific city or specific factory? Or were all cities/factories dedicated to a certain type of industry or military objective classified under the same code? Why did they teach you this code number growing up?
I'm really fascinated by this. Fantastic story overall, can't wait for part 2!
The existence of such a large and conspicuous secret might seem bizarre to the post-cold-war mind, but it was fairly common in the West too. For example, the British Telecom Tower in central London stands at 189 metres tall and had a revolving restaurant that was open to the public, but was also a designated site under the Official Secrets Act.
I'm interested in the laborers who did the work, not just the scientists who designed everything.
Thanks for your story.
Edit: And what a great read, thank you!
The coincidence with the HTTP error code is purely accidental, yet incredibly poetic—because for decades, this city literally could not be found on any public map.
I see even China hasn't tested in decades and so my chances of doing this are close to nil, but I ask because your answer could tell more how you feel about the technology and its future. My physics professor told me to study supernovae instead.
My perspective on 'the nuclear' is purely emotional and sensory—I simply find it terrifying. I resonate much more with the raw, human suffering described in Svetlana Alexievich’s Voices from Chernobyl than I do with the scientific future of nuclear power.
I was curious about this part and lingering perspectives among Chinese citizens. How do they regard the past mass starvations and deaths in the 1900s? Are these events well known? Are they seen as a catastrophe? Do they blame someone (like the government) or is it seen as the cost of progress or a natural disaster? Do old and young people see these events differently from each other?
Thanks for writing and sharing!
I remember when I was 4 or 5 years old, my mother told me stories about those years. As a child, I didn't understand the historical context; I thought mass starvation was something that happened cyclically, like the seasons. I vividly remember asking her: 'Does this happen every few years? Should we start stockpiling food now just in case?'
Even today, you will see older Chinese people who cannot bear to see a single grain of rice left on a plate. It’s not just frugality; it’s a ghost from 1959.
A common criticism of Chinese people is that they 'eat everything,' but a major reason for this is that China has endured more famines than almost any other nation in human history.
It's definitional in Gwh of productive, usable energy produced per tonne of damage to health. It also demands a lot of rigour against other forms of embedded energy regarding fugitive gas leaks, unassociated third party injury which is usually an externality. And of course it predates the rise in general efficiency of solar and wind and may no longer be true unless very specific criteria are applied like constancy.
But, awful though the trail of tears is behind example contamination events, including Chernobyl and Three Mile Island and Fukishima, counting death in coal or oil demands recognition of a huge problem in life shortening from contamination and injury at large in the whole cycle mine to chimney.
More people died from translocation consequences than direct nuclear radiation consequences in all three of the above. Not to minimise their deaths but if you move a million people in a rush, some die who otherwise would have lived.
"Modern" here is > 1949 and < "whenever wind and solar and batteries got so good"
Visiting my grandparents I remember we had to go through a sort of border control to get there.
My mom told stories of how the government would change the asphalt every year in that city to cover the nuclear dust.
I don't mind AI translation at all. The style comes off as a bit weird indeed, but I just took it off as a style I'm not used to because it comes from a different culture than mine. I wouldn't mind much the naysayers, I'd like to see them posting something in chinese and see how they'd like it ;)
I really enjoyed the writing style actually, all these different anectodes condensed in shorter sentences, without fluff or trying to connect them in a single narrative. Maybe this is not the correct way to put it, but I'm also not a native English speaker nor I have any classical training in writing.
Yours is the first substack I ever subscribed to and can't wait to read part two. It actually pushed me to start writing some of my childhood experiences.
Thank you again. Absolutely fascinating.
Unfortunately the ever-present desire for the moneys made folks use LLMs to produce lots and lots of slop, polluting not just the web but even the trust to each other. The default nowadays when reading a piece of text that has even the slightest LLM vibe is to assume it's made-up slop. That's very sad, but necessary, because it's just everywhere.
It's so sad because the tech could really bring people together. Creating almost seemless translations. That's why your work is such a great example for the good this could bring if we'd not have so many greedy people among us.
Line 418 was the most profitable. When the post office opened, the clerk assumed “418” was the town name, not the factory line number. By the time anyone noticed, mail was flowing, checks were signed, and no one wanted to correct the federal government. The factory closed in the 1950s. The town shrank but remained oddly proud of its name. Residents leaned into it without explaining it.
The name origination is however much less interesting but still entertaining
“Eighty Four was originally named Smithville. Due to postal confusion with another town of the same name, its name was changed to "Eighty Four" on July 28, 1884. The origin of the name is uncertain. It has been suggested that the town was named in honor of Grover Cleveland's 1884 election as President of the United States, but that occurred after the town was named. Another possibility is the town's mile marker on the Baltimore and Ohio Railroad. Another is that the town was named after the year the town's post office was built, by a postmaster who "didn't have a whole lot of imagination."
edit: 418.. I've been had.
April 4th is an informal city holiday, "404 day".
Lots of artists and companies make "404" branded stuff, and you generally see the number all over the city:
https://mondaynightbrewing.com/beer/404-atlanta-lager/
https://sneakernews.com/2025/03/21/adidas-superstar-404-day-...
I was extra surprised that there's a chain of crepe places called "Atlanta Crazy Crepes" which as far as I can tell are located only in Akihabara in Tokyo and Eastridge Mall in SJ.
It’s fascinating how industrial logic can accidentally become a place's identity, whether it’s a production line in Pennsylvania or a secret code in the Gobi Desert. The fact that residents remained 'oddly proud' of a name that was essentially a clerical error resonates deeply with me.
In 404, our pride was tied to a secret mission; in 418, it was tied to a factory's success. Both show how humans can find a sense of home and belonging in the most 'functional' or even 'accidental' labels. This is exactly the kind of connection I hoped this post would spark.
I promise that the story of 404 is my own, lived experience. Just as the whole article, they are translate by LLM.
My IELTS score is 7.5, but my writing band is 6.0.
I write my thoughts and comments in Chinese first and then use AI to translate them. The entire article was also translated from my original Chinese manuscript.
Thank you very much for the article, it was super interesting. The mystery in the story draws people in, and people surely won't mind a couple of grammatical mistakes. But you have to watch out: the use of AI makes it easy for people to suspect that the story might've been embellished. For the second part, it might be better to try translating it manually; the same goes for writing replies.
Thank you for sharing your story. It makes the world a better place.
Low-level English is normal and accepted.
On the job, I never speak above the level of a 13-year-old.
AI generated English is hated.
Consider using English, not software translation.
I've been seeing this take on HN a lot recently, but when it comes to translation current AI is far, far superior to what we had previously with Google Translate, etc.
If the substack was written in broken English there's no way it would even be appearing on the front page here, even less so if it was written in Chinese.
Of course, even this can be faked, sadly.
That's why translation is a job in the first place and you don't see publishers running whole books through Google translate. No one, least the authors, would accept that.
Contrast this with the faux polite, irritating tone of the AI, complete with fabrications and phrases the author didn't even intend to write.
Authenticity has value. AI speech is anything but authentic.
The author acknowledged they used AI to translate. Is the translation they decided to publish among the given tools they had available to them not by definition the most authentic and intentional piece that exists?
All of this aside, how do you think tools like Google Translate even work? Language isn't a lookup table with a 1:1 mapping. Even these other translation tools that are being suggested still incorporate AI. Should the author manually look up words in dictionaries and translate word by word, when dictionaries themselves are notoriously politicized and policed, too?
Maybe or, most likely this is the same for writing: there are people that think correct grammar and punctuation and no help on achieving this, means writing.
> This can avoid the taste of AI, but it may be very bad to read, I first used machine translation translation, many parts become very wordy, and at the same time puzzling.
Perfectly clear and comprehensible. It's not fluent English, there are comma splices everywhere, and it translated "machine translation翻译" as "machine translation translation", but I understand it – and I'm confident it's close to what you actually meant to say. I can spot-check with my Chinese-to-English dictionary, and it seems like a slightly-better-than-literal translation. My understanding of your comment:
> This can avoid the smell of AI, but it may be a struggle to read. I initially used a dedicated machine translation system, but many parts became verbose (/ very wordy) and incomprehensible.
Generative models don't solve the 令人费解 problem: they just paper over it. If a machine translation is incomprehensible, that means the model did not understand what you were saying. Generative models are still transformer models: they're not going to magically have greater powers of comprehension than the dedicated translation model does. But they are trained and fine-tuned to pretend that they know what they're talking about. Is it better for information to be conspicuously lost in translation, or silently lost in translation?
Please, be willing to write in your native language, with your own words, and then provide us with either the original text, or a faithful translation of those words. Do you really want future historians to have to figure out which parts of this you wrote yourself, and which parts were invented by the AI model? I suspect that is not the reason you wrote this.
The core algorithm behind modern generative AI was developed specifically for translation, the task which arguably these chatbots are the most suited! It’s the task that they’re far the best at, both relative to older translation algorithms (which were also AI), and relative to their capabilities other tasks that they’re being put to. These LLMs are “just” text-to-text transformers! That’s where the name comes from!
“Stop using the best electric power tool, please use the outdated steam powered tool.” is what you’re saying right now.
You’re not even asking for something to be “hand crafted”, you’re just being a luddite.
Indeed! And yet, generative AI systems wire it up as a lossy compression / predictive text model, which discreetly confabulates what it doesn't understand. Why not use a transformer-based model architecture actually designed for translation? I'd much rather the model take a best-guess (which might be useful, or might be nonsense, but will at least be conspicuous nonsense) than substitute a different (less-obviously nonsense) meaning entirely.
Bonus: purpose-built translation models are much smaller, can tractably be run on a CPU, and (since they require less data) can be built from corpora whose authors consented to this use. There's no compelling reason to throw an LLM at the problem, introducing multiple ethical issues and generally pissing off your audience, for a worse result.
Because translation requires a thorough understanding of the source material, essentially up to the level of AGI or close to it. Long-range context matters, short-range context matters, idioms, short-hand, speaker identity, etc... all matters.
Current LLMs do great at this, the older translation algorithms based on "mere" deep learning and/or fancy heuristics fail spectacularly in the most trivial scenarios, except when translating between closely related languages, such as most (but not all) European ones. Dutch to English: Great! Chinese to English: Unusable!
I've been testing modern LLMs on various translation tasks, and they're amazing at it.[1] I've never had any issues with hallucinations or whatever. If anything, I've seen LLMs outperform human translators in several common scenarios!
Don't assume humans don't make mistakes, or that "organic mistakes" are somehow superior or preferred.
[1] If you can't read both the source and destination language, you can gain some confidence by doing multiple runs with multiple frontier models and then having them cross-check each other. Similarly, you can round-trip from a language you do understand, or round-trip back to the source language and have an LLM (not necessarily the same one!) do the checking for you.
It's like writing something with a commodity Bic ballpoint vs a fancy fountain pen with expensive ink. The style of the prose itself is not the valuable artifact, at least not here (it may be in certain places e.g. poems and novels), unless you think well-written/well-spoken people are automatically more veritable or intelligent, which is just as shallow as lookism.
The witch-hunt style comments where people accuse an author of using LLMs as if it's some big gotcha that discredits everything they said need to stop. It only derails the discussion.
I think this simplifies the entire discipline of literary criticism and I suppose every other related science. You can write the same prose with both the Bic and the fountain pen; the quality of pen only affects the material quality of the writing—the ink—but not the style (rhetoric? eloquence?) of the writing (i.e., the contents of language, how it’s conveyed, etc.). We aren’t arguing whether it’s appropriate to depreciate writing generated by an LLM to using speech-to-text as opposed to using a keyboard.
The style of the prose does contribute to the value of the artifact and speaks to the repute of the reader in addition. Readers care about what you say and how you say it too.
Nonetheless I as well as others have good reason to interrogate the intrinsic value of LLM-assisted writing especially when it refers to writing like the one being discussed which I reckon qualifies as a part of the “literary non-fiction” genre. So it’s apt that we criticize this writing on those grounds. Many here have even said that they would prefer the 100% genuinely-styled version of the author’s experience which is apparently only 1.5 points lower than whatever their verbal acumen is. [1] Which I imagine places them around the rank of your average American...and I assume so with charity toward the Americans.
While I think some LLM accusations are lazily applied I think communities such as this one benefit from these discussions when waged critically. Especially when status and social capital are of implicit interest.