I.

The building has no windows. That was the first thing I noticed when they gave me my key and showed me to the dormitory where I'd be sleeping for the next two weeks. Fourteen bunks in a room that used to be a coal storage area, the walls still stained black from a century of mining, the air smelling of diesel and sweat and something chemical that I couldn't identify. They called it the Data Mine because they built it on top of an abandoned coal mine, and the servers that processed the world's information sat in the dark where the miners used to sit.

My job was simple: replace the cooling fluid every six hours, check the pressure gauges, clean the dust filters, report anything that looked wrong. I did this for twelve hours a day, six days a week, and on the seventh day I slept in a bunk that still smelled like the man who had slept in it before me.

I was forty-three years old when I started this job. Before that, I was a coal miner in the same mine for eighteen years. Before that, I was a coal miner in a different mine for six years. Before that, I was a kid from a town so small it doesn't show up on most maps. The mine closed in twenty eighteen, and when it closed, three hundred families lost their income and the town lost its reason to exist and I lost the only thing I had ever been good at.

Six months later, a recruiter from Ohio showed up at the community college with a presentation about the Data Mine. He said the pay was good, the work was easy, and the company provided housing and health insurance. He said it was a chance to start over. I signed the papers.

The servers were cold to the touch and loud as thunder. They filled the underground chambers—huge black cabinets stacked four high, their indicator lights blinking in patterns that meant nothing to me. I didn't know what data they processed. I didn't care. I changed the fluid, cleaned the filters, and went home to my bunk and drank beer and watched the ceiling.

II.

Old Dan had been here three years. He was the closest thing I had to a friend, which was saying something because I'm not a talkative person and he wasn't either. We communicated in nods and glances and the occasional shared cigarette in the loading bay during break.

One day, about a month after I started, he was standing in the loading bay, smoking, and he said, "You ever look at the data?"

"What data?"

"The stuff the servers process. You can see it on the terminal in the control room. Just a little bit. Nobody stops you."

"I don't want to see it."

"You don't know what you're missing."

"I know exactly what I'm missing. I'm missing a job that doesn't require me to look at anything. That's the deal. I change the fluid, they pay me, I don't look."

Dan laughed. It was not a happy laugh. "You think they pay you not to look? They pay you because you don't look. You think that terminal is for maintenance? It's for watching. And if you watch too much, they'll send you home."

He finished his cigarette and crushed it under his boot. "Torres watched too much. Young guy, twenty-six, data entry. One morning he just started crying at his desk. Silent tears. They let him go the next day. 'Psychological inability to continue employment.' But we both know why."

"Why?"

"Because he saw something he shouldn't have. And he told someone. And telling someone is the mistake. You see something, you keep it. You don't tell anyone. That's the rule. The rule isn't written down. You learn it."

I went back to the server floor and changed the cooling fluid and cleaned the filters and tried not to think about what was happening inside those black cabinets. Two million surveillance points across the country, feeding data into machines that sat in the dark beneath Ohio, processing the information of living people—their movements, their conversations, their purchases, their searches. Not the grand data of governments and corporations. The small data. The intimate data. The data that tells you who your neighbour is sleeping with, whether your boss is embezzling, whether your daughter is depressed.

I tried not to think about it. I failed.

III.

I looked at the terminal on a Sunday night, when the night shift was thin and the control room was mostly empty. The screen showed a dashboard with scrolling feeds: transaction records, location pings, communication metadata. I didn't know how to read it, but I knew enough to know that it was real.

I typed in my employee ID and hit enter. The system showed me a list of names—employees, I assumed—and next to each name, a category. Most said "standard." One said "elevated."

I clicked on "elevated."

The screen filled with information. Not about me—about the people around me. Old Dan: two DUIs, a bankruptcy, a daughter in Cleveland who sent him money every month and called him once a week and never asked why he was working in a mine that wasn't a mine anymore. Councilman Henderson: three payments from a construction company that had received city contracts without bidding. Mrs. Gable, who lived two doors down from my apartment in the worker housing: prescription records for antidepressants, pharmacy visits at 2 AM, a therapist in Youngstown who specialised in grief counselling.

And my own name.

I stared at the screen for a long time. My file was thin—four pages, nothing dramatic. Employment history, tax records, a traffic violation from 2012. But on page four, there was a section labelled "family health data" and inside it, a single line: Donovan, [redacted], Major Depressive Disorder, diagnosis date 2023-11-14, treatment status active.

My daughter. She had gone to a clinic in Cleveland six months ago. She had not told me. She had called me from a number I didn't recognise and said she was fine and she was going to be okay and her voice sounded like it was made of glass and if I cracked it, it would shatter.

I sat in the control room and stared at the screen until my eyes burned. I thought about what Dan had said: the truth is not a single thing. The truth is a million tiny things, all of them true, all of them terrible.

I closed the terminal. I went back to the server floor and changed the cooling fluid.

IV.

The next morning, I woke up at six, showered in water that was too cold, ate breakfast in the cafeteria with people who didn't look at each other, and went back to the server floor. I changed the fluid. I cleaned the filters. I checked the pressure gauges. I did exactly what I had been hired to do.

At five in the evening, I clocked out, drove my truck to a bar in the nearest town—there wasn't much else to do—and ordered a whiskey. I drank it slowly, watching the condensation form on the glass, thinking about nothing.

When I drove home, the Data Mine was visible from the road: a row of low, windowless buildings sitting on the hill where the coal mine used to be, their ventilation stacks releasing plumes of warm air into the cold night, their security lights casting long shadows across the empty parking lot. Inside those buildings, machines were processing the secrets of millions of people, and nobody was looking.

I parked my truck, walked to my bunk, and lay down on the thin mattress. The man who had slept in the bunk before me was gone now—transferred or fired or broken, I didn't know which. His name tag was still on the hook by the door. I took it down and threw it in the trash.

Outside, the servers hummed. Inside, I closed my eyes and listened to them.

The truth was down there, in the dark, processing itself in endless, silent loops. And I was up here, in the light, drinking whiskey and pretending I didn't know what it contained.

That was my choice. Not a heroic choice. Not a brave one. Just a choice.

The servers hummed. I drank my whiskey. I went to sleep.

---END_OF_STORY---

Objective Tensor Coding System v2 (OTMES) — Code: DIRTY-REL-20260528-005 作品变体编码: The Data Mine (V-05 Dirty Realism) ------------------------------------------------------------ MDTEM Parameters: V (Destruction Value): 0.30 I (Irreversibility): 0.40 C (Innocence/Suffering): 0.70 S (Scope): 0.10 (individual) R (Redemption): 0.00 TI (Tragedy Index): 42.1 TI Level: T5 Suffering

Tensor Coordinates: M1_Tragedy: 6.0 | M2_Comedy: 0.5 | M3_Satire: 5.0 M4_Poetic: 3.5 | M5_Scheming: 3.0 | M6_Suspense: 3.0 M7_Horror: 1.0 | M8_SciFi: 2.0 | M9_Romance: 0.5 | M10_Epic: 1.0 N1_Active: 0.20 | N2_Passive: 0.80 K1_Individual: 0.80 | K2_Collective: 0.20

Direction Angle: theta = 180.0° (Cold Objective 冷峻客观) Style: Dirty Realism (风格E) Narrative: First-person male (unnamed narrator) Word Count: ~1290 OTMES Hash: DIRTY-REL-20260528-005-F1B7