/Technology

Nobody made an official announcement. There was no headline saying "this job ended today." It was more like a tide going out slowly — one day you look and the water just isn't reaching where it used to, but you can't point to the exact moment it pulled back. That's kind of what happened to a bunch of jobs tied to writing, translation, and content production over the last two or three years. And I think almost nobody talks about it clearly because those roles weren't cut. They just stopped existing. Nobody fires a freelancer who was never hired in the first place. Nobody posts on LinkedIn that they "didn't renew" a contract that had already ended on its own. Freelance translators doing technical content and subtitling felt it first, because the work is literally what language models do well: take a text in one language, spit out an equivalent in another. Localization agencies that used to hand out batches of files to dozens of translators now use three or four people just to review what the AI already translated. Different job, paid differently, needing a tenth of the manpower it used to. Junior copywriters at agencies had a similar fate, just quieter, because the cut didn't show up as mass layoffs. It showed up as a job that never opened. Agencies still hire copywriters, just fewer of them, and almost nobody comes in anymore as a pure junior writing that generic blog post just to drive traffic. The "10 tips for X" kind of thing — that's literally what a generative model spits out in seconds, so why pay someone to learn on the job doing that. One thing barely anyone mentions is transcriptionists and captioners. It's not just that AI transcribes well, it's that it transcribes fast enough to wipe out any advantage of outsourcing to humans, even in countries with cheap labor. Companies that used to send hours of recordings to teams in India or the Philippines now run everything automatically and only review the parts that came out wrong. And then there's the more hidden case of stock illustrators, who lost income not because AI is technically better, but because it's infinitely cheaper for the kind of generic image most clients actually need — someone smiling in an office, a "success" icon, some random landscape to illustrate an article. Nobody needs art with soul for that, they need visual filler. And that's exactly what the models deliver well. What connects all these cases, I think, is that it's fragmented work, paid per piece, historically invisible in formal employment stats. That's why the public conversation still circles around "AI will replace programmers" or "AI will replace lawyers" — visible, organized professions — while the real cut has been happening for a while now among people who never had a formal contract or a union to complain to. Which leaves an annoying question: if this kind of erosion is hard to measure precisely because it hits informal and freelance work, how many other sectors are already going through the same thing without anyone noticing the pattern yet?