ChatGPT at One Year: What Changed in Translation?

On November 30, 2022, ChatGPT launched. A year on, the prediction that "the translation industry will be disrupted" has met reality: the industry has not disappeared, and the global language services market is still growing — but its workflows, pricing models and talent standards are permanently changed.

The substantive changes of this year make a long list: LLM translation entered mainstream CAT tools and TMS platforms, and "AI draft + human review" became the mainstream production mode for business content; per-word pricing loosened as tiered services and subscriptions rose; post-editing methodology upgraded from "correcting" to "verifying"; prompt engineering entered the translator's skill set; translation programs began rebuilding curricula; and the EU AI Act wrote rules for generative AI.

Equally noteworthy is what did not happen: no mass unemployment of professional translators — on the contrary, demand for reviewing AI output created new roles; no collapse of high-end pricing — expert rates in legal and medical fields rose rather than fell; and no disappearance of clients' need for "accountability for quality" — the more ubiquitous AI becomes, the more precious traceable, accountable human service grows.

The first anniversary's lesson can perhaps be summarized thus: technological revolutions never eliminate professions — only the parts of professions that refuse to evolve. The translation industry's second year of the AI era has begun.