When Google Translate launched in 2006, professional translators saw their obituaries being written. The narrative was simple: machines would render human translation obsolete within a decade. Two decades later, the translation industry employs more people than ever before, though what they do bears little resemblance to the work of their predecessors.
The great unbundling
Machine translation didn't kill the profession; it unbundled it. Where translators once owned the entire process from source to target text, AI carved the work into discrete specializations. Post-editors clean up machine output. Localization engineers manage translation memory systems. Cultural consultants ensure AI doesn't accidentally insult entire nations. Quality assurance specialists catch the subtle errors machines make when they confidently mistranslate legal contracts or medical instructions.
The economics shifted dramatically. Raw translation—converting words from language A to language B—became nearly free. But the demand for translation exploded as the cost plummeted. Companies that once translated their websites into three languages now target thirty. Legal firms that avoided international cases due to document costs now take them routinely. The pie grew faster than AI could eat it.
The human premium
What AI revealed was that translation was never really about converting words. The highest-paid translators today barely translate in the traditional sense. They're cultural bridges for corporate boards, diplomatic interpreters who navigate unspoken tensions, literary artists who capture an author's voice across linguistic chasms. A senior pharmaceutical translator explained the shift: "AI handles the molecule names perfectly. I make sure the dosage instructions won't kill anyone in Japan."
The stratification is stark. Bottom-tier work—user manuals, routine business correspondence—pays pennies per word to post-editors who clean up AI output. Top-tier work—literary translation, high-stakes legal interpretation, real-time diplomatic service—commands higher fees than before AI. The middle disappeared.
Our take
The translation industry's transformation previews how AI will reshape most knowledge work: not through wholesale replacement but through radical restructuring. The survivors aren't necessarily the most skilled at the old craft; they're the ones who recognized early that their value lay not in the mechanical act of translation but in the judgment, cultural fluency, and specialized expertise that wrapped around it. Every profession convinced AI will automate it away should study what happened to translators. The machines came for their jobs and accidentally created new ones—just not the ones anyone expected.
