Being human helps': despite rise of AI is there still hope for Europe's translators?

"He would test whether AI could put him out of work. Gentric had been grappling with a short non-verbal sentence that described the book's protagonist's feelings upon opening a window: Bright, sharp night air, bracing. He put the prompt into DeepL, a neural-network-powered machine translation engine that regularly outperforms Google Translate in accuracy assessments. The proposed translation was reassuring, with his job security in mind: L'air de la nuit, vif et vif, etait vivifiant (The night air, lively and lively, was enlivening.) AI had translated the sentence's meaning but was seemingly unaware that the repetitions rendered the line absurd."

"He repeated his experiment this spring, however, the outcome made him feel less at ease: L'air nocturne etait vif, pur et vivifiant, DeepL suggested this time. The online translator still lost the sentence's stylistic trait by adding a verb, but it had learned to use three different words that even had a musical ring to them. I don't know if it's just chance or a fine-tuned algorithm at work, but nocturne and pur is not bad, said Gentric."

"Chatbots running on large language models (LLMs) neural networks trained on vast amounts of text to generate natural-sounding language are rapidly infiltrating every aspect of our work and leisure lives. But few professional sectors are being disrupted by the technology as rapidly as the translation industry in Europe, home to more than 200 languages and a booming tech sector."

A literary translator tested whether AI could replace his work by translating a short non-verbal sentence describing a protagonist’s feelings when opening a window. An initial DeepL output captured the meaning but repeated words in a way that made the line absurd, producing an inferior result compared with his later published translation. He repeated similar tests in later years and found different outputs, showing inconsistency. A newer attempt still added a verb and missed a stylistic trait, but it used different word choices that sounded more musical. Large language model chatbots are spreading quickly across work and leisure, with European translation facing rapid disruption due to many languages and a strong tech sector.

Read at www.theguardian.com