How long will humans be needed for translation?
AI dominates the translation business, but for high-stakes work, humans are still needed.
The French expression "Il pleut des cordes" translates to "it's raining ropes," which confuses AI translators. François Yvon, a research director at the French National Centre for Scientific Research, explains that AI can interpret the phrase as "it's raining cats and dogs" in English, but faces challenges when translating into languages like Indonesian.
The use of AI for translation has surged, with Google reporting a trillion words translated per month, according to John Zimmerman, founder of OKA, a Swiss non-profit that translates Wikipedia articles into various languages. For internal documents or product listings, AI translation is often sufficient, but for critical applications such as encyclopedias, legal documents, medical translations, or crucial contracts, AI translation may not be as reliable.
While AI-generated drafts are increasingly accurate, human translators are still required to review and correct errors. A study by OKA found that 27% of AI-generated text still needed correction, with 5 to 6% of errors altering the original text's meaning. Legal expert Ariella Baltié highlights the importance of accuracy in legal translations, as mistakes can have serious consequences.
In medical translations, errors can be life-threatening, and AI translators have been known to misinterpret Greek letters in dosage instructions. As AI translation systems advance, spotting errors becomes more challenging due to their natural-sounding output.
Ultimately, while AI translation is undoubtedly beneficial, eradicating the need for human translators entirely remains a challenge due to the inherent complexity of language and the crucial role of context in accurate translation.
Written by urgent.news from BBC Business's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
