Translation is one of the areas of life where AI has already become indispensable. One of the most obvious advantages of machine translation is how it helps us to communicate with people all over the world, regardless of their and our linguistic background. In this way, AI-powered tools break down barriers that used to feel insurmountable. AI can translate or simplify written and spoken text almost instantly, which makes content more understandable and accessible to a broader audience and can support language learners. Yes, when it comes to translating, AI made my life easier!
Inevitably, these advancements lead to upheaval in the translation industry. Professional translators and interpreters fear to be replaced by AI sooner or later. Recently, it was reported that translators in the Netherlands have to deal with falling fees and AI was readily identified as the culprit. In the following, I would like to nuance this picture a bit.
Undeniably, the profession of translators is evolving. Many of them find themselves in the role of post-editors, tasked with reviewing and correcting machine-generated translations. Apart from the fact that many translators experience this work as more daunting and uninspiring (“this is not what we were trained for”), it also pays less. Many businesses are experimenting with machine translation to save costs and seem to accept lower quality of translations. And when human translators are hired after all, they should justify the expense.
That humans are still critical for double-checking reveals some of the limitations of machine translation. AI is far from infallible (for a particularly striking example, have a look at my last blog), and machines simply can’t do every task we’d like them to do for us. Specifically, in many cases, AI-powered tools are not equipped with the contextual and cultural understanding that professional human translators can provide. For complex and sensitive issues – think of medical, legal, or military contexts – AI is therefore too unreliable: errors can potentially have life-threatening consequences. Apart from that, we should also not forget privacy concerns. Furthermore, AI, if considered to be creative at all, can only produce artificial creativity. Skilled humans will be needed for, e.g., literary translation to ensure subtlety and cultural appropriateness. Basically, for the ‘human touch’, human involvement is indispensable.
So yes, translators are confronted with challenges posed by AI, and they will have to adapt. But, crucially, and by now maybe somewhat less surprisingly, the translation industry is not expected to face a massive job loss in the coming years! Instead, in spite of (and perhaps also thanks to) the rise of AI, a slight growth of the sector is predicted (globally; this article provides the numbers for the US). While the fear of job displacement nevertheless appears real and valid to me (not everyone will be able to adapt), I recognise that the demand for human translators will persist.
Initially, I intended to portray the translation industry as a test case for the impact of AI on a certain sector with regard to potential job loss. After all, AI is being part of translation for a couple of years already, and translating might be one of AI’s strongest powers. Now, I doubt whether the above described observations help to predict employment in other economic sectors that are confronted with AI. By all means, the impact of AI is complex, and some of us are affected more than others. Technological progress is accompanied by challenges. I am curious to see how we will manage to deal with them.
You’re right, the translation industry is increasingly likely to be replaced by AI, but when it comes to the “human touch,” AI can’t quite compete with humans. Real translators are better at providing context-aware translations that are more attuned to local culture and nuances. Unless AI translation can achieve a more localized approach through machine learning in the future, I still have confidence in human translators for now.