I have written about getting up and out of tier two of the translation business (where most translators are) elsewhere, and the need to think about doing that has clearly grown since my presentation on that topic back in 2019. With the subsequently accelerated migration of agencies to AI followed by post-editing, that presentation is more relevant today than it was five years ago.
Lately, a significant number of translators fear that AI will take away much of their work. Let me say this: That fear is justified, and is playing out already.
In response to this fear, numerous translators can be heard claiming that their sensitivity to cultural differences will set them apart from—and place them above—AI translation. For most translation, however, this is an empty argument, as most of the translation that is sold is not related to or affected by cultural differences.
Translation of creative works (including games, anime, and other entertainment content) is a tiny demand segment, compared to domains that generate very little buzz but bring in great amounts of income for people in the translation business. Even adding marketing doesn’t tip the volume much toward subject matter that would require cultural sensitivity.
Similar to the ineffectiveness of talking about culture is the invoking the translator-created buzzword transcreation, which is quite irrelevant when the subject matter being translated is not culture-specific or “creative.” For most translation, invoking the transcreation mantra is no more meaningful or effective than repositioning products and services as solutions.
Translation in the fields of finance, engineering/manufacturing, medicine, pharmaceuticals, patents, and legal, for example, are natural targets for AI, and they are already being done to a considerable extent by AI. And a considerable amount of the JA-EN discovery document translation demand for litigation in the US, which is arguably a subset of legal translation, has been handled in recent years by translation brokers in China and is now in the process of migrating to AI systems in various venues. Seldom does translation in these various commercially important fields have any cultural aspects that deserve special attention. What should a translator of these buzz-challenged subjects do?
I think it’s time for professional translators to recognize and exploit their true advantages, the things that AI cannot provide.
One of the strengths of a professional translator is the ability to care. AI doesn’t care and cannot care. It cannot care to ask pointed questions about a source text. It cannot care about the purpose or the presumed readers of a document, and it cannot care about context, either within a document or within a stream of documents of which the document currently being translated is a part. More seriously, it cannot care whether it is even capable of translating a particular text. It just spits out a translation, with the owner or user of the AI assuming that a cheap-enough post-editor will be able to make the translation good-enough to sell.
For many types of translation, the above level of AI indifference might not create serious problems, particularly when considered together with the deep reduction in cost and blinding speed brought by AI.
It remains, therefore, for the translator to identify and acquire clients that value the care that a professional translator can bring to the translation process, and value translators willing to walk the extra mile and, for example, get on the phone and discuss issues that have arisen. The universe of that type of client will not include the existing large agencies that have supported and continue to support the great majority of translators. What that means for translators and what translators should do to survive is clear.
Another strength of a professional translator is end-to-end accountability. Whereas the provenance of a translation produced by AI can be laundered by a translation seller claiming that an expert “linguist” takes responsibility for the accuracy and overall quality of a translation because they post-edited it, professional translators know that such claims are often at serious odds with the actual quality of what results from post-editing at extremely low rates. If professional translators understand this reality, they need to convey it to clients who will care about it. Again, this leads translators away from the agencies that traditionally have provided most translators their livelihoods. Those agencies, like the AI they are relying more on these days, cannot be bothered to care about accountability, as long as the brokers can fashion a believable story about how a translation was produced.
There you have it. Well, not exactly, because it is up to each individual translator not wanting to exit translation to identify and acquire specific clients and people to speak with at those clients who will care that professional translators care and that they take end-to-end responsibility for the translations they produce, something that most agency users of AI cannot and will not do, because to do so is inconsistent with the profitability of their business model and the demands of their client demographic.