Well, maybe not friends, but colleagues for sure. Not frenemies either. We need to keep it professional between us.

Let's elaborate on the very burning question in the era of GenAI: does Machine Translation (MT) threaten professional linguists and the overall quality of translation services?

How about we overhear a conversation between Localization Engineer Denis and Linguist Mary, two colleagues at our Alconost Team:

Mary: Is there a possibility that the job of a translator won’t be needed anymore with all these AI tools?

Denis:  From our point of view, if AI tools are ever to come close to draw significantly near the quality of human translations, it won’t happen at least in this decade. We strive to honor the essential role of professional translators in delivering top-quality translations for our localization services, particularly when handling creative content, intricate target audience preferences, brand voice and style.

Mary: I can wrap my head around using machine translation for technical and legal documentation when it has a defined structure. But how successful does it work for creative content like games, user-generated content (UGC), advertising, etc?

Denis: Creative content poses unique challenges! MT has limitations in delivering work that truly resonates on a human level, especially across various languages and cultural contexts. That's why we emphasize the irreplaceable role of human translators in bridging these gaps. Machines alone fall short. MT simplifies and accelerates the process of providing the initial translation (not by the magic of clicking “send” button to the Gen AI) which is ready for human-powered professional post-editing procedure. This includes a whole stack of work: fixing the style, tone, context and culture-specific nuances.

Mary: What does it mean when they say machine translation post-editing (MTPE) can reach the same quality as translations done by native speakers?

Denis: MTPE is a great tool that works alongside the skills of human translators, helping to increase efficiency and reduce costs of translation for specific types of content. However, we don't believe this process can replace the unique touch of a translator in making content feel natural, culturally relevant, and idiomatic.

Summary:

There are lots of myths and rumors around GenAI. It’s a good reminder of our responsibility to communicate the strengths and limitations of machine translations with integrity and transparency. And with that responsibility, we want to emphasize that that GenAI should be used selectively. The value of human expertise remains essential.

So yes, colleagues it is!