The future of translation: on and offline


Type “future of translation” into a search engine and you’ll be presented with pessimistic articles questioning whether translators and interpreters will all be unemployed in a few years and whether machines will completely replace human translators. But what is the situation actually like? Is machine translation (MT) really putting thousands of jobs at risk?

Here is what we know about the future of translation on and offline!

What does the future of translation look like?

In our digital times, businesses don’t have a minute to waste! So, they use new technologies to translate their company websites, newsletters and communications materials.

If they didn’t, their competitors would have the edge.

But it’s important to distinguish between two types of translation tools.

First, you have translation tools with an integrated dictionary and rules. These are known as computer-assisted translation (CAT) tools and, just like humans, they have a translation memory. This means they can save words and phrases to support and facilitate the translator’s work. SDL Trados Studio, MemoQ and Wordfast are the best-known CAT tools.

Second, you have machine translation engines like DeepL and Google Translate. While these can be handy if you want to translate a non-technical text, they remain flawed and a long way from being powerful enough to replace human and professional translators altogether…

The future of translation: a homogeneous mix of humans and new technologies

Humans will always be better at detecting cultural subtleties in a text than machines. However, human translators do need to climb on the bandwagon and use translation tools to save clients time and money.

Many translation agencies are already using technical software to automate the translation process and CAT tools are constantly improving. Does that make the human translator obsolete?

Not yet! For the moment, tools support translators in their work by allowing them to work faster. This would be impossible without artificial intelligence.

The importance of artificial intelligence

Translation and artificial intelligence (AI) are developing alongside each other. As a reminder, AI means machines that imitate human intelligence to carry out a range of virtual tasks.

Based on neural algorithms (NMT), machine translation is improving daily because it can learn. Machines can now translate over 7000 languages and dialects successfully.

Thanks to this technology, translation agencies can supply reliable translations more quickly.

Futuristic translation services

Researchers and experts are currently working on futuristic translation services like speech translation and neural machine translation (NMT).

Let’s explore these two futuristic solutions in the translation sector.

Speech translation

Speech translation is still in development. Why? Because this innovation requires three things:

  1. Speech recognition and discourse analysis;
  2. Translation (using a tool with a translation memory);
  3. Text-to-speech delivery.

These three things must be meticulously carried out on all words spoken within a few tenths of a second.

Neural machine translation

Neural machine translation (NMT) is a technology that uses artificial neural networks. It has come on hugely in recent years. NMT uses data corpora: the more data there is in a corpus, the more reliable it is. This data includes documentation resources, terminology databases, glossaries, etc.

Neural machine translation is now more like a human translator than a CAT tool.

What do these changes mean for translators?

Machine translation is the future. When it has progressed far enough, the translation profession may disappear over time to leave machines doing all the translation work.

At that point, humans will exclusively be in the role of post-editor. As a key element involved in producing faithful texts in a careful style, post-editors will never completely disappear.

In future, humans’ task will be detecting grammatical errors, mistranslations and misinterpretations in texts translated by a machine. To do this work, a post-editor needs to have real linguistic skills in the source and target languages. Thanks to him or her, machine translation will generate the same quality as a human translator.

The future of translation: things to remember

Although machine technical skills are increasingly approaching those of humans, nothing can replace the knowledge of a native translator. Texts can be highly technical and without having studied the topic at hand, they are impossible to translate.

Of course, NMT can contextualise a text so the most appropriate word can be chosen. However, terminological and grammatical checks are and will remain an essential part of the translation process in every language pair.

In short, CAT tools are currently support tools that are becoming more and more important in the translation world. That said, post-editors should always be there to check the output.

New Call-to-action

You may also like
Expert opinions, In the news, TextMaster