Machine translation has been used since the 1970s and was once reserved mainly for governments and large companies. Today it is experiencing significant growth, mainly thanks to advances in artificial intelligence. Indeed, the recent application of neural networks to automated translation has produced revolutionary results in a multitude of language pairs.
Gone are the days when automated translation offered only a more or less comprehensible overview of the source text. Companies like DeepL harness the power of neural networks and deliver stunning results, from terminology accuracy and consistency to fluent translations.
The biggest winners in this period of continuous improvement in automated translation are customers, that is, those with translation needs. Companies, individuals, non-profit organizations, everyone benefits. The transition from a traditional translation market to an automated translation post-editing market reduces the cost per word translated, and consequently the price paid by the client. In addition, translations are delivered faster, as average translator productivity increases and the entire translation process (post-editing, revision, layout, etc.) inevitably accelerates. Finally, we must consider that the automated translation engines will continue to improve in the coming years, accelerating linguistic tasks all the more.
Obviously, this technological tsunami is not without consequences for the translation industry, which must adapt quickly to the new state of affairs. Increased competition, on prices, but also on delivery times; accelerated learning of the operation of new software and new technologies; support for the workforce in these new tasks (post-editing, i.e. revision and correction of texts pre-translated by automated translation). The list is long and impact is substantial.
Some would say that many types of texts do not lend themselves to automated translation, and that is quite true. However, most of the translated texts are technical in nature or general documents, for which automated neural translation gives excellent results. The need for translation is constantly increasing, due in part to the increasing internationalisation of business and commerce. Individuals and businesses buy and sell products and services around the world today.
If automated translation has been able to make so much progress in just a few years, we must honestly ask ourselves: what does the future hold for us? It is reasonable to expect that in the near future, major upheavals will occur in the translation industry around the world. Translators are likely to be required to wear several hats, such as those of post-editor and reviser, but also new ones such as machine translation engine expert or termbase manager.
To learn more about neural translation, consult our blog article on neural machine translation here.