I can use Google Translate, why should I pay for a translation service?

la creación de Adán

A few weeks ago, Slator—the reference media for the translation industry—published a very interesting article on a paper by two recognized institutions: Oxford University’s Future of Humanity Institute and Yale’s Department of Political Science. A group of experts on artificial intelligence (AI) were asked about the future for some tasks currently performed by humans. Specifically, they asked: “When Will AI Exceed Human Performance?” They covered many areas, including translation. The very same question they were asked gives a very useful tip of where opportunities lay for translation companies:

  • Regarding translation (Amateur Translation): When AI will be able to perform translation about as good as a human who is fluent in both languages but unskilled at translation, for most types of text, and for most popular languages (including languages that are known to be difficult, like Czech, Chinese and Arabic)?

By the way, the forecasted timeframe is 2024, but this is for languages where MT (Machine Translation) has not yet reached an advanced stage. For certain language combinations such as English into Spanish or certain domains such as IT, the future is almost here. We periodically check how Google is doing on our main language pairs and it is learning fast! MT, including neural MT, keeps improving and the actual content in the web keeps growing. In view of this, a list of questions comes easily to our minds: Where do we have opportunities ahead? What shall we tell to translation college graduates? Is translation offering job opportunities? What value can we provide to our clients when translation is readily available over the internet or, in the case of more sophisticated companies, available via a customized MT system?

We should focus on the “Human Touch”. Once you have considered all repetitive tasks as ready for automation, where are the opportunities for us?  Let us look at some of them:

  • Dealing with different formats. Although many CAT applications deal with multiple formats, a human being still needs to prepare the source or target files and carry out a quality control before they are published in other languages, especially if they change from left-to-right to right-to-left or use different character sets.
  • Specialisation. The more specialised a text is, the less suitable a machine translated output will be, particularly when translating new content or innovative products or services. We can be a strong support for a client that needs to search for linguists in a given industry. By being based locally, it is easier for us to find subject-matter experts.
  • Creative writing. AI is said to even write film scripts but the creativity of a human is still to be beaten by a machine.
  • Professionalism. Blame instant messaging and 140-character restrictions, but writing properly in some languages is becoming a good asset. As Language Professionals, part of our role nowadays is to be guardians of language quality. If we do not care about language, who will?
  • Quality at source. More and more often we see that the source text needs “polishing”, either due to shorter time to market (or to post) or because the staff at international departments must write in a language they might be fluent at but is not their mother tongue.
  • New language combinations. As the world becomes more and more interconnected, more companies require translation from and into other languages than English, German or French as source or target. For instance, we have lately seen an increase in the demand for Chinese into Spanish and Spanish into Chinese.

 

These are some of the opportunities we are seeing. What is your experience?