It’s a given that artificial intelligence (AI) is entrenched in our world. It is already omnipresent in many industries, in the form of personal digital assistants, search engines that steer users toward targeted content, self-driving vehicles, or smart cameras that can perform visual inspections on an assembly line. Given that these technologies promise to save us time and money, and be more productive, it will be increasingly difficult to ignore the phenomenon.
Will we shortly be seeing jobs disappear in law, transportation, manufacturing, services and communications, to name just a few fields? What about professional translation? Will automated translation drive translators into extinction? According to researchers at Oxford and Yale, AI will exceed human performance in translation by 2024. Has the time come to close translation schools?
Let’s put things into perspective. Not that long ago, translation was essentially done on paper, ideally close to a library that could be used for document and terminology research. Technology gradually emerged that made life a lot easier for translation professionals: word processing software, terminology databases, access to countless resources on the Internet, and computer assisted translation (CAT) systems. Those who adopted these tools agree that their productivity improved, enabling them to make a decent living even though rates were stagnating.
It may be otherwise with AI, a tool unlike the others. Time will tell, but we have to look at the evidence. There is no point in fighting it. On the contrary, we have to engage with it rather than leaving the technology’s development solely in the hands of the IT experts. Many translators find the prospect of spending their days doing post-editing unappealing, so it is up to us, the language professionals, to promote our expertise. While literary translation is less threatened by the dangers of automation, practical professional translation is not ready to become the prerogative of machines that do not “understand” the meaning of the texts they are processing. What these machines do is perform language transfers according to programming rules. They do not actually translate. This is where professional translators can come in, by becoming the experts who should know better than anyone how to use this technology.
In any event, artificial intelligence’s impact on professional translation is being felt in every part of the profession, including the teaching of translation: a post-editing course has been added to the curriculum at Université de Montréal. We can therefore hope that the profession is not dying out, but rather undergoing a change that will likely involve taming artificial intelligence.
*Text translated by Alison Newall.
 Grace, K. J. Salvatier, A. Dafoe, B. Zhang and O. Evans, When will AI exceed human performance? Evidence from AI Experts, 3 May 2018, page consulted on August 22, 2018