Future Citizen News
Globalisation and Language: machine-generated translations
A world with greater international mobility will increasingly experience the need for fast and reliable translations for purposes of knowledge-building and to facilitate communication (think of the special glasses that may help foreign visitors to the 2020 Olympics in Japan help decipher Japanese menus). From that perspective it is quite shocking that, according to Stephen Doherty, "analysts from the translation industry report that only a tiny amount of digital content, less than 0.1%, is currently being translated".
 
 
However, major developments are currently taking place in the translation industry, so that an exponential increase of the number of translations is to be expected. A major breakthrough was accomplished with the recent move from the statistical method to neural machine translation. The latter method uses artificial intelligence learning processes that closely resemble the way the neural networks in the human brain work. It can take into consideration the context of a written text when translating a particular word or sentence, so that translations of words with multiple meanings or ambiguous sentences do not lead to bizarre outcomes. (The famous example of a Russian computer supposedly translating ‘the spirit is willing, but the flesh is weak’ as ‘the wodka is strong, but the meat is rotten’ apparently is a myth.)
 
 
While we may at first sight expect that machine translation will leave many translators without a job – as part of the predicted process of substantial job loss due to automatization– experts in the field predict that the translator’s job is not at risk but will change to post-editing high-quality machine-generated translations. The European Union is at the forefront of this development. As explained by the European Commission, the main goal of its eTranslation project launched in 2017 "is to help European and national public administrations exchange information across language barriers in the EU, by providing machine translation capabilities that will enable all Digital Service Infrastructures (DSIs) to be multilingual".
 
Author: Dr. Olivier Vonk
Previous