In our August article on the subject of translation apps at the Rio Olympics we searched for the best machine translation apps currently available to download. Of course any translator knows that when it comes to professional translation, machine translation will often not be associated with the adjective “best”; but hey, the machines need some credit!
Often a source of humour, the texts produced by machine translations are constantly evolving as, for some reason, we humans believe that one day machines will be able to do absolutely everything for us! In today’s society it’s rare that anybody does anything without some type of machine, even professional translators use CAT tools powered by machines to aid their levels of productivity and quality management; but unlike Machine Translation, CAT tools don’t produce an entire translation for you.
Machine translation systems spit out all sorts of absurdities. Whether it is used by lazy students, restaurant owners, or tourists, machines give instant results sometimes accompanied by undesirable consequences.
Apart from the mistakes highlighted in the image below, there are many other inaccuracies within the text that make it an inconsistent and low-quality translation:
Machine translation systems can be useful for gist purposes or if you have knowledge of both languages, but for those who do not have the latter it can have disastrous consequences. There is also a question of privacy and security, once entered into an online machine translation service such as Google Translate, your client’s or your own sensitive data is no longer privacy protected.
In September, Google announced a great advance in their technology: Google Neural Machine Translation system (GNMT). The system is claimed to provide results that are so accurate it is difficult to tell, upon comparison, which is the machine translation and human translation. However, we can assure you that for the foreseeable future, Human Translation is the way forward for accurate and high-quality results.
Although we support the constant evolution of technology, and as professional translators we may be considered slightly biased, we believe that machines will never be able to fully replace human translators, nor will machines ever replace the need for language learning. There are simply too many constantly-changing facets of complexity in a language to successfully be interpreted solely by a machine without the intuition of a human being.
Sources: Wikipedia Commons; pexels.com