The advances in translation technology in the last decade have been vast. There has been a race between big companies such as Google and Microsoft to create real time translation wearables, a la Star Trek, and programs that can translate within chat apps.
The aim of this technology is to break down barriers. With the help of machine learning software is advancing at a an exciting and terrifying rate. Terrifying that is for the future of the human translator. But can such an important role be rendered obsolete by technology?
The current translation industry attitude toward machine translation
Translation is a vast subject and the world of literary translation is very different to that of business or law, yet again marketing takes a different approach to that of medical translation. Translation agencies that work primarily within the business sector have in the large part embraced aspects of machine translation, explaining to consumers how they use machine translation within their work..
Machine translation speeds up the translation process, allowing companies to achieve a quick turn around – however the belief is that a machine translation alone, at this point in time, could not do a satisfactory job. Human finesse and judgement is needed to make a translation passable, never mind perfect.
Literary translator Daniel Hahn said on BBC World Service’s The Forum, that the leaps and bounds made in translation are more exciting than alarming. Though this attitude may in part come from the fact that machine translation could never reasonably replace literary translation, it is an attitude shared by many business translators too.
Translation technology predictions
Though translation machines are vastly improving, there are so many little known languages in the world that it will be along time before it’s possible to retire human translators. This is particularly true for the less popular languages that do not necessarily have ‘commercial value’ but instead cultural value.
However more popular languages are seeing a lot of activity in machine learning and translation technology. Microsoft has rolled out Skype translator over the last year which can now function both in the instant messenger, and translate speech, in 7 languages. The 7 languages are, quite unsurprisingly, English, Spanish, French, German, Chinese, Italian and Portuguese.
Through using deep learning is is expected that machine translation will reach dizzying heights in terms of accuracy, allowing websites to be translated easily to help global business. According to Moore’s law this technology will have doubled in capacity within the next year which gives the idea of real time translation wearables – which already exist – credence to become more widely used.
Adaptation, not extinction, for translators
There a certain translation industries many are already certain will not be affected too harshly by the technology advancements. As mentioned before literary translation is one of these, but also sensitive translations such as diplomatic or law related translations will need to be handled by human beings who will understand the nuance necessary.
It’s also arguable that marketing materials will not thrive if done through only machine translation. Localisation is such a huge part of effective marketing and depends very much on the nuance mentioned above. Marketing to a different market will likely always require a translator native to the tongue of that market, rather than machine translation alone. Basic translation may move in the direction of machines but anything high end and nuanced will need a human.