DeepL, Skype interpreting and friends

Current advertising of the big tech giants are trying to make us believe that the era of human translators has come to an end. Did you know that scientists and engineers in the area have been making the claim that machines will soon replace human translators since the 1960s? But it has to be said, where we used to ridicule Google Translate and others for their funny translation results until recently, there has definitely been a huge jump in quality in the last few years, caused by the use of neural networks and machine learning.

The results of modern AI-based translation tools are very good, especially in what is termed the “large” languages. This could well lead us to believe that human translators have now actually become superfluous, after decades of claiming it. But is that actually true?

You may call me biased, given that I am a translator myself, but I would argue that: no, we have not. Why? That is what I would like to explore in this three-part post series.

This month, in part 1, I will have a look at the advantages that machine translation can provide. In part 2 in March, I will be looking at the problems that machine translation still faces. And the final post of the series, part 3, will explore where machine translation is a good choice and where it would be recommendable to work with a human translator.

Advantages of machine translation

Machine translation is fast, cheap and often quite good these days—at least in languages like English, German, Spanish or French (but not so much for smaller languages such as Icelandic, for example).

It only takes you a few clicks to copy and paste your text into a translation window and receive an almost instantaneous result. There is no need for an extensive search for qualified human translators and almost no waiting for the results. You can go and immediately process the translated text, e. g. by copying it into an email program to contact that lead you have been in touch with recently.

Machine translation is also great to get a rough idea of the content of websites or documents. If you are a researcher who speaks only English, but have found a promising paper on the web, machine translation will give you the opportunity to check whether the content is actually what you think it is.

Something that especially appeals to our fast-paced modern lifestyle is the quick turnaround. Long texts that would take a human days or weeks to translate can be done within minutes or seconds by online translation tools.

Translation tools might also beat some humans in “user friendliness”. Rather than dealing with a human and “risking” complicated queries or negotiations, a software tool just delivers results without asking any questions.

But the main argument for many is money: professional human translations can be perceived as quite expensive. When there are free-of-charge tools everywhere on the web that sometimes deliver great quality, you might think there is no reason to work with human translators at all.

But as mentioned in the introduction to this post: life is not that simple. Machine translation does definitely have its place, but it is not always the best choice for your specific needs. Even modern intelligent technology still encounters problems, which are down to… well, human nature. That is why in my next post in March, I will highlight aspects that humans still deal with better than neural networks and AI.