With the emergence of artificial intelligence and rapid improvement of AI-based translation tools the question arises: will human translators cease to exist shortly? Let’s try to analyze the facts and trends to find an answer.
According to Ahrefs’ keywords analyzer, in the US people search for an AI translator 3.2K times per month. While those looking for a professional translator are 3 times less (1.1K searches monthly). It may mean, that a lot of users find tools rather helpful in resolving their translation issues.
What languages are in trend now? According to Google Trends, Russian is super popular. The search term ‘translate to Russian’ has been trending during the last year. Look at the graph below.
Probably, this interest can be explained by the current war initiated by the Russian Federation 2 years ago. So, let’s dwell on a comparison between machine translation (MT) and human translating services in this particular language pair – English and Russian.
Machine VS Human Translator
Volume:
Free tools translate limited pieces of text at a time (for example, Google Translate allows 5K characters). So, such apps are good for vocabulary and short translations. But there’s hardly a way to feed it a few hundred pages of text.
Humans handle any project irrespective of its length. It means that for a book, website and any other long text translation professional services are more appropriate.
Complexity:
You can use tools to roughly understand the meaning of a Russian text. They are really good for things that can be directly translated without human involvement (like “The sky is blue”=”Небо голубое”). However, tools can get pretty wacky even for simple text at times. Look at this screen:
And what can we expect, if the text is more complicated? So, reliable English to Russian translation doesn’t work with these services.
Professionals usually do a better job. In-depth knowledge of Russian helps grasp the full semantic range. That includes abbreviations (like on the screenshot above: ‘Св.’ is a shortened ‘свежий’ which means ‘fresh’ but not ‘Saint’. A human translator will never make such a mistake!
Context:
The tone of voice, emotion, subtext aren’t considered by AI translators. MT can translate a hundred different sentences from a hundred different films in the same way.
A professional translator can translate a sentence from a scene in a film in a hundred different ways depending on the context, subtext, intonation, etc.
Proverbs, idioms:
Tools translate idiomatic expressions and proverbs literally or by using stylistically neutral equivalents. For example, according to the popular Google Translate English expression ‘out of the blue’ corresponds to Russian ‘внезапно’ (suddenly).
Professional Russian translators with real-world experience and deep knowledge of the 2 languages will render ‘out of the blue’ by finding the most appropriate idiom in Russian – ‘как снег на голову’ (lit. ‘like snow on one’s head’)
Sarcasm, humor:
Figurative meaning, metaphors, word-play etc. – that’s what machines can’t understand. If the source makes a pun on an animal that you can see on the screen, MT can’t understand the whole context and will translate word for word inevitably losing the punchline.
If you have a target language that, say, expresses sarcasm with a specific word to make it clear the sentence is sarcastic, but the source text has no such word, you need a human to know that word must be added.
Visual information:
Right now computers can’t distinguish between a rifle and a broom in a video to translate accordingly. You need to see the image to know how to translate the source text correctly, but a computer can’t see it. The proof of this is Captcha, a simple test to detect where a robot or a human is trying to, for example, sign in by choosing pictures with, say, bicycles.
Human translator takes into consideration everything they can perceive and understand: from cultural background to nuances of speaker’s tone. That helps to pick up the very best word or expression which conveys the intended meaning with utmost accuracy.
Creativity:
The areas involving creating writing and especially literature – that’s where human translators will always be in demand. The fact is, we cannot teach a machine to feel, at least for now. So, for now a machine will never do the same work a poet can do when translating a poem from English to Russian. There have been experiments done with a humanoid robot writing poetry. Yes, AI can generate poetry but can these derivate samples be compared to outstanding masterpieces of talented poets? No, and the fundamental reason is that AI systems lack an essential ingredient: imagination.
What Might Be Happening In The Translation Field In The Near Future
Shifts In Specialization
In most fields human translators don’t do much of translations from scratch even now. They often just copy-paste a machine translated text, and make corrections so that it reflects the original meaning. This process is called post-editing of machine translation (MTPE) and it has been a reality for several years.
We can predict the increasing prevalence of post-editing in the future. That may mean translators are being deskilled and reduced to linguist post-editors.
Of course, the translation market will always require professionals specializing in high-value high-risk content (literary, legal, healthcare, finance, localizaion). So, a translator will need to have some areas of expertise, the more niche, the better. A human will always be required to ensure accuracy in translations, no matter how advanced AI becomes.
Also, professionals will still be needed for small-market languages where the cost of AI training outweighs the cost of human labor. So, in the future linguists will have to focus on translation for lesser-known languages.
Transcreation will still be needed, like for example translating a movie name or a company’s slogan. That is especially true for professional poet-translators. However, the more creative translations make up a small portion of the market.
All these factors will probably become the reasons why lots of translators will be leaving the industry.
English To Russian Translation Rates
Human translators won’t cease to exist but there won’t be enough well-paid work for all of them. However, highly specialized or creative ones will still have jobs and earn enough to make a living. The rest will struggle with reduced rates as they’re forced to do MTPE. Downward pressure on prices will continue.
Having said that, the job market will definitely shrink. The small and mid-sized agencies will keep trying to reduce the costs. That’s what we are observing now already. Some companies offer English to Russian translation rates per word starting from only $0.06 (check out, for example, RussianLanguage.Services).
Conclusions
AI translators are the best choice for clients picking quantity over quality. However, we should bare in mind that modern apps can screw you from time to time. So, for now they mostly work as a way to get a general idea.
Human translators will continue to exist but their work will sometimes be very different from what it has been traditionally. And the times of professional translators doing a huge volume of straightforward work are pretty much over. Probably, translators will focus more and more on editing a machine translation to get the exact meaning across. AI needs them to work properly.
The sad truth is that AI will probably replace humans in many fields of translation. Sooner or later AI is going to be able to solve the problems regarding lack of real-world experience, context etc. Because it basically feeds on us and as long as we keep training it, it will learn.
However, that will happen not in every field and language pair. In rare languages and high-risk areas, and for projects requiring creativity humans will still be in demand. There are a lot of reasons why a company or a person might opt out of AI and go to an actual translator. Most probably, at some point only the really big players will be willing to pay more to maintain the reputation of consistent high-quality output. Regarding the English to Russian translations, accurate, contextually precise and culturally relevant results will be key in order to give a true insight into the war from both sides.
To sum up, the possibility of the translator’s job being completely wiped out by a machine is really low. Translators will still exist for a long time because only humans can understand all the nuances of human communication.
