AI translations – why everyone should be satisfied

 

The speed and quality of services will increase, and pricing will feature much more segmentation in the interests of customers

 

Orlin Georgiev, owner and managing director of Prevodachi.eu

 

I suppose not many people know that in 2026 we will celebrate the 20th anniversary of one of the most popular platforms for free online translations – Google Translate. In April 2006, the start is “modest,” but it shows Google’s true scale – a statistically based platform that uses UN and EU documents and transcripts to quickly and efficiently generate a database through which translations into the respective world language can be produced. And English is once again confirmed as lingua franca, because the translation passes through it before it is directed to the target translation language. Many of us probably remember naive jokes we made with the “awkward translation” – for example, in the summer of 2009 the Bulgarian for “I’m just asking” was, more or less – “I am only the king of z.dnicite.” Yes, I know – you can’t verify it at the moment, and I didn’t take a print screen! Maybe because at that very moment I already embraced Google’s vision and the possibilities of SEO as a very serious tool for business development, and the success of our business showed that I sensed it correctly, while many of our competitors at the time considered the online direction of our business to be… extravagant and somewhat unlikely!

Ten years later (2016) Google took the step that defines the direction of development today – neural machine translation (NMT) began to translate entire sentences and the final result was not 100% literal, but carried characteristics of “professional translation”. The current integration of platforms (e.g., DeepL) with large language models (LLMs) is not merely a continuation, but the next generation of the NMT concept (as far as we can nominally compare them) – including broad multilingualism and zero-/few-shot, i.e., a real system for contextual translation and revision. Translate, on the other hand, is on the way to realize Douglas Adams’s Babel Fish from The Hitchhiker’s Guide to the Galaxy, experimenting with AI-based simultaneous translations and even language-learning elements within the app.

At first glance it seems that from the rapid development of AI all of us in the translation business are on the verge of losing it… and my intuition tells me once again that the opposite is true – we can develop it to the benefit of trusted relationships with our clients and to advance additional services related to the effectiveness of the language used! Of course, provided that we are aware of the nature and the “magnitude” of the changes and we are ready to change – both quickly, but also freeing ourselves from the prejudice that we can rely on people’s weak information or their distrust of modern technologies. Honestly, no translation business has set itself the goal to “monopolize” every translation in a single global world and information society of even 2016 – and there is hardly any translation agency that can do this in 250 world languages, in real time! In fact, our strength lies in something else – that we can adequately and in a short time have a competent translator or a team of translators review the translation (most often – specialized), apply their edits and ensure that the client can blindly trust the translation and simply forward it. Especially when it comes to important documents such as epicrises, contracts, court rulings and more.

When I founded Prevodachi.eu, we began using software that learns, i.e., helps the translator translate faster and not deviate from terminology. In this sense, our team is prepared for this and we can use opportunities for faster responses to client orders and optimized pricing. My internal sense is that this successful synergy between AI and experienced translators with continually increasing competencies (linguistic and technological) represents a successful and fully feasible development path, which in the next 5–10 years will create an entirely new economic sphere, in which part of our services will overlap with fields such as education, marketing, information technology, creation of media content, and so on.

Linguistics is a science that many do not classify as “exact sciences” – and in fact, the strive of every business, organization, institution or public figure to appear adequately in search engines and in AI tools shows that it deals precisely with THE COMBINATION OF QUANTITATIVE AND QUALITATIVE INSTRUMENTS OF LANGUAGE, which must synchronize with the algorithms of the digital world. Since the time of Harold Lasswell, who during World War II used content analysis to predict Hitler’s plans by analyzing his speeches and achieving exceptionally high effectiveness in forecasting the future actions of the aggressor, the potential of language as a social instrument makes it extremely important. Every technological innovation does not diminish, but increases the importance of the skillful use of language – this is not an exception. Our willingness not merely to adapt but to devise new approaches and services is not a “lifesaver” – I would liken it more to a “Noah’s Ark,” in which we will gain greater respect for language and its function in culture and in our relationships with one another.

And whether this is true – depends on us…