Rise of the Robots: chatbots vs human translators

Onuba Comms guest blogger Sophia Georgiou

By Onuba Comms’ guest blogger Sophia Georgiou, edited by Ronan Malt

Technology within the translation and localization world is constantly evolving. From Machine Translation tools such as Google Translate, to advanced Computer Assisted Translation (CAT) tools used by professional translators, there is a wide variety of machine-based functionalities that allow both linguists and non-linguists to disseminate information across the world.

A fairly new functionality that is emerging is the translation chatbot. Although the monolingual chatbot has been around for a while, training a bot to understand multiple languages has proved to be rather complex. The person building the bot must understand the grammar, syntax and other complex aspects of the language, and as is the case with machine translation, context-specific issues, cultural references and idioms are never understood and translated as well as they are by a human translator

However, AI developments have allowed for deep neural learning for natural-language processing, which means that these bots will eventually be able to learn more colloquial aspects of languages that are being used online.

Since the online market place is ever-growing, it is estimated that 80% of businesses will be using online chat bots by 2020, so it seems essential to be able to offer a wide range of language-speaking bots for customers who prefer to communicate in their native language. This is an important localization solution that is going to aid growth in online business and communication. 

Last year, Facebook announced that they were introducing this functionality into their marketplace messenger, starting with English-Spanish and eventually introducing more languages in the future. Not only as a way for people to connect interculturally, but Facebook also believes this will help people sell goods faster.

As of 2019, the Instant Translator Bot that Facebook and Viber have both integrated into their instant-messaging platforms has been able to work with 19 different languages. Although it seems like there is a long way to go with how AI is being trained in language-learning, we can expect to see more and more of these translator bots being integrated into our online experiences in the near future, with increasing accuracy.

However, what exactly do all of these developments mean for us as translators and the future of our profession? Well, whilst we are having to become more accustomed to introducing technology within our workflow – as the attempts to cut costs and improve productivity within the industry continue apace – human intervention is still undoubtedly necessary. 

This particularly applies to translations within our area of expertise, multilingual sport-related content. There is often a lot of reworking to be done, specialist terms to interpret and repetition to be avoided when it comes to translating sports content and machine-based tools are simply not yet up to the challenge.

¿Necesitas Ayuda? Escríbenos
Need help? Write to us