AI in Translation

Type ‘translator’ into most search engines, and you’d be hard-pressed to find human translators included in the populated results. Pictures of machine translators and AI systems are instead what flood your screen. It appears that search engines are confirming the concerns long held by many human translators about the bleak present and future of human translation. This can deter any translation student with aspirations to establish a career within the field. This need not be the case, though; artificial intelligence (AI) in translation is here to stay, and the best approach is not to turn our backs on it but instead learn how to work with it and make it work for us. 

There are currently two main forms of digital translation in the market: machine translation and AI translation. Machine translation (MT) is the traditional form most of us are already used to. MT produces word-to-word automated translation using computer software. Modern MT analyses text elements to produce a more coherent translation (What is Machine Translation? 2024). Examples of MT systems are Google Translate, Amazon Translate and Siri. AI translation takes it many steps further, incorporating sophisticated algorithms and utilising neural networks (a machine learning program, or model, that makes decisions in a manner similar to the human brain (IBM, n.d.)) to produce more nuanced, accurate and natural-sounding translations (Habash, 2024). Some examples of AI-powered translation tools are ChatGPT, DeepL and Worldly.ai.

How is Artificial Intelligence Changing the Translation Services Industry?

Globalisation has contributed to the demand for bridging between languages and cultures resulting in the boom of translation and interpreting services. Annual growth within this industry is estimated at 7% and is forecasted to reach $90.8 billion by 2027 (Institute of Translation and Interpreting, n.d.). For many years, multilingual humans were the go-to for all language-bridging requirements. However, in recent years, as the world transitions into a technological society, the introduction of AI in the language industry has caused a lot of upheaval. 

AI has the capacity to improve the efficiency and accuracy of human translators. It can automate and expedite some elements of the translation process, suggest alternative translations and identify errors. As a result, the impact of AI on the translation landscape has been monumental. As the world becomes more interconnected, coining the term ‘global village’, fast, accurate and efficient communication assumes centre stage in businesses and individual relations. This, in turn, creates opportunities for language professionals to expand their services, resulting in business growth and increased revenue (Carrier, 2023).

Another area of AI-powered translation that is garnering attention is within language learning. International students and language learners are utilising the instant output that AI translation provides. Previously, schools would employ professional interpreters or translators to assist international students in acclimating to their new life abroad, most international students today make use of AI software to navigate the new language and culture. Tools such as Camb.ai and Notta translate speech, making spoken content accessible across various languages. Owing to the marked advancement of AI in recent years, speech translation tools now demonstrate high precision in their understanding of cultural nuances, colloquial speech, slang and accents (Hanji and Haiqing, 2019). As such, many language learners and international students employ the use of this software. Outside of academic use, voice translation tools are often used when transcribing, dubbing or subtitling videos. AI-powered voice translation tools offer human translators similar financial opportunities, much like text-based AI tools (Carrier, 2023).

Literary translation, unlike business or technical translation, is not as straightforward. A translator often mulls over synonyms, various phrases and ideas for hours, days or weeks on end to effectively communicate the source text to the target readers. The creative nature of translating literature results in translators leaving an imprint of their voice and style in the narratives (Literary vs. Business Translation, n.d.). This unique variation noted in every literary translator's work cannot be replicated by AI translation, and herein lies the problem. AI-assisted translations produce generic and often uninspired texts. Despite these issues, some publishers employ the use of AI-powered tools to translate books. However, some companies, like Nuanxed, employ human translators to post-edit AI-assisted book translations. This undercuts the translator’s pay by offering ‘savings to publishers and “market rates”’ to linguists. This is a genuine concern for many literary translators, leading to organisations such as the European Council of Literary Translators’ Associations calling for translators to avoid such jobs. The hope in doing so is that much like the value of organic and hand or homemade food items, the quality of human translations can garner more appeal over AI-assisted translations. This, in turn, will benefit human translators in the long run (Aslanyan, 2024). 

Overall, the proliferation of AI tools in translation can be utilised to expand a translator’s business; unfortunately, it can also result in translators being undercut. Nevertheless, AI is here, and it is here to stay. Instead of being apprehensive about using AI, let’s reflect on its advantages. Translating a text can be arduous, and we need all the help we can get. AI can be a tool that assists us in navigating the many tapestries of language to reach beautifully curated translations. 

How can we use AI to increase our productivity as translators?

Mariana Silva (2023), a fellow translator specialising in marketing, cosmetics, tourism, and software translations, wrote a blog post on this topic for the Institute of Translation and Interpretation. Translation as a career does not end with text production. As professionals, we generate business, communicate with clients, research, and manage projects. Mariana proposes five ways we can boost our productivity using AI:

1. Maximise research 

Those hours we spend contemplating precise terms or phrases we need to communicate an idea can be significantly reduced through AI. AI tools can compare and analyse terms and their nuances to help us find the exact wording we are looking for.

2.   Enhance content creation

A translator’s work isn’t limited to translating but can, at times, comprise transcreation, SEO keyword research, copywriting and content creation; AI tools such as ChatGPT can help generate new perspectives based on our drafts and help to enhance our ideas.

3.   Efficient client communication

AI tools such as ChatGPT can draft clear and concise emails using a few words about the subject (do not include personal information). This way, you can dedicate most of your time to translation tasks.

4.   Resolve technical issues faster

When facing difficulties regarding the technicalities of translation software such as CAT tools, AI can assist in resolving these issues by providing multiple solutions.

5.   Task support

AI can also support general writing tasks such as reducing word count, rephrasing, and rearranging things alphabetically or numerically. This way, you can direct your focus to core tasks such as translating the text.

Three Most recommended AI tools for translators

1. ChatGPT

2. DeepL

3. Worldly.ai

Ethical Considerations

As with anything to do with AI, it's important to remain ethical when using it in translation. AI in translation is still relatively new, and without a standardised ethical and legal framework in place, many organisations have established their own for commissioned translators. Copyright issues resulting from the collaboration between humans and AI in translation is one of the major ethical issues in discussion and merits every translator’s attention (Bo, 2023). 

Final Thoughts

While AI is transforming the translation landscape at a rapid speed, it is essential to recognise its limitations and, therefore, the continual need for human translators. Without the creative touch of humans, AI-generated translations will remain generic and unmemorable. However, the collaborative work of AI and human translators has the potential to produce well-curated and refined translations of any scale. As AI has fully entered the chat, let’s welcome it and learn how to utilise it. 

References:

Aslanyan, A. (2024) 'AI translation: how to train ‘the horses of enlightenment’', Guardian, 15 March. Available at: https://www.theguardian.com/books/2024/mar/15/ai-translation-literature (Accessed: 15 April 2024). 

Bo, L. (2023) 'Ethical Issues for Literary Translation in the Era of Artificial Intelligence', Babel, 69(4), pp. 529-545. Available at: https://doi.org/10.1075/babel.00334.li.

Carrier, T. (2023) 'The Future of Translation: How AI is Changing the Game' [LinkedIn] 31 May. Available at: https://www.linkedin.com/pulse/future-translation-how-ai-changing-game-thibault-carrier/ (Accessed: 15 April 2024).

Habash, F. (2024) ’AI Translation’, Blend, 5 February. Available at: https://www.getblend.com/blog/ai-translation/ (Accessed: 15 April 2024). 

Hanji, L. and Haiqing, C. (2019) 'Human vs. AI: An Assessment of the Translation Quality Between Translators and Machine Translation', International Journal of Translation, Interpretation, and Applied Linguistics Preview, 1(1), pp. 1-12. Available at: https://doi.org/10.4018/IJTIAL.2019010104. 

IBM (n.d.) What is a neural network? Available at: https://www.ibm.com/topics/neural-networks (Accessed: 16 April 2024).

Institute of Translation and Interpreting (n.d.) The industry in context. Available at: https://www.iti.org.uk/discover/about-the-profession/the-industry-in-context.html (Accessed: 16 April 2024).

Literary vs. Business Translation (n.d.) Available at: https://localization.blog/2019/03/24/literary-vs-business-translation/ (Accessed: 15 April).

Silva, M. (2023) 5 ways generative AI can increase your productivity as a translator. Available at: https://www.iti.org.uk/resource/5-ways-generative-ai-can-increase-your-productivity-as-a-translator.html (Accessed: 15 April 2024). 

What is Machine Translation? (2024) Available at: https://aws.amazon.com/what-is/machine-translation/ (Accessed: 15 April 2024).