Abstract

The translation industry is undergoing a significant transformation due to advances in artificial intelligence. Many translators fear replacement by AI, some complain about having to correct AI-generated translations, and others work to adapt to this new tool. Our study aims to investigate whether artificial intelligence can produce translations that match the quality and reliability of human translators. By “quality”, we mean the degree to which translations are adequate, fluent, and culturally fit for their intended audience. We selected excerpts from four education-related texts (a scientific article, a catalogue, a blog entry, and a video) and compared their translations which were generated exclusively by traditional and generative AI agents with those produced by a human translator. We assessed which translations are more accurate and fluent, identifying their respective strengths and areas for improvement. Even though none of the translations is perfect, those generated by the human translator still prove to be the most accurate. In any case, human translators should not be afraid of using AI as a tool that can help them in the translation process.

This work is licensed under a Creative Commons Attribution 4.0 License.