From Cloud Captioning to Cloud Revoicing in Times of AI: Testing the Accessibility and Usability of the TRADILEX Platform
摘要
In a world in which technological advances have a direct and significant impact on every dimension of our lives, it is important to highlight the need for preparing students for a complex job market. In this regard, when it comes to translation, generative AI and the spread of large language models (LLM) have had negative effects in terms of employment insecurity, as the industry shows an increasing tendency to offer precarious employment based on bad practices. In this regard, translators’ trainers should be aware of that fact and equip their students with skills that can hardly be replaced by generative-AI or machine translation. As literature confirms, creativity is the key skill for translators as LLM cannot deal with certain aspects of language linked to pragmatics, cognition and culture. The TRADILEX platform offers a reconceptualization of modern language courses included in Translation and Interpreting programs because these courses can be of great utility to foster creativity through didactic audiovisual translation (DAT) at the same time that students are acquiring language skills. DAT consists of the pedagogical use of translation applied to audiovisual texts. Therefore, students learn the language by dubbing or subtitling videos themselves, so they become prosumers. The TRADILEX platform offers sequences of lesson plans of different audiovisual translation (AVT) modes (subtitling, dubbing, voice-over, audio description, and subtitling for the deaf and hard of hearing), which have been tested and piloted and have offered positive results. It is a free online resource that allows cloud captioning and revoicing and that, apart from the classical combinations of direct and reverse translation, offers lesson plans on creative translation. The present study consists of a first step to understanding the perception of trainee translators regarding the TRADILEX platform in terms of usability, accessibility and language gains, since a positive perception of these three variables implies that the platform is an inclusive resource. The data was gathered through a questionnaire following a pre-experimental post-test-only design and the results show a positive perception in terms of the three aforementioned dimensions.