In the latest episode of ADAPT’s podcast series, Professor Andy Way, an international expert in machine translation with over three decades of experience, delves into the rapidly evolving world of AI and its implications for the field of translation. The discussion centers on whether AI, particularly large language models, poses a threat to human translators.
The use of AI in translation has been increasingly prominent, with advancements in technology continually pushing the boundaries of what’s possible. Professor Way, Deputy Director of the ADAPT Centre at Dublin City University, addresses key aspects of AI translation, its successes, limitations, and the general hype surrounding technologies like ChatGPT.
During this talk, Professor Way highlights how AI and ChatGPT are being utilised for both positive and negative purposes and stresses that the overhyping of AI technologies needs to be tempered to avoid disappointment and ensure responsible, explainable AI. He notes that “for languages lacking robust machine translation systems, the need for human translators remains” and recommends that “developers should shift to multilingual large language models to avoid redundancy in their systems”.
Professor Way also talks about the significance of the European Language Equality Project. Ultimately, there is a crucial role for human intelligence in leveraging AI systems.
Professor Andy Way has been a pivotal figure in shaping machine translation globally, having served as Editor for the journal ‘Machine Translation’ and President of both the International Association for Machine Translation and the European Association for Machine Translation. He has numerous awards and over 400 peer-reviewed publications, highlighting his impact on the field.
For further insight into Professor Way’s role in translation and other fields, catch HumanAIse on SoundCloud, iTunes, and Spotify.
ADAPT Radio: HumanAIse is ADAPT’s newest podcast series providing an in-depth look at the future of AI, automation and the implications of entrusting machines with our most sensitive information and decisions.
