DiSCourse Seminar with Katharina Walter on Sustainable Use of AI in Translation

Talk given on 3 May 2024 by Katharina Walter from the Department of Translation Studies at the University of Innsbruck.
Abstract:
My presentation examines what a sustainable interaction between humans and machines in translation might look like in the age of AI. Apart from potential ethical and legal questions around copyright, this presentation also addresses the possibilities and limitations of sustainable human-machine interaction regarding the quality of translation output. Recent research shows that it is not only irresponsible to use AI-based applications such as DeepL or ChatGPT without human supervision in high-risk situations, which may occur, for instance, in the healthcare sector. In literary translation, my main area of focus, we also have at least an ethical obligation towards original authors and their potential readership to produce accurate and high-quality translations. As any text generated or translated with AI is normally also used to train AI tools, keeping humans in charge is the only way to counteract an otherwise inevitable degeneration in the overall quality of our linguistic environment.