Show Notes
Our guest is Professor Isabelle Augenstein from the Copenhagen University computer science department, where she heads a team dealing with natural language processing. Professor Augenstein shares her deep insights into how large language models work and what we as users can expect – or not – from the output.
Music: Title:Days Past. Author: In Closing. Source: In Closing - Days Past (auboutdufil.com). Licence: https://creativecommons.org/licenses/by/3.0/deed.fr
More from DGT podcast: Languages and Technology
How can we make sure AI works fairly across Europe’s languages?
In this episode, we explore translation, terminology and artificial intelligence (AI), with a particular focus on the Irish language. We discuss how new AI technologies are changing the work of language professionals, and what it means to work with a language that has fewer digital resources than many larger languages.
How does a European Commission document make its way from an initial draft to 24 equally valid language versions?
Transcreation sits where translation meets copywriting, creativity and cultural adaptation. In this episode, Anna Holmén talks to Italian transcreation expert Claudia Benetello about why human nuance, humour, voice and individual flair still matter in a world increasingly shaped by AI – especially in marketing and creative language work.
Multilingualism is a democratic right and a core part of the EU’s identity. Christos Ellinides, Director-General for Translation, talks about what this means in the era of AI – and what we do to safeguard EU language diversity. Why is DG Translation a good place to develop AI tools? What is the role for linguists in all this? And are we really building an LLM of our own? Hear the answers to these questions and many more.