Human Language Meets AI - A Conversation on Language and Large Language Models

Speakers: Dr Simon Fries and Dominik Lukeš

Activity Type: Discussion
Time: 18:00-19:00
Levett Room, Wolfson College & Online (Teams link).

About the speakers:
Dr Simon Fries: Prior to taking up my position as a Researcher in Comparative Philology at Oxford I worked as a researcher at the Institute of Linguistics and the Collaborative Research Centre on "Prominence in Language" at the University of Cologne, where I also received my doctorate in Historical-Comparative Linguistics.
I take great interest in all aspects of language change and language typology, and have published on particular processes and patterns of change in many languages, including ancient and modern Indo-European languages such as Ancient Greek, Latin, Sanskrit, Lithuanian, and Latvian, as well as other languages, notably the Yeniseian languages spoken in Siberia.
Dominik Lukeš is an AI consultant at the University of Oxford's AI Competency Centre (Oxford e-Research Centre), where he designs and delivers training on generative AI for researchers. His background is in cognitive linguistics, and his work focuses on how Large Language Models relate to human cognition, language, and learning.

Abstract:
How does human language work and how does it compare to the language of AI? This event brings together two expert speakers for a concise and engaging exploration of human language and Large Language Models (LLMs). Dr Simon Fries, Researcher in Comparative Philology, presents “How does language work and change? A brief introduction into the structure and historical development of human language”, offering a brief introduction to the structure of human language, including the grammar, production, and how languages evolve over time. Dominik Lukeš, AI Consultant, follows with “Language without structure? How can LLMs speak all the languages without knowing any of the rules?”, examining how LLMs generate fluent, multilingual language without explicitly learning grammatical rules, and what this reveals about language itself. Together, the talks open a space for discussion on the similarities and differences between human and AI language.