The rapid rise of AI and the emergence of Large Language Model (LLM) technologies have cast uncertainties over the relevance and utility of human linguists. Even the very existence of translators has been called into question, with some fearing that their profession may soon go extinct.

In this talk Marina Pantcheva addresses this pessimistic outlook and presents strategies for preparing linguists for a future filled with opportunities where their skills remain relevant.

The presentation will start out by providing an in-depth understanding of what Large Language Models (LLMs) can and cannot do well. This exploration of LLM capabilities is based on real-life research and practical applications of linguistic AI in localization processes rather than on mere theoretical speculations. By drawing parallels between the linguistic capabilities of AI and that of humans, the presenter will identify the areas where the human language faculty still remains unmatched by AI.

Equipped with an understanding of where the human linguistic faculty outperforms or complements that of LLMs, Marina will then proceed to explore the ways and domains where AI and humans can work together effectively. She will focus specifically on how the role of human linguists must evolve and what specific skills and knowledge linguists must possess in a world where their work is augmented by AI. In doing this, she will also present the challenges that linguists may face during the (re-)training process, and later on when working with linguistic AI.

The talk will provide invaluable guidance on how to enable linguists to not simply adapt to the new AI technologies but to harness their power and use them in a productive way.

Host organization: RWS