GDB3 | The Role of Large Language Models (LLMs) in Transforming Geosciences: Friend, Foe, or Fad?
Thu, 14:00
The Role of Large Language Models (LLMs) in Transforming Geosciences: Friend, Foe, or Fad?
Convener: Jens Klump | Co-conveners: Alice-Agnes Gabriel, Philip Stier
Programme
| Thu, 01 May, 14:00–15:10 (CEST)
 
Room E1
Thu, 14:00
Large Language Models (LLMs) and other advanced AI tools are reshaping how researchers and practitioners approach data analysis, synthesis, and communication across Geosciences. This Great Debate at the EGU General Assembly 2025 aims to critically examine the impact of LLMs in fields such as Seismology, Atmospheric Sciences, and Earth and Space Science Informatics. As these models grow increasingly capable of interpreting complex datasets, generating predictive models, and even drafting research manuscripts, they raise profound questions about their role and limitations in scientific workflows.
The discussion will address key issues, such as the potential of LLMs to accelerate discovery through data-driven insights, their capacity to support real-time decision-making during natural hazards, and the ethical considerations of AI-generated content in peer-reviewed publications. Are LLMs a transformative tool that can bridge knowledge gaps and facilitate interdisciplinary research, or do they risk introducing bias, reducing transparency, and undermining traditional scientific rigour? The panel will provide diverse perspectives on these questions.
We will also address the need for proper training, education, and guidelines to avoid misuse.

Programme: Thu, 1 May | Room E1

The oral presentations are given in a hybrid format supported by a Zoom meeting featuring on-site and virtual presentations. The button to access the Zoom meeting appears just before the time block starts.
Chairpersons: Jens Klump, Alice-Agnes Gabriel, Philip Stier
14:00–14:05
14:05–14:25
14:25–15:10

Speakers

  • Mariana Madruga de Brito, Helmholtz-Centre for Environmental Research, Germany
  • Paul Cleverley
  • Steven Zimmerman
  • Ahmed Elbanna, University of Illinois Urbana Champaign, United States of America