SORTEE Webinar: The Risks and Rewards of Large Language Models in (Open) Scientific Research

The integration of Large Language Models (LLMs) into the scientific research workflow presents a dual-edged sword. The promise of rapid and automated workflows also introduces the risk of introducing methodological  errors with potentially far reaching consequences for our global state of knowledge within scientific fields. This talk aims to critically examine both the technical utility along with the inherent challenges of using LLMs as a research tool. Here I will present a series of case studies on the use of LLMs, weighing up both the pros (utility) and cons (challenges) in these different scenarios. Additionally it is important that we open a discussion on the role of LLMs within an open science framework. How can we ensure transparency, reproducibility, and equitable access when core components of our research is generated by (oftentimes) proprietary ‘black-box’ tools? This session aims to foster a balanced perspective on  LLMs and open a discussion on how we can facilitate the responsible adoption thereof.

Speaker:

Dr. Tanya Strydom, Ecology and Evolutionary Biology, School of Biosciences, University of Sheffield, UK

Requirements:

None. The content of the talk will mostly focus on the general use of LLM’s in general research workflows as opposed to specific methodological advances in any one field.