HIDA Lecture Series on AI and Large Language Models
Speaker: Dr. Jörg Pohle, Head of Research Programme: Data, actors, infrastructures: The governance of data-driven innovation and cyber security, Alexander von Humboldt Institute for Internet and Society, Berlin
Date: 14.02.2024, 11 am
Title: Ally or Adversary: Examining the Impact of Large Language Models on Academics and Academic Work
Abstract
The unveiling of ChatGPT sparked extensive discussions on its implications for science. This lecture will delve into the dynamic interplay between Large Language Models (LLMs) and the academic landscape, exploring their potential benefits as well as the risks they pose to the science system, academic profession and scholarly practice. Through the examination of empirical data, ethical questions concerning this technology and its use will be unpacked, in addition to the ways LLMs may contribute to the creative, analytical and administrative tasks in academic work. Through this discussion, we will reflect on broader questions concerning the changing role of academics in the increasingly digitalized world, the increasing datafication and quantification in how scientists, their work as well as their institutions are assessed and evaluated, and what should be the mission of academics in the future that deserves appreciation.
Dr. Jörg Pohle
Jörg Pohle is a postdoc and head of the research programme “Data, actors, infrastructures: The governance of data-driven innovation and cyber security” at the Alexander von Humboldt Institute for Internet and Society in Berlin. He is a computer scientis by training, with a background in law, political science and sociology. He received his doctorate from Humboldt-Universität zu Berlin for a dissertation on the history and theory of data protection and implications on the design of ICT systems. His research centres on societal negotiations in the field of surveillance, IT security, privacy and data protection, and their implementation in technology and organizational practice.
In cooperation with
