Minerva and the parrot

Generative IA and linguistic model in the digital humanist laboratory

Abstract

Since its release by OpenAI last autumn, the ChatGPT linguistic interaction system based on the GPT-4 linguistic model has risen to the headlines and debate, both in the specialist and the public domain. In this paper, after examining its principles, I intend to explore the possible effects of this class of systems on scientific research practices in the humanities and, in particular, in the digital or computational humanities. The thesis I propose is that ChatGPT, already in the state it is in today, 1) represents an excellent support tool for the design and implementation of research workflows; 2) could have a significant impact on the quality and effectiveness of digital/computational humanities research.