Journal Articles Nature Machine Intelligence Year : 2024

The dangers of using proprietary LLMs for research

Abstract

The release of ChatGPT at the end of 2022 has thrust large language models (LLMs) into the limelight. In this paper, we discuss the merits and limits of using proprietary LLMs for (social) scientific research.

Keywords

Fichier principal
Vignette du fichier
NMI_Resubmission_Nov30_2023.pdf (431.06 Ko) Télécharger le fichier
Origin Files produced by the author(s)

Dates and versions

hal-04825849 , version 1 (08-12-2024)

Identifiers

Cite

Étienne Ollion, Rubing Shen, Ana Macanovic, Arnault Chatelain. The dangers of using proprietary LLMs for research. Nature Machine Intelligence, 2024, 6 (1), pp.4-5. ⟨10.1038/s42256-023-00783-6⟩. ⟨hal-04825849⟩
35 View
39 Download

Altmetric

Share

More