Aug 26, 2024
Misinformational AI-Generated Academic Papers
Aaron Sandoval, Akash Kundu, Layla Adam
This study explores the potential for generative AI to produce convincing fake research papers, highlighting the growing threat of AI-generated misinformation. We demonstrate a semi-automated pipeline using large language models (LLMs) and image generation tools to create academic-style papers from simple text prompts.
Lucas Hansen
Very cool idea! This is solid progress towards auto-generation of research papers. I particularly liked (1) including figures & (2) starting with a many similar papers as a baseline.I think it would’ve been nice to include some research papers that you consider problematic as examples.
Épiphanie Gédéon
Good idea!I feel like it would have been better to generate a paper proving a point of view generated by the user, as the frontend only allows downloading the generated paper, which doesn’t prove automation.
Adam Binksmith
Nice work! Showing rendered LaTeX is a nice direct way to demonstrate LLMs’ ability to generate it. For next steps, I’d be interested to see a more interactive version of this demo, and more support for the user in understanding where the generated paper is or isn’t plausible.
Cite this work
@misc {
title={
@misc {
},
author={
Aaron Sandoval, Akash Kundu, Layla Adam
},
date={
8/26/24
},
organization={Apart Research},
note={Research submission to the research sprint hosted by Apart.},
howpublished={https://apartresearch.com}
}