ChatGPT Wrote This Paper, But I Helped

This page hosts the evolving drafts of my paper exploring how large language models (LLMs), like ChatGPT, are changing academic writing. The paper is both an analysis and a demonstration of human-AI collaboration.

📄 Drafts

Draft 1a (Initial AI-generated content) Draft 2 (Incorporated voice and tone) Draft 3 (Streamlined and revised) Draft 4 (Final pre-submission) Draft 4a (Submission sent to human colleagues for comments.) Submission to first journal

💬 Chat Log

Full ChatGPT Conversation that led to Draft 4a and the creation of this page

📔 Status Updates

March 27, 2025: Draft 4a ready for review. Sent to 7 colleagues in department, offered to send to AI Advisory Committee (9 members). Asked for feedback by April 7. 1 colleague has replied offering to provide a review (CS). Another (DS) sent a review and I made edits in response to their points.
April 8, 2025: Colleague CS replied with their comments. Biggest question was "How do you credit ChatGPT? Should it be an author?". Planning to add more on that point, since it's mentioned in the draft but isn't explored as much as it could be. Also going to point out in next draft that the point of ChatGPT as a "junior colleague that can review" is very important to those in small groups - I invited 16 people to review this manuscript and gave them 10 days to do so. Only 2 took me up on it, with 1 other mentioning that they had seen my email. The other 13 said nothing. Finding humans to help you review your academic writing can be challenging! Anyway, after integrating feedback, and a final polish, the paper was submitted to the first journal.