/AI13h ago

Academics Debate AI-Generated Text in Research Papers

835324.6K
David Head@DavidHeadPhd

This is the attitude I didn't know was widespread. Writing as totally instrumental. No wonder people are eager to offload it onto a machine.

5:02 AM · Jun 7, 2026 · 3K Views
Sentiment

Negative users dismissed the advocate supporting fully AI-written papers with disclosure by accusing them of being an LLM bot themselves.

Pos
0.0%
Neg
100.0%
1 comments with sentiment.
Cluster Engagement
Posts from X
Most Activity
Most Activity
VIEWS1.6KBOOKMARKS1LIKES15REPLIES2RETWEETS2
N8 Programs@N8Programs

Despite by pangram-crusading, I have no issue w/ academic research papers being fully AI-written, provided their is disclosure. My problem is solely with pasing AI writing off as your own - it's like having a ghostwriter do your paper and refusing to credit them.

12hViews 1.6KLikes 15Bookmarks 1
N8 Programs@N8Programs

@astralarkitekt (i detect this response may be AI)

Sorry, "ghostwriter" was sloppy - I meant *writer*, and I typed *ghostwriter* cause ghostwriters do what I'm describing.

8hViews 13Likes 1
Nicholas Hyperion@nickhistgeek

@N8Programs Agree w/ the disclosure take. These detection tools have awful false positive rates. Bad foundation for any threshold.

11hViews 24
Astral Arkitekt@astralarkitekt

I genuinely agree with you about disclosure being appropriate - but... you used "ghostwriter" and "credit" in the same sentence. Credit to a ghostwriter is antithetical to their purpose. Thus the critical "ghost" sticking-point of the term. The intent with using ghostwriters is NOT to credit the writer so that the contracting "author" CAN pass things off as their own. What you've stated that you are disagreeing with has nothing to do with AI and everything to do with merit, due-credit and transparency. This is not an AI problem. It's a problem as old as publishing.

And that's a tough one to work through, imo, because then you get into biases and gatekeeping before you've ever considered the potential merits of the work.

8hViews 16
N8 Programs@N8Programs

@astralarkitekt claude is rubbing of on all of us we all use 'genuinely' these days LOL. if anything, my error above is a good example of why human-intuition for AI detection shouldn't be used asa discriminating factor (though the limited type of anecevidence that means very little).

8hViews 3Likes 1
Astral Arkitekt@astralarkitekt

@N8Programs Sorry, no AI here, just a thinkin cap

8hViews 4
Robert E. Wright@robertewright

@DavidHeadPhd Actually heard multiple Ph.D.s call research and writing "work."

10hViews 8
Astral Arkitekt@astralarkitekt

@N8Programs You're right to push back. "Genuinely" is one of those charged words these days 😜

I wonder sometimes what a proper solution to this will be like once we get it figured out. The tools and tech we have today aren't very good.

8hViews 5
Brumafriend@Brumafriend

@nickhistgeek @N8Programs Pangram doesn't + you are an LLM bot

8hViews 4