Re: Hallucinated citations are polluting the scientific literature. What can be done?

On Mon, 18 May 2026 18:37:21 -0600, phoenix <[email protected]>
wrote:

>Steve Hayes wrote:
>> A human author understands human emotions and can make a better guess
>> at human reactions. LLms have no empathy, so can't do this. Someone
>> says something that most people would see as insulting, but the bot
>> writes that the hearer took comfort from it:
>> 
>> "I don't feel brave."
>> "You don't look it either."
>> 
>> LLMs can be no better than the masterial they work with. They are good
>> with factual material, but sometimes have difficulty in distinguishing
>> between facts and factoids. Of course humans have the same difficulty:
>> as you say, ask yourself if you'd avoid the same mistake if you didn't
>> have the knowledge that you do.
>> 
>The reason I feel comfortable continuing to speak is because I have 
>attended seminars on AI and I see no credentials among the others 
>speaking here.
>
>If I were to engage in an AI novel, I would talk back and forth a bit in 
>order to design a framework of maybe ten chapters.
>
>Then, I might possibly jumble it mildly just so any preconceived notions 
>the AI had built about it finishing it were dashed. This is in the 
>manner of someone who jumbled a Rubik's cube for me, they jumbled it 
>real good and I had to drop back to techniques that I had learned at the 
>beginning but not used recently.
>
>Finally I would feed that framework into the AI and tell it to flesh it 
>out into the book.
>
>What I see you doing wrong is allowing the AI to construct the plot, and 
>the climax and the finish. The more you have it do on its own the more 
>generic it will get and will be just like other users engaging in 
>"slopnoveling."

Well yes.

What I wanted to see was whether so-called AI could write an academic
paper or a novel, as some have claimed it can. I wanted to see if AI
could construct the plot, the climax and the finish, and have, like
you, concluded that using these tools like that is "doing wrong". 

>YMMV but if you want a really good book, do the hard work and then edit 
>it, but this is probably too much for the get rich scheme being pitched.
>
>2. Rinse.
>3. Repeat.

Exactly. 

An LLM bot isn't going to do it for you. GIGO.

A bit earlier I used an LLM bot to generate a specification for a
cover illustration for a book I had written. 

To do that I described the physical characteristics of the main
characters, and after 8 or 9 iterations it had produced a result that
was close enough. 

I still haven't found an artist competent enough and willing enough to
try, but I think the graphis specification produced by the bot could
be useful, and the non-visual characteristics were included in words.
If interested, you can see the result here:

<https://methodius.blogspot.com/2025/01/fantasy-adventure-stories-for-kids.html>

I have also found LLM bots useful to gather relevant information about
a particular topic from widely scattered sources. NOT to *write* an
academic paper but to give me necessary background information in
order to *read* one. 













-- 
Steve Hayes from Tshwane, South Africa
Web:  http://www.khanya.org.za/stevesig.htm
Blog: http://methodius.blogspot.com
E-mail - see web page, or parse: shayes at dunelm full stop org full stop uk