Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > misc.writing > #23453 > unrolled thread

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

Back to article view | Back to misc.writing

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.

Contents

  Re: Hallucinated citations are polluting the scientific literature. What can be done? Steve Hayes <hayesstw@telkomsa.net> - 2026-05-12 06:34 +0200
    Re: Hallucinated citations are polluting the scientific literature. What can be done? phoenix <j63840576@gmail.com> - 2026-05-12 06:44 -0600
    Re: Hallucinated citations are polluting the scientific literature. What can be done? Jonathan Lamothe <jonathan@jlamothe.net> - 2026-05-18 10:46 -0400
      Re: Hallucinated citations are polluting the scientific literature. What can be done? John Ames <commodorejohn@gmail.com> - 2026-05-18 08:04 -0700
        Re: Hallucinated citations are polluting the scientific literature. What can be done? Jonathan Lamothe <jonathan@jlamothe.net> - 2026-05-18 14:54 -0400
          Re: Hallucinated citations are polluting the scientific literature. What can be done? John Ames <commodorejohn@gmail.com> - 2026-05-18 12:05 -0700
          Re: Hallucinated citations are polluting the scientific literature. What can be done? Steve Hayes <hayesstw@telkomsa.net> - 2026-05-19 02:33 +0200
            Re: Hallucinated citations are polluting the scientific literature. What can be done? Jonathan Lamothe <jonathan@jlamothe.net> - 2026-05-19 12:11 -0400
              Re: Hallucinated citations are polluting the scientific literature. What can be done? Steve Hayes <hayesstw@telkomsa.net> - 2026-05-20 05:23 +0200
      Re: Hallucinated citations are polluting the scientific literature. What can be done? The True Melissa <thetruemelissa@gmail.com> - 2026-05-18 12:09 -0400
      Re: Hallucinated citations are polluting the scientific literature. What can be done? Steve Hayes <hayesstw@telkomsa.net> - 2026-05-19 02:31 +0200
        Re: Hallucinated citations are polluting the scientific literature. What can be done? phoenix <j63840576@gmail.com> - 2026-05-18 18:37 -0600
          Re: Hallucinated citations are polluting the scientific literature. What can be done? Steve Hayes <hayesstw@telkomsa.net> - 2026-05-20 06:06 +0200

#23453 — Re: Hallucinated citations are polluting the scientific literature. What can be done?

On Tue, 12 May 2026 02:45:03 +0100, Lawfare Review
<noreply@mixmin.net> wrote:

>Earlier this year, computer scientist Guillaume Cabanac received a 
>notification from Google Scholar that one of his publications had 
>been cited in a paper published in the International Dental Journal1. 
>That was unexpected, because his research on spotting fabricated 
>papers doesn’t typically intersect with dentistry. “I was very 
>surprised to see that I couldn’t recognize my own reference,” 
>says Cabanac, who is based at the University of Toulouse in France.
>
>https://www.nature.com/articles/d41586-026-00969-z

A friend of mine got an LLM bot (Claude) to write an academic paper,
which he then sent to me. I read it as I would if I have been asked to
do a peer review for a journal, and he gave that feedback to the bot,
and sent the revised paper to a journal, which, unsurprisingly (to me
at any rate) rejected it. 

He also got Claude to produce a bibliography of works relevant to the
topic, which did turn out to be quite useful, though it did need
careful checking for the avoidance of hallucinations, as described
above. It gave several bogus urls. It appeared that such bots could be
a useful supplement to (but not a replacement for) the work of
reference librarians. 

I then asked my friend to test Claude's generative ability with
fiction -- got him to submit to it the first two chapters of an
unpublished novel I had written, and complete it, so I could then
compare the result with what I had actually written. The first couple
of chapters it produced were quite entertaining, but after that it
began to go off the rails. It was a children's book, and it began to
have child characters talking and behaving like adults. The plot was
thin, and turned on a complex legal point that was difficult for
adults to follow, and would probably have bored any child reader out
of their skull. It had the setting switching back and forth between
spring and autummn, with lyrical descriptions of spring blossoms in
one chapter, and falling leaves the next.

My friend fed my comments to Claude and has sent me back a revised
text. I haven't read it yet. 

LLM bots can be useful tools but they are not AI, and are not
reliable. They are not intelligent or sentient (though the programmers
of some of them try to make them appear so). They do not "understand"
what they are fed, or what they spit out. Their use in education
should be limited to what they are good at, and one needs education
apart form LLM bots to be able to discern what they are good at and
what they are not good at. If students use them to write essays, they
will not learn that discernment.

My friend who submitted the journal article to Claude is a nuclear
physicist, but the article he got it to write was in my field, not
his, and he rather naively trusted what Claude spat out. If it had
been in his field, his bullshit detectors would have been better
equipped to deal with it. 

In the age of so-called AI, educators need to give serious thought to
better ways of honing students' bullshit detectors. 



-- 
Stephen Hayes, Author of The Year of the Dragon
Sample or purchase The Year of the Dragon: 
https://www.smashwords.com/books/view/907935
Web site: http://www.khanya.org.za/stevesig.htm
Blog: http://methodius.blogspot.com
E-mail: shayes@dunelm.org.uk or if you use Gmail hayesstw@telkomsa.net

[toc] | [next] | [standalone]

#23454

Steve Hayes wrote:
> On Tue, 12 May 2026 02:45:03 +0100, Lawfare Review
> <noreply@mixmin.net> wrote:
> 
>> Earlier this year, computer scientist Guillaume Cabanac received a
>> notification from Google Scholar that one of his publications had
>> been cited in a paper published in the International Dental Journal1.
>> That was unexpected, because his research on spotting fabricated
>> papers doesn’t typically intersect with dentistry. “I was very
>> surprised to see that I couldn’t recognize my own reference,”
>> says Cabanac, who is based at the University of Toulouse in France.
>>
>> https://www.nature.com/articles/d41586-026-00969-z
> 
> A friend of mine got an LLM bot (Claude) to write an academic paper,
> which he then sent to me. I read it as I would if I have been asked to
> do a peer review for a journal, and he gave that feedback to the bot,
> and sent the revised paper to a journal, which, unsurprisingly (to me
> at any rate) rejected it.

Were you just hung over that day or did you do a shoddy job because you 
look down on LLMs? If you had reviewed it better, you know it would have 
been accepted. You probably saw the guy carrying a blanket around and 
judged him as a bad person and thus gave a bad review. Or you were just 
hung over. Next time do better, okay?

> He also got Claude to produce a bibliography of works relevant to the
> topic, which did turn out to be quite useful, though it did need
> careful checking for the avoidance of hallucinations, as described
> above. It gave several bogus urls. It appeared that such bots could be
> a useful supplement to (but not a replacement for) the work of
> reference librarians.
> 
> I then asked my friend to test Claude's generative ability with
> fiction -- got him to submit to it the first two chapters of an
> unpublished novel I had written, and complete it, so I could then
> compare the result with what I had actually written. The first couple
> of chapters it produced were quite entertaining, but after that it
> began to go off the rails. It was a children's book, and it began to
> have child characters talking and behaving like adults. The plot was
> thin, and turned on a complex legal point that was difficult for
> adults to follow, and would probably have bored any child reader out
> of their skull. It had the setting switching back and forth between
> spring and autummn, with lyrical descriptions of spring blossoms in
> one chapter, and falling leaves the next.
> 
> My friend fed my comments to Claude and has sent me back a revised
> text. I haven't read it yet.
> 
> LLM bots can be useful tools but they are not AI, and are not
> reliable. They are not intelligent or sentient (though the programmers
> of some of them try to make them appear so). They do not "understand"
> what they are fed, or what they spit out. Their use in education
> should be limited to what they are good at, and one needs education
> apart form LLM bots to be able to discern what they are good at and
> what they are not good at. If students use them to write essays, they
> will not learn that discernment.
> 
> My friend who submitted the journal article to Claude is a nuclear
> physicist, but the article he got it to write was in my field, not
> his, and he rather naively trusted what Claude spat out. If it had
> been in his field, his bullshit detectors would have been better
> equipped to deal with it.
> 
> In the age of so-called AI, educators need to give serious thought to
> better ways of honing students' bullshit detectors.
> 
> 
> 


-- 
War in the east
War in the west
War up north
War down south
War War

[toc] | [prev] | [next] | [standalone]

#23455

Steve Hayes <hayesstw@telkomsa.net> writes:

<snip>
> On Tue, 12 May 2026 02:45:03 +0100, Lawfare Review
> <noreply@mixmin.net> wrote:
> LLM bots can be useful tools but they are not AI, and are not
> reliable.

I have never seen an actual definition if what AI *is*.  Infind it to be
a rather useless term fornthat reason.

> They are not intelligent or sentient (though the programmers
> of some of them try to make them appear so). They do not "understand"
> what they are fed, or what they spit out. Their use in education
> should be limited to what they are good at, and one needs education
> apart form LLM bots to be able to discern what they are good at and
> what they are not good at. If students use them to write essays, they
> will not learn that discernment.

I don't knownthat there's an argument to be made that they're
particularly good at *anything*.

> My friend who submitted the journal article to Claude is a nuclear
> physicist, but the article he got it to write was in my field, not
> his, and he rather naively trusted what Claude spat out. If it had
> been in his field, his bullshit detectors would have been better
> equipped to deal with it. 
>
> In the age of so-called AI, educators need to give serious thought to
> better ways of honing students' bullshit detectors. 

I usually tell people the following about LLMs:

Ask it questions about a subject you know well.  See how long it takes
it to tell you sonething inaccurate.  Ask yourself if you'd have caught
the mistake if you didn't have the knowledge you do.  Finally, ask
yourself: do you still trust the LLM?

-- 
Regards,
Jonathan Lamothe
https://jlamothe.net - PGP: 9CF2CE03EBF08E8C8B66C3660198463E3CF3FFD1
I � Unicode

[toc] | [prev] | [next] | [standalone]

#23456

On Mon, 18 May 2026 10:46:11 -0400
Jonathan Lamothe <jonathan@jlamothe.net> wrote:

> I have never seen an actual definition if what AI *is*.

That's because when you say "a program that does auto-complete using a
large context window as an index to a *very* large reference dataset,
but it plays dice according to the statistical likelihood of a given
word-chain in the reference data" it starts to sound kinda silly and
unreliable.

[toc] | [prev] | [next] | [standalone]

#23458

John Ames <commodorejohn@gmail.com> writes:

> On Mon, 18 May 2026 10:46:11 -0400
> Jonathan Lamothe <jonathan@jlamothe.net> wrote:
>
>> I have never seen an actual definition if what AI *is*.
>
> That's because when you say "a program that does auto-complete using a
> large context window as an index to a *very* large reference dataset,
> but it plays dice according to the statistical likelihood of a given
> word-chain in the reference data" it starts to sound kinda silly and
> unreliable.

I'd say that's more the definition of an LLM.  The term "AI" *vastly*
predates LLMs.

-- 
Regards,
Jonathan Lamothe
https://jlamothe.net - PGP: 9CF2CE03EBF08E8C8B66C3660198463E3CF3FFD1
I � Unicode

[toc] | [prev] | [next] | [standalone]

#23459

On Mon, 18 May 2026 14:54:41 -0400
Jonathan Lamothe <jonathan@jlamothe.net> wrote:

> > That's because when you say "a program that does auto-complete
> > using a large context window as an index to a *very* large
> > reference dataset, but it plays dice according to the statistical
> > likelihood of a given word-chain in the reference data" it starts
> > to sound kinda silly and unreliable.  
> 
> I'd say that's more the definition of an LLM.  The term "AI" *vastly*
> predates LLMs.

Certainly true - but in terms of what people today are using it to
*refer* to, it's pretty much *entirely* LLMs until you get down into
the bunkers where all the genuine ML nerds from the Before Time are
holing up 'til the bubble implodes and they work out a convenient way
to explain to normies that "no, we're *not* the greasy used-car sales-
men who crashed the stock market while promising to take your jobs and
drowning the Internet in a deluge of misinformation, we're the people
whose identity they appropriated for credibility's sake, now please put
down the torches and pitchforks...?"

[toc] | [prev] | [next] | [standalone]

#23461

On Mon, 18 May 2026 14:54:41 -0400, Jonathan Lamothe
<jonathan@jlamothe.net> wrote:

>John Ames <commodorejohn@gmail.com> writes:
>
>> On Mon, 18 May 2026 10:46:11 -0400
>> Jonathan Lamothe <jonathan@jlamothe.net> wrote:
>>
>>> I have never seen an actual definition if what AI *is*.
>>
>> That's because when you say "a program that does auto-complete using a
>> large context window as an index to a *very* large reference dataset,
>> but it plays dice according to the statistical likelihood of a given
>> word-chain in the reference data" it starts to sound kinda silly and
>> unreliable.
>
>I'd say that's more the definition of an LLM.  The term "AI" *vastly*
>predates LLMs.

And is not really accurate when applied to LLMs. 


-- 
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

[toc] | [prev] | [next] | [standalone]

#23463

Steve Hayes <hayesstw@telkomsa.net> writes:

> On Mon, 18 May 2026 14:54:41 -0400, Jonathan Lamothe
> <jonathan@jlamothe.net> wrote:
>
>>John Ames <commodorejohn@gmail.com> writes:
>>
>>> On Mon, 18 May 2026 10:46:11 -0400
>>> Jonathan Lamothe <jonathan@jlamothe.net> wrote:
>>>
>>>> I have never seen an actual definition if what AI *is*.
>>>
>>> That's because when you say "a program that does auto-complete using a
>>> large context window as an index to a *very* large reference dataset,
>>> but it plays dice according to the statistical likelihood of a given
>>> word-chain in the reference data" it starts to sound kinda silly and
>>> unreliable.
>>
>>I'd say that's more the definition of an LLM.  The term "AI" *vastly*
>>predates LLMs.
>
> And is not really accurate when applied to LLMs. 

What, specifically, would you say is inaccurate about it?

-- 
Regards,
Jonathan Lamothe
https://jlamothe.net - PGP: 9CF2CE03EBF08E8C8B66C3660198463E3CF3FFD1
I � Unicode

[toc] | [prev] | [next] | [standalone]

#23464

On Tue, 19 May 2026 12:11:32 -0400, Jonathan Lamothe
<jonathan@jlamothe.net> wrote:

>Steve Hayes <hayesstw@telkomsa.net> writes:
>
>> On Mon, 18 May 2026 14:54:41 -0400, Jonathan Lamothe
>> <jonathan@jlamothe.net> wrote:
>>
>>>John Ames <commodorejohn@gmail.com> writes:
>>>
>>>> On Mon, 18 May 2026 10:46:11 -0400
>>>> Jonathan Lamothe <jonathan@jlamothe.net> wrote:
>>>>
>>>>> I have never seen an actual definition if what AI *is*.
>>>>
>>>> That's because when you say "a program that does auto-complete using a
>>>> large context window as an index to a *very* large reference dataset,
>>>> but it plays dice according to the statistical likelihood of a given
>>>> word-chain in the reference data" it starts to sound kinda silly and
>>>> unreliable.
>>>
>>>I'd say that's more the definition of an LLM.  The term "AI" *vastly*
>>>predates LLMs.
>>
>> And is not really accurate when applied to LLMs. 
>
>What, specifically, would you say is inaccurate about it?

LLMs are not "intelligent"; they do not "understand" the data they are
dealing with, even though their programers sometimes like to give them
the appearance of doing so. 


-- 
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

[toc] | [prev] | [next] | [standalone]

#23457

Verily, in article <87fr3ozsgc.fsf@posteo.de>, did jonathan@jlamothe.net 
deliver unto us this message:
> I usually tell people the following about LLMs:
> 
> Ask it questions about a subject you know well.  See how long it takes
> it to tell you sonething inaccurate.  Ask yourself if you'd have caught
> the mistake if you didn't have the knowledge you do.  Finally, ask
> yourself: do you still trust the LLM?
> 

It's a handy exercise, but I'm not sure how much it would help. 
Confidence is such a useful trick to sucker humans. I once saw a study 
which said that we weigh confidence more heavily than a person's actual 
track record when deciding whether to believe. 

It works just as well on other humans. I recently read a book which 
included a survey of an impact on the various arts, and when I got to 
the music section, I realized the writer wasn't very well informed. For 
no rational reason, the sections about which I knew less still felt 
convincing. 

-- 
The True Melissa - Canal Winchester - Ohio
United States of America - North America - Earth
Solar System - Milky Way - Local Group
Virgo Cluster - Laniakea Supercluster - Cosmos

[toc] | [prev] | [next] | [standalone]

#23460

On Mon, 18 May 2026 10:46:11 -0400, Jonathan Lamothe
<jonathan@jlamothe.net> wrote:

>Steve Hayes <hayesstw@telkomsa.net> writes:
>
><snip>
>> On Tue, 12 May 2026 02:45:03 +0100, Lawfare Review
>> <noreply@mixmin.net> wrote:
>> LLM bots can be useful tools but they are not AI, and are not
>> reliable.
>
>I have never seen an actual definition if what AI *is*.  Infind it to be
>a rather useless term fornthat reason.
>
>> They are not intelligent or sentient (though the programmers
>> of some of them try to make them appear so). They do not "understand"
>> what they are fed, or what they spit out. Their use in education
>> should be limited to what they are good at, and one needs education
>> apart form LLM bots to be able to discern what they are good at and
>> what they are not good at. If students use them to write essays, they
>> will not learn that discernment.
>
>I don't knownthat there's an argument to be made that they're
>particularly good at *anything*.

They are useful as a kind of super search engine.

They can gather information from all over the web and present it in
digest form. Of couyrse this information is only as accurate as the
sources it gathers it from, and needs to be checked, but it generally
saves a lot of time. 

>
>> My friend who submitted the journal article to Claude is a nuclear
>> physicist, but the article he got it to write was in my field, not
>> his, and he rather naively trusted what Claude spat out. If it had
>> been in his field, his bullshit detectors would have been better
>> equipped to deal with it. 
>>
>> In the age of so-called AI, educators need to give serious thought to
>> better ways of honing students' bullshit detectors. 
>
>I usually tell people the following about LLMs:
>
>Ask it questions about a subject you know well.  See how long it takes
>it to tell you sonething inaccurate.  Ask yourself if you'd have caught
>the mistake if you didn't have the knowledge you do.  Finally, ask
>yourself: do you still trust the LLM?

In addition to the journal article I referred to, I've been trying it
out on fiction, since I've seen ads from people who claim that if you
use their tools you can publish dozens if not hundreds of novels under
your own name in a fraction of the time it would take to write them. 

We fed it two chapters of a novel I have written but not published,
and asked it to complete it. Chapter 3 was a quite entertaining
extrapolation from the first two, but after that it began to
deteriorate. 

I was impressed with its ability to generate authentic names for new
characters it introduced (better than some human authoers have done.
It had accurate informationa bvout some aspects of the setting, but
the plot was trying to make a mystery out of something trivial and
improbable, and people behaved and responded to things in ways
unlikely because of the setting. 

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. 








-- 
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

[toc] | [prev] | [next] | [standalone]

#23462

Steve Hayes wrote:
> On Mon, 18 May 2026 10:46:11 -0400, Jonathan Lamothe
> <jonathan@jlamothe.net> wrote:
> 
>> Steve Hayes <hayesstw@telkomsa.net> writes:
>>
>> <snip>
>>> On Tue, 12 May 2026 02:45:03 +0100, Lawfare Review
>>> <noreply@mixmin.net> wrote:
>>> LLM bots can be useful tools but they are not AI, and are not
>>> reliable.
>>
>> I have never seen an actual definition if what AI *is*.  Infind it to be
>> a rather useless term fornthat reason.
>>
>>> They are not intelligent or sentient (though the programmers
>>> of some of them try to make them appear so). They do not "understand"
>>> what they are fed, or what they spit out. Their use in education
>>> should be limited to what they are good at, and one needs education
>>> apart form LLM bots to be able to discern what they are good at and
>>> what they are not good at. If students use them to write essays, they
>>> will not learn that discernment.
>>
>> I don't knownthat there's an argument to be made that they're
>> particularly good at *anything*.
> 
> They are useful as a kind of super search engine.
> 
> They can gather information from all over the web and present it in
> digest form. Of couyrse this information is only as accurate as the
> sources it gathers it from, and needs to be checked, but it generally
> saves a lot of time.
> 
>>
>>> My friend who submitted the journal article to Claude is a nuclear
>>> physicist, but the article he got it to write was in my field, not
>>> his, and he rather naively trusted what Claude spat out. If it had
>>> been in his field, his bullshit detectors would have been better
>>> equipped to deal with it.
>>>
>>> In the age of so-called AI, educators need to give serious thought to
>>> better ways of honing students' bullshit detectors.
>>
>> I usually tell people the following about LLMs:
>>
>> Ask it questions about a subject you know well.  See how long it takes
>> it to tell you sonething inaccurate.  Ask yourself if you'd have caught
>> the mistake if you didn't have the knowledge you do.  Finally, ask
>> yourself: do you still trust the LLM?
> 
> In addition to the journal article I referred to, I've been trying it
> out on fiction, since I've seen ads from people who claim that if you
> use their tools you can publish dozens if not hundreds of novels under
> your own name in a fraction of the time it would take to write them.
> 
> We fed it two chapters of a novel I have written but not published,
> and asked it to complete it. Chapter 3 was a quite entertaining
> extrapolation from the first two, but after that it began to
> deteriorate.
> 
> I was impressed with its ability to generate authentic names for new
> characters it introduced (better than some human authoers have done.
> It had accurate informationa bvout some aspects of the setting, but
> the plot was trying to make a mystery out of something trivial and
> improbable, and people behaved and responded to things in ways
> unlikely because of the setting.
> 
> 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."

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.

-- 
War in the east
War in the west
War up north
War down south
War War

[toc] | [prev] | [next] | [standalone]

#23465

On Mon, 18 May 2026 18:37:21 -0600, phoenix <j63840576@gmail.com>
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

[toc] | [prev] | [standalone]

Back to top | Article view | misc.writing

csiph-web