https://phys.org/news/2024-08-junk-ai-scientific-publishing.html
Several examples of scientists using AI to write papers with AI
generated mistakes that passed peer review. I noted before that ChatGPT could be used to write the introductions of papers, sometimes, better
than the authors had done. One example of a figure manipulation
indicates that some authors are using it to present and discuss their
data. That seems crazy. ChatGPT doesn't evaluate the junk that it is given. It just basically summarizes what they feed into it on some subject. I used a graphic AI once. I asked it to produce a picture of
a chicken walking towards the viewer. It did a pretty good job, but
gave the chicken the wrong number of toes facing forward. Apparently
junk like that is making it into science publications.
With these examples it may be that one of the last papers that I
reviewed before retiring earlier this year was due to AI. It was a good introduction and cited the relevant papers and summarized what could be
found in them, but even though the authors had cited previous work doing
what they claimed to be doing, their experimental design was incorrect
for what they were trying to do. The papers they cited had done things correctly, but they had not. I rejected the paper and informed the
journal editor that it needed substantial rewrite for the authors to
state what they had actually done. What might have happened is that the researchers may have had an AI write their introduction, but it was for
what they wanted to do, and not for what they actually did. English was likely not the primary language for the authors, and they may not have understood the introduction that was written. If they had understood
the introduction, they would have figured out that they had not done
what they claimed to be doing. Peer review is going to have to deal
with this type of junk. The last paper that I reviewed in March came
with instructions that the reviewers were not to use AI to assist them
with the review, but it looks like reviewers are going to need software
that will detect AI generated text.
Ron Okimoto
On 10/08/2024 22:32, RonO wrote:
https://phys.org/news/2024-08-junk-ai-scientific-publishing.html
Several examples of scientists using AI to write papers with AI
generated mistakes that passed peer review. I noted before that ChatGPT
could be used to write the introductions of papers, sometimes, better
than the authors had done. One example of a figure manipulation
indicates that some authors are using it to present and discuss their
data. That seems crazy. ChatGPT doesn't evaluate the junk that it is
given. It just basically summarizes what they feed into it on some
subject. I used a graphic AI once. I asked it to produce a picture of
a chicken walking towards the viewer. It did a pretty good job, but
gave the chicken the wrong number of toes facing forward. Apparently
junk like that is making it into science publications.
With these examples it may be that one of the last papers that I
reviewed before retiring earlier this year was due to AI. It was a good
introduction and cited the relevant papers and summarized what could be
found in them, but even though the authors had cited previous work doing
what they claimed to be doing, their experimental design was incorrect
for what they were trying to do. The papers they cited had done things
correctly, but they had not. I rejected the paper and informed the
journal editor that it needed substantial rewrite for the authors to
state what they had actually done. What might have happened is that the
researchers may have had an AI write their introduction, but it was for
what they wanted to do, and not for what they actually did. English was
likely not the primary language for the authors, and they may not have
understood the introduction that was written. If they had understood
the introduction, they would have figured out that they had not done
what they claimed to be doing. Peer review is going to have to deal
with this type of junk. The last paper that I reviewed in March came
with instructions that the reviewers were not to use AI to assist them
with the review, but it looks like reviewers are going to need software
that will detect AI generated text.
Ron Okimoto
I can understand why journals would not want to authors to use AI in
writing papers*, but why would they not want reviewers to use AI tools
if they can assist in reviewing the paper?
* Even so, the AI rubric includes translation tools (authors might write
text in their native language, and use AI for a first pass translation
into English), and the spelling/grammar/style checker Grammerly now
includes AI features.
On 8/11/2024 11:25 AM, Burkhard wrote:
On Sun, 11 Aug 2024 10:29:40 +0000, Ernest Major wrote:
On 10/08/2024 22:32, RonO wrote:
https://phys.org/news/2024-08-junk-ai-scientific-publishing.html
Several examples of scientists using AI to write papers with AI
generated mistakes that passed peer review. I noted before that
ChatGPT
could be used to write the introductions of papers, sometimes, better
than the authors had done. One example of a figure manipulation
indicates that some authors are using it to present and discuss their
data. That seems crazy. ChatGPT doesn't evaluate the junk that it is >>>> given. It just basically summarizes what they feed into it on some
subject. I used a graphic AI once. I asked it to produce a picture of >>>> a chicken walking towards the viewer. It did a pretty good job, but
gave the chicken the wrong number of toes facing forward. Apparently >>>> junk like that is making it into science publications.
With these examples it may be that one of the last papers that I
reviewed before retiring earlier this year was due to AI. It was a
good
introduction and cited the relevant papers and summarized what could be >>>> found in them, but even though the authors had cited previous work
doing
what they claimed to be doing, their experimental design was incorrect >>>> for what they were trying to do. The papers they cited had done things >>>> correctly, but they had not. I rejected the paper and informed the
journal editor that it needed substantial rewrite for the authors to
state what they had actually done. What might have happened is that
the
researchers may have had an AI write their introduction, but it was for >>>> what they wanted to do, and not for what they actually did. English
was
likely not the primary language for the authors, and they may not have >>>> understood the introduction that was written. If they had understood >>>> the introduction, they would have figured out that they had not done
what they claimed to be doing. Peer review is going to have to deal
with this type of junk. The last paper that I reviewed in March came >>>> with instructions that the reviewers were not to use AI to assist them >>>> with the review, but it looks like reviewers are going to need software >>>> that will detect AI generated text.
Ron Okimoto
I can understand why journals would not want to authors to use AI in
writing papers*, but why would they not want reviewers to use AI tools
if they can assist in reviewing the paper?
* Even so, the AI rubric includes translation tools (authors might write >>> text in their native language, and use AI for a first pass translation
into English), and the spelling/grammar/style checker Grammerly now
includes AI features.
If any of you are in Edinburgh right now, I'm on a panel on
this topic at the International Bookfestival, presenting the outcome
of two research projects we had on this, and some workshops
with publishers.
https://www.edbookfest.co.uk/the-festival/whats-on/page-against-the-machine >>
I'm on the more relaxed side on this myself, and agree in particular
with
Ernest that nobody worries about some routine tasks like spell-checking
(translation raises some really interesting issues "at the margins" -
e.g got some pushback when publishing in its latest list of languages
also
Romani, without checking with the community, and many are unhappy as
they
considered the "quasi-secret" nature of the language a historical
survival tool)
Very interesting questions also on the copyright for translations etc
For the use by academics, it often depends on the details. GenAI is a
glorified autocomplete tool, keep that in mind and you'll be fine. So
helping
write the review, once you decide on the content, is much less of
an issue than outsourcing the actual analysis eg.
And be aware of hallucinations... as some lawyers found to
their detriment when they submitted files to the court that had made-
up precedents in them
One recent paper that I recall reading indicated that AI halucinations resulted from feeding the AI, AI generated summaries. The AI started
making things up when it had to deal with AI generated material.
Ron Okimoto
On 8/11/2024 8:09 PM, JTEM wrote:
RonO wrote:
Peer review has it's flaws, but there is absolutely no doubt that it is
the best means we have for giving research it's first pass evaluation.
It's irredeemably flawed. There needs to be transparency.
The biggest danger, and it does happen, is good science being killed
off by "Peer Review."
You are just delusional.
There are so many journals publishing similar science that peer
review is about the last thing that is going to kill off good science.
The current situation is that there are journals damaging the integrity
of the science by being paper mills, and publishing junk if the authors
are willing to pay them.
How to stop it? Transparency. Let the rejected papers see the light
of day.
When I review a paper, I always check the box that gives the journal
the right to name me as one of the reviewers, and to forward my reviews
to other journals if they think that the paper would be better suited
to those journals, when journals have that policy. My recollection is
that pretty much all journals warn reviewers about reviewing papers
where they have a conflict of interest, and pretty much all of them
have the reviewers claim no conflict.
There really are so many journals at this time, that the suppression
that you claim, just doesn't exist.
Bad junk gets rejected from all legitimate journals.
Peer review can be manipulated (Sternberg and Meyer), and groups of
researchers have been exposed for recommending each others papers for
peer review (some journals ask the authors to recommend possible peer
reviewers in their field).
Less concerned about bad science making it through. Science is self
correcting. Science is repeatable or it isn't science. We can
reasonably expect garbage to self correct. But the opposite isn't
true. Good science that is kept from seeing the light of day is a
loss to the world.
Science is not narrowly focused, and quite dispersed with many journals publishing similar science. The fact that science is self correcting
is the reason that you don't have to worry about peer review. Things
that aren't worth publishing get published all the time. They just get buried in the junk pile, and do not get noticed. My guess is that the
rate of rejection is pretty low for most journals. I was an associate
editor for around a decade (off and on) since the 1990's, and have
reviewed papers from a wide range of journals, and not just that one,
and I have only outright rejected 2 papers, all the rest were sent back
for revision, and most were eventually accepted.
Ron Okimoto
On 8/17/24 1:15 PM, RonO wrote:
On 8/17/2024 10:42 AM, JTEM wrote:He neither knows nor cares. Ignore the chump.
RonO wrote:
This seems
You. Could. Not. Raise. A. Single. Objection.
Not one.
Not even after I humiliated you for it.
I already did, but you have just removed the material and kept plodding
on with the same misdirection. Even you can't figure out what you are
responding to in this post let alone the previous posts.
Ron Okimoto
| Sysop: | Keyop |
|---|---|
| Location: | Huddersfield, West Yorkshire, UK |
| Users: | 376 |
| Nodes: | 16 (3 / 13) |
| Uptime: | 26:30:51 |
| Calls: | 8,036 |
| Calls today: | 6 |
| Files: | 13,034 |
| Messages: | 5,829,399 |