Login to account Create an account  


  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Damnable research fraud - AI is making it even worse
#1
I offer a pair of content sources here... Both expound on the massively growing research fraud cases in the global scientific community.  The first is a seven- or eight-minute video which treats the subject with some clarity.
 
Alarming: Fraud spreads in Science -- and I fear it will become worse
(with my apologies for the embedded advertisement - it's brief enough though)

In the video, hostess Sabine Hossenfelder explains that the enormous explosion in the number of published research papers being retracted after seem to share a few commonalities which don't speak well of the people who are responsible for reviewing scientific research... meaning many may have been bribed into authorizing papers for publication.  Further, the insipid practice of off-loading the burden of review to 'third party' contracted 'editors' made it possible for the fraud to be carried out without any consequences for the perpetrators.

Also, artificial intelligence tools are used to generate repurposed data manipulated to "pass initial muster" which creates career benefits for any scientist willing to pay money to claim the 'credit' they ostensibly receive from having "published."  Sometimes it is potentially detectable because AI is still clumsy at 'rewording' things - and some examples are shared (like rewording breast cancer into "bosom peril" or kidney failure into "renal disappointment, et. al.)

She also explained how, up until recently, the bulk of the fraud seemed to be coming from the eastern hemisphere, and eastern Europe.  Apparently, the publishing firm Wiley's subsidiary named Hindawi is a volume producer of these kinds of fraudulent papers.

Second on my list, is an article from the Royal Chemistry Societies online reporting in Chemistry World, which focuses on China's extraordinary situation...,

China conducts nationwide audit of research misconduct after thousands of papers retracted last year
 

...
China has the highest retraction rate globally, including conference papers, exceeding 20 per 10,000 articles. Hindawi, a subsidiary of Wiley, retracted over 8000 articles involving a Chinese co-author in 2023, a record-breaking year for retractions that resulted in Wiley announcing that it would stop using the Hindawi brand and fold the journals into the rest of its portfolio.\
...
The Chinese government is worried that the high number of retractions is damaging its academic reputation. In 2022, China was the world leader in scientific research output and ‘high impact’ studies, accounting for 23.4% of the world’s research output, according to a report published by Japan’s science and technology ministry. However, analysis conducted by Nature revealed that there were almost 17,000 retraction notices on papers with China-affiliated authors since January 2021.
...


We need to be concerned.  Scientific integrity has suffered a lot of damage in the past years... and this stuff isn't helping one bit.  Many of these papers are being cited in current research... how exactly, I can't imagine, given that they are practically fiction... but somehow researchers (even the legitimate ones) don't seem to see it until it's too late.
Reply
#2
I look to see if it was repeated with the same results, with those that could not be repeated exactly then it's suspect. There's really not much more to it than that IMO.
"The real trouble with reality is that there is no background music." Anonymous

Plato's Chariot Allegory
Reply
#3
(02-18-2024, 10:42 PM)quintessentone Wrote: I look to see if it was repeated with the same results, with those that could not be repeated exactly then it's suspect. There's really not much more to it than that IMO.

I can't imagine how disappointing it must be for a scientist working on a research project to discover that some cited sources turned out to be fraudulent.  It must be even worse than finding one of your data sources has been simply plugging in useless make-believe data that 'satisfies' the PI's hypothesis, but I imagine that usually gets called out before publication... or at least I would hope it does.

Some folks have no shame.
Reply



Possibly Related Threads…
Thread Author Replies Views Last Post
  Newborn's T cells aren't better (or worse) Maxmars 1 250 02-24-2024, 07:36 PM
Last Post: Nugget
  Cloned rhesus monkey created to speed medical research putnam6 8 407 01-20-2024, 02:05 PM
Last Post: putnam6