Too much of it is fraudulent. It is a problem; in the short term there is a lot of fraud. In the long run... Fraudulent research is 'destroying trust in science'
Organized networks are infiltrating the academic publishing system to promote fake science, say experts investigating research fraud. A new study highlights the major challenge for modern science.
Researchers get rewarded for publishing. They get to add to their curriculum vitae, or CV. It is basically the academic's résumé. They get grants. They get to keep their jobs. They get hired for better positions. It is not in their best interests to pursue the truth; it is in their best interests to publish anything they can.
Since they have to publish, they publish a lot of schlock.
A new study has found that networks of bad actors work together to publish bogus research. The findings, published in the journal PNAS this week, came from analyzing more than 5 million scientific articles published across 70,000 journals.
"There are groups of editors conspiring to publish low-quality articles, at scale, escaping traditional peer review processes," said the study's lead author Reese Richardson, a social scientist at Northwestern University in the US.
And this isn't just a problem for the Universities.
For example, researchers found evidence of image manipulation in a landmark study about Alzheimer's disease. The paper was eventually retracted and the lead scientist resigned, but [Anna] Abalkina [a social scientist at the Free University of Berlin] said billions of dollars in research funding and years of research had already been invested from one bad study.
So when someone tells you to "trust the science," ask them exactly what science they are talking about, since a lot of it is fraudulent.
Click thru. It isn't a long article.
It is important to note that despite the cuts from D.O.G.E, the US government does pay for a lot of scientific research, and as you can see from the example with Alzheimer's disease, it does impact a lot of things. Years of research, and billions of dollars wasted because of a "manipulated" image and a bad study.
You and I have been talking about this for about as far back as I can recall. As is always the case, a couple of amateurs who aren't in the network can't affect it.
ReplyDeleteI think this might be the "one in a million" situation where AI can do something positive. Once the examples of how badly AI screws up start getting spread around, people are going to start ignoring anything that's based on an AI study. As it stands now, people have started counting fingers in any picture of people, because of how badly the early AI image creators did with that, so that now if I see any carefully composed and technically well-executed photograph, I'll see someone commenting to ignore it because it's an AI image.
They seem to have the fingers thing figured out, because some clearly AI-generated ads have started having characters who hold up their hands, so you can count the fingers.
DeleteI don't have faith in AI the way a lot of people do, because LLMs are not good at truth. Maybe they will get better, but they still even fail at telling you if a given number is prime, even though there is a well established algorithm