Boffins warn that AI paper mills are swamping science with garbage studies – Research flags rise in one-dimensional health research fueled by large language models
https://www.theregister.com/2025/05/13/ai_junk_science_papers/
Boffins warn that AI paper mills are swamping science with garbage studies – Research flags rise in one-dimensional health research fueled by large language models
https://www.theregister.com/2025/05/13/ai_junk_science_papers/
9 Comments
Well, color me shocked—AI churning out junk science? Who’d have thought letting machines play scientist would backfire. Guess we’ll need human brains to clean up the mess. Time to dust off the old peer review, folks. /oldtimer
Gosh, if only we could start checking that people who submit important scientific papers are real, exist and have a verifiable educational and scientific discourse record.
What’s the fecking point in “peer review” if you don’t even know that your peers that are reviewing even exist, let alone are human, let alone are a reliable and trusted expert in their particular field?
It’s time for CREDENTIALS of scientists to be public knowledge.
I hung around for a few years with a bunch of PhD graduate researchers, lecturers, etc. at some the world’s most prestigious universities. Turns out many of them often just cite each other’s papers even when they’re irrelevant and/or end up doing each other’s peer review because so few people work in that field in that country… which completely destroys the whole point of that system.
Thought boffins were birds like penguins or something. Was wondering why they were warning us about ai…
Wild how the same tech meant to speed up discovery is now flooding the system with junk. It’s not progress if no one can trust the results.
How many of them died to bring us this information?
Any moment now it’s going to be AGI and not just a DDOS on our society. Any moment now. Just give it a minute.
As-if the previously human submitted studies were not junk.
Funny. Because a lot of artists are feeling the same
Well this is depressing. Also sucks that 90% of the crap is coming out of China (according to the article — I’m not making this up). As an academic I already struggle with a negative perception of Chinese research. I automatically doubt papers with strings of Chinese authors, because (unfortunately) I’ve read so much dodgy stuff from there. I hate how this reputational damage impacts the genuinely brilliant researchers out there. AI crap is going to dial that problem up to 11.