The top economics journal in the world nearly published an article that, to all appearances, is 100% bullshit. It claimed that AI accelerates scientific discovery. An MIT PhD student appears to have used generative AI to fabricate the whole thing, data and all.
Comments
Absolutely wild fake university facade they got hey
It's sick how they do this wilfully all to benefit the already wealthy
From the MIT press release:
"This student is no longer at MIT"
every time you conflate "AI" with "Generative AI", Sam Altman gets a little bit richer
please shut the fuck up
This could've been another "Transgressing the Boundaries"
https://en.m.wikipedia.org/wiki/Sokal_affair
It needs work. But when it gets used properly, it works well.
That’s not bullshit.
Anyone who thinks about where the data they use comes from should know better. Just think about it for five minutes.
What.
“AI wrote this headline, didn’t it”
in any case seems like a really profound and thorough bit of scientific fraud, all the way down to faking a Corning website to source it from!
https://econospeak.blogspot.com/2025/05/artificial-intelligence-creates-more.html
It demeans and cripples the human imperatives of critical thinking and creativity.
Generative A.I. needs to be cast into an open volcano for the sake of our species's future.
https://www.reuters.com/legal/new-york-lawyers-sanctioned-using-fake-chatgpt-cases-legal-brief-2023-06-22/
as well, if this guy caught the issues why wouldn’t peer reviewers? he’s smart and they’re all dumb?
Peer review likely would have caught this if it the editor had recruited at least one material scientist to review alongside the economist reviewers.
https://bsky.app/profile/kevinjkircher.com/post/3lperiyasss24
It would be useless without that capability wouldn't it?
(I do review their code, make sure that if I’m coauthor someone has reviewed it if not me, & urge them to make code review groups for solo work)
People across all disciplines are nowhere near sceptical enough of this nonsense.
https://bsky.app/profile/katzish.bsky.social/post/3lpefgijoxk2t
This is hilarious after the week I have had debating the importance of checking the sources even if "AI said so." 🤣😂🤣😂
If artificial insanity becomes intelligent
It will commit suicide
No advisor, editor or peer reviewer caught this apparent MIT fraud. It was only caught because the preprint went viral.