MIT paper on AI for materials research found to be fraudulent
8 comments
·May 17, 2025readthenotes1
I am shocked, "shocked*, to find fraud in the Science community!
The best thing from Science for the last 15 years or so has been the focus on discovering and publicizing it.
feverzsj
Just like most AI papers.
trollied
pera
This is actually an in-depth analysis on the fraudulent paper and it's much more interesting than the WSJ article. One for dang to decide I guess.
ChrisArchitect
And it's shared there adding to large discussion. The discussion is over there.
atleastoptimal
The ironic thing is I'm sure AI for materials research isn't too far from being avaliable or possible, they were just a bit too early.
whatshisface
There are a lot of legitimate papers, this is just one fraud.
> I also think that if comments were enabled on arxiv preprints, this could have led to a much more rapid conclusion to the fraud. Probably a materials scientist who read the paper realized this was fraudulent but wasn’t able to get that view quickly to the economists who were actually reading and discussing the paper. A well-written arxiv comment explaining why the data on materials similarity, for example, couldn’t be true, would have gone a long way.
could this work outside/on top of arxiv?