Show threadLê Nguyên HoangNov 28
I believe that the AI research failures exposed by the #ICLRLeaks are illustrations of broader societal concerns. While "innovation" is glorified, with their authors earning millions, regulation (here in the form of reviewing) is botched, automated and under-funded. This cannot be sustainable.
Lê Nguyên HoangNov 28
Yesterday, an #OpenReview vulnerability led to the leak of reviewer identities of all the major academic AI conferences, including the ongoing #ICLR2026 conferences. #ICLRLeaks This is both a huge disaster, and an opportunity to tackle the serious flaws of AI research. eu.36kr.com/en/p/3572028...

Academic Circle in Uproar: ICL...
Academic Circle in Uproar: ICLR Reviewers Reveal Identities, Low Scores Given by Friends

True Open Review: Unveiling Transparent and Authentic Evaluations