In a surprising turn of events surrounding the International Mathematical Olympiad (IMO), OpenAI announced its results prematurely, causing controversy. Initially scheduled for July 28, Google DeepMind has advanced its announcement timeline to today. OpenAI’s research scientist, Noam Brown, stated on X that he informed an IMO organizer about the announcement, who requested it be held until after the closing ceremony to respect participants. This claim was contradicted by an IMO coordinator who labeled OpenAI’s actions as “rude and inappropriate,” emphasizing that OpenAI did not collaborate with IMO on model testing.
Since its inception in 1959, the IMO has posed challenging mathematical problems requiring deep reasoning rather than mere computation. Contestants from over 100 countries tackle six proof-based problems in two rigorous 4.5-hour sessions. The unexpected nature of OpenAI’s results raises industry speculation, particularly as predictions assigned only an 18% chance for any AI to achieve gold by 2025.
Source link
