Super HN

New Show
   New secret math benchmark stumps AI models and PhDs alike (arstechnica.com)