A new math benchmark just dropped and leading AI models can solve 'less than 2%' of its problems... oh dear
Sometimes I forget there's a whole other world out there where AI models aren't just used for basic tasks such as simple ...