Math problems have long proven difficult for AI. Scientists have speculated that this may very well be because AI systems can’t recognize their very own limitations, while others have surmised that the difficulty is that AI are focused on language and never on numbers, resulting in just a few stumbles along the way in which.
The Epoch AI experiment has demonstrated that AI is becoming more proficient at among the trickier math issues. Within the test, GPT-5.2 Pro was presented with problems from various branches of math.
Joel Hass, a professor within the department of mathematics at University of California, Davis, contributed certainly one of the issues solved by GPT-5.2 Pro. He told Epoch AI he was impressed with the way in which it cracked his topological challenge. “GPT-5.2 Pro solved the issue with correct reasoning. Notably it was capable of recognize the precise geometry of a surface defined by a polynomial in the issue statement,” he said.