OpenAI's Experimental AI Achieves Gold at Math Olympiad

OpenAI has created a new AI model that is very good at solving math problems, even complex ones like those found in the International Mathematical Olympiad (IMO).

The model was tested on IMO problems, similar to how human contestants would be tested, and performed at a gold medal level, solving most of the problems.

This is a big deal because these math problems need creative problem-solving skills.

Researchers are surprised that LLMs can do so well.

It's not clear exactly how the model was trained or how much human help it needed.

OpenAI said they won't be releasing this technology soon.