GPT-5.6 Closes a 30-Year Gap in Convex Optimization

GPT-5.6 used a prompt to close a 30-year gap in convex optimization

I am sharing news that OpenAI's GPT-5.6 reportedly used a prompt to solve a decades-old problem in convex optimization. This breakthrough suggests that advanced AI models are now capable of tackling complex mathematical challenges that have stumped researchers for thirty years. The announcement has sparked significant discussion within the math community about the evolving role of large language models in scientific discovery.

"GPT-5.6 used a prompt to close a 30-year gap in convex optimization."