GPT-5.6 Used Similar Prompt to Close 30-Year Gap in Convex Optimization
AI-generated image · Bay Street Wire
OpenAI’s GPT-5.6 reportedly used a prompt akin to the CDC proof announcement, with the result verified in Lean.
According to a report from Hacker News, OpenAI’s GPT-5.6 used a prompt similar to the one used in the company’s CDC proof announcement to close a 30-year gap in convex optimization, verified in Lean.