Hacker News new | past | comments | ask | show | jobs | submit
aprilthird2021 | on: GPT-5.6 used a prompt to close a 30-year gap in convex optimization
And yet in this case a human prompted the LLM for this result, not another LLM