Ten advances in mathematics and theoretical computer science
Ten advances in mathematics and theoretical computer science A few days ago it was Anthropic discovering cryptographic weaknesses with Claude using Mythos Preview, spending $100,000 on tokens and with prompts that included "again we are not looking for low hanging fruit, we want proper research to find genuinly hard findings." Now it's OpenAI's turn to flex. They set "an internal version of…
Ten groundbreaking advancements in mathematics and theoretical computer science have been achieved by OpenAI, according to recent wire material. The tech giant spent less than $2,000 on each problem at GPT-5.6 Sol token prices, far less than Anthropic's $100,000 investment. OpenAI's internal version of Astra, their next major model, discovered solutions to ten long-standing mathematical problems that had gone unsolved for at least a decade.
The results have been formalized using Lean 4 and documented in a paper. OpenAI also released an additional LLM-generated PDF detailing the proof reconstruction based on unpublished reasoning traces, showcasing a commendable level of transparency. Mathematicians have been impressed by these results, reflecting on the potential for AI to revolutionize the field and usher in a new era of "big mathematics" through large-scale, decentralized collaborations.
Written by urgent.news from Simon Willison's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
