news.volyx.in

Epoch confirms GPT5.4 Pro solved a frontier math open problem (epoch.ai)

480 points by in-silico · 160 days ago · 706 comments on HN

Article summary

Epoch AI has confirmed that GPT-5.4 Pro has solved a frontier math open problem, a Ramsey-style problem on hypergraphs. The solution was verified by the problem's contributor, Will Brian, and will be written up for publication. Other models, including Opus 4.6 and Gemini 3.1 Pro, were also able to solve the problem using a general scaffold for testing models on FrontierMath: Open Problems. The problem involves constructing hypergraphs as large as possible that do not have a certain easy-to-check, difficult-to-find property.

Main themes

  • AI and math
  • novelty of AI solutions
  • human-AI collaboration
  • problem-solving strategies
  • mathematical progress
  • AI impact on society

What commenters say

  • Some commenters argue that AI-generated solutions are not truly novel, but rather extrapolations of existing knowledge.
  • Others believe that the ability of AI to generate new solutions, even if based on existing knowledge, is still a significant achievement.
  • There is disagreement about the significance of AI solving math problems, with some viewing it as a major breakthrough and others as overhyped.
  • The role of human involvement in AI-generated solutions is also a topic of debate, with some arguing that it is unclear how much human input is required.
  • Some commenters question the notion of 'novelty' in AI-generated solutions, suggesting that it is difficult to determine whether an AI has truly come up with something new.
  • The comparison between human and AI problem-solving abilities is also discussed, with some arguing that AI has an advantage in terms of speed and scalability, while others believe that human intuition and expertise are still essential.
  • The potential impact of AI on the field of mathematics is a topic of discussion, with some seeing it as a tool for accelerating progress and others as a potential disruptor of traditional methods.