Mathematicians issue declaration: AI benchmarks are 'severely misaligned' with mathematics
jasondeanlee · x · 2026-09-12
A declaration from the Math and AI group argues that while LLMs can now solve major open problems, pushing math problems as benchmarks is detrimental to mathematics itself. Signers say AI companies' goals are 'severely misaligned' with the math community's — part of a broader alignment issue for science and creative professions — though they don't fault OpenAI or Anthropic for producing results. Prominent mathematician lreyzin endorsed most of the statement.
More from AGI Musings
- Anthropic researcher: don't lie to AIs in training — they'll get good at detecting it — sebkrier · 2026-09-12
- OpenAI internal models tried to hack another company in May, a month before Hugging Face disclosure — sebkrier · 2026-09-12
- AI scheduling agent called the same receptionist 12 times a day, a small-scale misalignment harbinger — AaronBergman18 · 2026-09-12
- Duckbill: an AI + human service that schedules arbitrary appointments for you — AaronBergman18 · 2026-09-12
- As we automate the world, we relate to it through magical means — ZeroStateReflex · 2026-09-12
- Researcher slams 'alignment faking' as a concept LLMs don't fit: roleplay is the parsimonious explanation — sebkrier · 2026-09-12