Paying mathematicians $1,000 a proof to spot errors for RL training data
ctjlewis · x · 2026-09-10
An X user revealed that their team pays mathematicians $1,000 per task to review proofs and find issues, producing high-quality training data. The hired mathematician was visibly exhausted yet still identified errors "like a professional." The author argues this should become a "professional norm of mathematics," hinting that frontier labs increasingly outsource expert-level review at high rates to support RL and proof-model training.
Related event: Mathematicians Paid $1,000 to Find Errors in AI-Generated Proofs(3 posts)→
More from Research
- Researcher flags unresolved data transparency issues in compositionality evals — najoungkim · 2026-10-27
- Separate Forms from Estimators: Why Steer by Difference of Means When You Can Use SGD? — johnhewtt · 2026-10-27
- EMNLP paper probes why some LLM biases survive rewording while others don't — radamihalcea · 2026-10-27
- Datology webinar: Curation Studio matched top data baselines with under 20% compute — josh_wills · 2026-10-27
- KANLab: an interactive browser playground for Kolmogorov-Arnold Networks — Lukas Niehaus · 2026-10-27
- CMU's LLM2Cert synthesizes stability certificates verified via SMT and Lean — CarnegieMellonU · 2026-10-27