AI Discovers Counterexamples in Statistics
teortaxesTex · x · 2026-07-15
- This post recounts a case of AI making a new discovery in statistics: by combining known concepts, a model generated a non-standard yet valid counterexample.
- The original text emphasizes that the output distribution of LLMs is broader than human thinking, making them better at generating such counterexamples. For statisticians, these are open problems that genuinely warrant manual construction.
- The quote also mentions that GPT-5.6 has already solved this conjecture. The author feels that while the AI did the job, they would have much preferred it to be solved by a human.
Related event: GPT-5.6 Reportedly Found a BH Counterexample(8 posts)→
More from Models
- DeepSeek cut KV cache per token from 389KB to 890B in two years as open-closed gap debate rages — teortaxesTex · 2026-10-27
- GPT-6 Astra controls dexterous hands zero-shot at just 5.2% success in 500 trials — stingning · 2026-10-27
- Unverified: fable 5.5 speculated at 4-5T parameters to outscale 1-3T Opus — haider1 · 2026-10-27
- Kimi K3 Is 'Open' Until You Succeed: License Caps Free Use at $20M Revenue — alex_verem · 2026-10-27
- Screenshot shows 'GPT-6 Astra Ultra' in ChatGPT model picker — unverified leak or joke — justalexoki · 2026-10-27
- DeepSeek Safety Eval Finds "Over-Attribution": Model Mistakes Ordinary Tasks for Tests — Much_Boysenberry9124 · 2026-10-27