🤖 AI Published 1h ago · 2 sources

Exponential Time Hypothesis Refuted by AI-Led Preprint

💡 Why it matters

If the proof holds, it would shrink the gap between P and NP for specific problem classes, potentially enabling efficient algorithms for real-world tasks like airline scheduling and protein folding.

OpenAI's math preprint claims that the Exponential Time Hypothesis — a bedrock of computational complexity — is false, based on new proofs derived by large language models. TechCrunch reports that the finding, if accepted, implies certain NP-hard problems may have.