Claude Independently Cracks 27-Year Exponential Time Hypothesis (ETH), Coincidentally on the Same Day as OpenAI's Discovery

Beating AI News Flash: OpenAI has just announced that AI has overturned the Exponential Time Hypothesis (ETH) proposed in 1999, and Anthropic has also publicly disclosed Claude's independent research findings. The two companies revealed that the discovery dates were surprisingly the same day: October 16.

ETH holds that some computational problems are extremely difficult, and no matter how algorithms are optimized, when encountering the hardest instances, computation time will still explode exponentially as the problem size grows. The two models separately found sub-exponential time algorithms that break through this limit, and can even precisely calculate how many satisfying answers a 3-SAT problem has, rather than merely determining whether a solution exists.

Anthropic disclosed that Claude independently found the core algorithm without any human mid-course input, and the entire research process output approximately 14 million tokens. Subsequently, Anthropic researchers collaborated with Claude to organize and refine the paper.

Both parties have now publicly released their papers and Lean formal proofs, awaiting further verification by the academic community.