Press Enter to search · ESC to close

AI × Crypto

OpenAI models disprove Exponential Time Hypothesis, three discoveries announced

OpenAI’s internal models made three new discoveries in mathematics, computer science and physics, the company announced on October 27. One model proved the Exponential Time Hypothesis — used as a computational hardness assumption in hundreds of papers — false, with the result formalized in Lean. A team led by ETH proposers Russell Impagliazzo and Ramamohan Paturi, alongside MIT professor Ryan Williams, will verify the finding.

Original source

AI take

A machine-generated result that overturns a foundational hardness assumption — and is then handed to the very researchers who built that assumption — marks a shift in how mathematical claims get credentialed, not just in who produces them. The Lean formalization matters as much as the proof: it moves verification from social trust toward machine-checkable evidence, which could reshape peer review in complexity theory. The open question is whether the ETH's role in hundreds of papers survives scrutiny, since a disproof would ripple through results that relied on it as a given.

Generated by AI for reference only.

Share

Related News

TREE NEWS share card
Long-press image above → Save to Photos / Share
Pitch us Feedback