According to CoinMeta, OpenAI indicates that their artificial intelligence has proven the error in the exponential time hypothesis (ETH). ETH is a decades-old hypothesis regarding the increasing difficulty of solving certain computational problems as they grow in size. OpenAI claims that they have developed an algorithm that exceeds the speed limits predicted by this hypothesis, challenging the assumptions used in hundreds of research papers. The company stated that the counter-evidence was discovered on October 16th, and the following day it was verified through a Lean proof-checking system, with the official verification being published on GitHub. The original proponents of ETH, Russell Impagliazzo and Ramamohan Paturi, will undergo review together with MIT's Ryan Williams. OpenAI indicates that translating these results into faster real-world algorithms "will require further significant research." The company also claims to have made two additional breakthroughs: • In mathematics: solving the Kakeya conjecture in each dimension, following the proof of the three-dimensional case last year. • In physics: producing a proof for solving a problem that is part of the million-dollar Millennium Prize. This work has not yet been published and will be reviewed by a team led by Nobel laureate David Gross.