OpenAI says its AI disproved the Exponential Time Hypothesis; experts will check

TL;DR

• On 26 October OpenAI said an internal model produced a disproof of the Exponential Time Hypothesis, a conjecture about how hard Boolean satisfiability is.
• A group led by Russell Impagliazzo, Ramamohan Paturi and Ryan Williams will check it in the open. Williams calls it a claimed refutation.
• OpenAI also claims a maths result and a physics result, each handed to outside experts.