So every big AI math result will now have its “me too”s tagging along claiming they were “almost there” and a crowd of people insinuating that the work was stolen. If I was Anthropic or OpenAI, I would just give the whole math proof thing a rest and focus on things that demonstrably produce value. If you need to prove a math theorem or two to get there, then do it as needed, but not for the sake of just proving theorems. It’s too much drama for very little reward. Fields that are heavily prestige based instead of value based always descend into these meaningless debates about primality of work. After Riemann, there’s not much besides PvNP that would make as much noise anyway.