The claim that OpenAI and Anthropic achieved the same partial case of the Hodge conjecture

A user claims OpenAI and Anthropic both achieved the same partial case of the Hodge conjecture. He calls it another instance of his “difficulty convergence” thesis, arguing that this is how much of the problem the current generation of models has unlocked.