So finally they reach a part of PhDs level. Current version rely human to integrate results from they model and writting the papper. If LLMs/AIs can do all thing above, we can exactly get a PhDs level model.

My good sir or madam, disproving a decades-old conjecture produced by Erdos that has had armies of people in that field have their go at it IS a novel mathematical result.

They mean "new math" in the sense of more than a novel mathematical result, a new math paradigm or so.

The gap between novel result and "new math" is as wide as the pacific ocean.