The function of LLM-based math “proofs”

There has been quite the uptick in “AI” labs’ claims of their LLMs “solving complex math problems”. I wondered why that is, especially given that LLMs are not exactly built for the kind of rigor that proving mathematical theorems requires. There’s a whole bunch of – really hard – math problems out there looking for solutions. For people who are not into that kind of thing: The issue is not that there is completely uncharted territory but that usually there is a hypothesis about some mathematical structure that so far could not be proven or disproven. So the hypothesized theorem can’t be used in other proofs. And there are some really load bearing hypotheses like the “hoping that P and NP are not equal” which a lot of our current cryptographic algorithms rely on. So why are the “AI” labs going so hard toward mathematical proofs? It’s not like that’s a use case actual people or businesses have in their everyday life so they are burning a lot of money doing something that many people…