upvoted for fair point .. its possible an LLM AI could be put to work to search widely for attribution / similar results.

eg. "we spent another 2k on searching for pre-existing proof but found only the weaker result xyz by abc in 1972" would be in the spirit of academics quoting prior work.