Ask HN: What if we funded math research like we fund AI labs?

The annual U.S. federal budget for basic math research is <$1B.

OpenAI alone just raised ~$122B, and frontier labs are increasingly concentrating enormous amounts of compute and human talent on open math problems.

We seem to be entering a strange situation where a handful of private labs have vastly more resources for attacking these problems than the academics traditionally responsible for doing so.

Have we had the capability all along, just not the hype? Please discuss :)