Having watched this and Numberphile’s explanation, I was pretty intrigued about how this may be applied to other problems. As I understand it, seems like the ability to restrict the search field to rational solutions could be extremely helpful for areas of research where continuous distributions are applied to necessarily discrete outcomes, especially in terms of saving compute resources and processing larger data sets.

Anybody have insight about how this performs computationally? Benchmarking “simple problems” like the one in the video? Numerical instability?

Disclaimer, not a math guy, so maybe this is a nonsense idea and I’ve misunderstood something along the way.