> they're probably going to be evolutionary improvements that build on what we have, rather than throwing everything away in favor of a brand-new approach.

Probably, yes -- but working on a problem starting from the beginning is something more people should do.

It is riskier, for sure. And I am glad their team is doing it.