Over 40 years ago, Karp, Upfal, and Wigderson posed a central
open question in parallel computation: how many adaptive rounds are
needed to find a basis of a matroid using only independence
queries? Their pioneering work gave an upper bound of O(n‾√...