A recent long-term population study finds no evidence for "directional selection," but plenty of evidence for "fluctuating selection." These findings present a serious challenge to the traditional assumptions that variations in biological traits are random and that adaptation results from an external mechanism. Though several kinds of selection are invoked to explain observed population dynamics, evidence for non-random variation renders the concept of selection unnecessary. Instead, observed adaptation is better explained by distributed problem solving. This mechanism also provides a plausible explanation for the post-Flood rapid diversification of life. It is time for creationists to discard natural selection.