In this note, I propose an explanation for the seeming contradiction between bioinformatics-based predictions of an essential function for ultraconserved DNA sequences, and the lack of an experimental demonstration of such function.