Science definitely involves explanation as well as prediction, otherwise we’d still be iterating on more and more complex epicycles as predictive models for astrophysics. Induction isn’t everything.