I will describe two asymptotic results: one concerning random matrix theory and one concerning the NLS equation. From the bowels of those considerations will emerge a crazy method for analyzing the asymptotic behavior of an integral. Then some open directions for research (aimed at graduate students) will be discussed.