Sparse Aperture-Mask Interferometry (SAM or NRM) behind Adaptive Optics (AO) has now come of age, with more than a dozen astronomy papers published from several 5-10m class telescopes around the world. I will describe the reasons behind its success in achieving relatively high contrasts ( 1000:1 at lambda/ D) and repeatable binary astronomy at the diffraction limit, even when used behind laser-guide star adaptive optics. Placed within the context of AO calibration, the information in an image can be split into pupil-plane phase, Fourier amplitude and closure-phase. It is the closure-phase observable, or its generalisation to Kernel phase, that is immune to pupil-plane phase errors at first and second-order and has been the reason for the technique's success. I will outline the limitations of the technique and the prospects for aperture-masking and related techniques in the future.