The recent radio observations (Mooley et al, 2018) of a superluminal radio afterglow following GRB 170817A are interpreted in terms of a jet impacting a baryonic cloak, which is presumably the material caught at the front of the jet as the latter emerges from a denser ejected material. Assuming that we the observers are located {bf at a viewing angle of $sim 0.2$ radians from the emitting material (perhaps slightly more from jet axis)}, we suggest that the Lorentz factor of the jet is $lesssim 20$ at the time of the prompt emission, and that, as suggested previously, it is accelerated to much higher values before finally decelerating during the afterglow phase. A less extreme example of a short GRB being observed off axis may have been GRB 150101b (Fong, et al., 2016). A feature of GRBs viewed from large offset angles is a large afterglow isotropic equivalent energy as compared to prompt emission, as predicted (Eichler, 2017), and this is born out by the observations of these two GRB. It is also shown that the prompt emission of GRB 170817A, if seen way off-axis ($theta gg 1/Gamma $), could not be made by internal shocks in the baryonic material that powers the afterglow.