Though the distance scale to Gamma Ray Bursts is now known, the energy is still subject to several orders of magnitude uncertainty due to the possible beaming of the emission. Since the flow is relativistic, with Lorentz factor $\Gamma \gg 1$, the emission is collimated to a narrow cone of half opening angle $1/\Gamma$ around the direction of motion. Observer is therefore unable to tell whether the emission is spherical, at which case the energies of these events is as high as 1054erg, or narrowly collimated with proportionally less energy. During the afterglow, the Lorentz factor is decreasing with time. Consequently, the collimation should decrease, and the true energy of the explosion is revealed. We discuss the afterglow theory of such collimated events. We present observational evidence that Gamma Ray Bursts are indeed collimated: the explosions emit jets. Interestingly, the jets opening angle is inferred to be very small (a few degrees) for apparently bright events. Once corrected for the beaming angle, the energy of a large sample of events is clustered around 1051erg, comparable to the mechanical energy of a supernova. Gamma Ray Bursts seem to be standard candles, with apparently broad luminosity distribution only due to variation in their collimation. We discuss how future polarization measurements, can farther strengthen this understanding.