Yesterday, a striking new paper arXiv:0903.4630v1 emerged on the arxiv by the well-known Renata Kallosh. Kallosh builds on the unitary cut and pure spinor methods which have established 3-loop superfiniteness and suggested the onset of divergences starting from the 9-loop order or even all loop ﬁniteness, respectively. She specifically investigates N=8 supergravity on the light cone and makes her prediction for the actual computations using Feynman light-cone supergraphs. It is noted that light-cone superﬁeld amplitudes have a non-local structure in the transverse directions leading to the increase of the delay of divergences with the increasing number of legs in the loop-amplitudes. The delay of divergences leads to the all loop ﬁniteness prediction for N =8 supergravity.