I is argued here that (at least light) nuclei may reside in a sweet spot: bound weakly enough to be insensitive to the details of the interaction, but dense enough to be insensitive to the exact values of the large two-body scattering lengths as well. In this scenario, a systematic expansion of nuclear observables around the unitarity limit converges. In particular, in this scheme the nuclear force is constructed such that the gross features of states in the nuclear chart are determined by a very simple leading-order interaction, whereas-much like the fine structure of atomic spectra-observables are moved to their physical values by small perturbative corrections. Explicit evidence in favor of this conjecture is shown for the binding energies of three and four nucleons. arXiv:1812.05327v1 [nucl-th]