The research frontier on economic networks has been advancing along two strands: one emanating from economics and sociology, the other from research on complex systems in physics and computer science. The socio-economic perspective puts emphasis on understanding how the strategic behavior of the interacting agents is influenced by -- and, reciprocally, shapes -- relatively simple network architectures. By contrast, the alternative perspective investigates how the structure of large networks affects their overall performance, i.e. in the exchange of knowledge, in trade, or investments. The key challenge is to identify paths through which the two strands of research may converge. The talk will adress some of the problems that should be tackled in this endeavor.