In this paper we present a dialogue between game theory and laboratory/fiel
d experiments using a number of examples from the economic systems design l
iterature. We observe that neither game theory or laboratory experimentatio
n is sufficient for economic design. Instead, what seems to be critical is
the ability of each to inform the other. We discuss the rapid emergence of
'smart' markets which are being engineered to extract the gains from exchan
ge in increasingly complex resource allocation environments. We describe so
me of the desirable properties of these markets.