The answer to your question is very sensitive to the time frame you are interested in. Assuming the costs and benefits you describe are referring to fitness (i.e., survival and reproduction), then the situation you are describing is one in which a particular behavior has lowered fitness (i.e., costs outweigh benefits).
There are several possible outcomes in this case.
Evolution. If the behavior is not plastic (i.e., it cannot change in a single individual), the there would be selection pressure against those individuals that exhibit the behavior. If the behavior is heritable then evolution would cause a reduction in the frequency of the behavior over time.
Lowered Fitness. If you are not interested in evolutionary time scales but the behavior is not plastic, then the outcome would be no change in behavior but a reduction in the survival and reproduction success of individuals with the behavior.
Change in Behavior. If the behavior is plastic (i.e., it can be changed by a single human) and the current situation provides the appropriate information about the state of the behavior, then the behavior will change. The key here (as with optimal foraging models) is that the appropriate cues have to be present to change the behavior. Often organisms use proxy cues to trigger behavioral shifts. For example certain caterpillars use day length to trigger shifts in developmental pathways (diapause vs pupation) when what really matters is temperature.
So the real answer to your question lies in the further questions of whether consumption is plastic (or hardwired as you say) and if it is, whether a situation provides the appropriate cues to indicate that we have acquired enough resources. Those are rather complex questions that, as far as I know have not been answered definitively in the general sense that your question implies.