In economic theory one typically discounts future benefits at a constant rate. An example of this is the celebrated model of endogeneous growth, originating with Ramsey (1928), which leads to the so-called golden rule in macroeconomics. There are now excellent reasons (intergenerational equity, for instance) to use non-constant discount rates. There is then a problem of time-inconsistency: a policy which is optimal today will no longer be so when the time comes to implement it. So optimization is pointless, and a substitute has to be found for optimal strategies. We will define such a substitute, namely equilibrium strategies, show how to characterize them, and investigate what happens to the golden rule. This is joint work with Ali Lazrak.