In this talk Andrew Lo addresses the problem of finding the right level of abstraction with which to think about economic phenomena. He compares economics to physics, with some surprising results.  For at least several decades economics theorists have assumed that the highest level of abstraction is the best. Lo argues that this assumption has stemmed from what he calls "physics envy": the enviable ability to explain a huge range of phenomena with a small number of rules or laws. Admittedly physics envy has led to the development of many useful ideas, including utility theory, game theory, and general equilibrium theory, among others. These have been real successes; however Lo takes up the argument recently made by some observers that the recent fiscal crisis is evidence that economics is by the nature of its subject matter not reducible to a small number of general laws. The possibility that a certain level of abstraction can be inappropriate for a given material raises the question of how to know what level is right where. Lo addresses this issue with a distinction originally made in the 1920's between risk and uncertainty which has been further developed into the Ellsberg Paradox. Both terms refer to unknowns, but in this formulation, "risk" refers to the category of unknown that is governed by defined probabilities. You do not know whether a flipped coin will land heads or tails, but you do know, with certainty, that the odds of either outcome is 50- 50. “Uncertainty" refers to unknowns in which the odds of a given outcome are not known and even cannot be known, such as the odds of running into an old friend you have not seen for years on the street tomorrow. Recently Lo and physicist Mark T. Mueller developed this polarity into a spectrum of types or kinds of unknowns. These run from well- defined cases like the coin toss, through types in which, while the odds of a given outcome are unknown now, it is pretty clear how to go about making them...