Physicists have been working for banks and hedge funds on applied problems in finance for more than two decades, and recently have doing academic research as well. This talk will survey academic research by physicists and contrast it with mainstream economics. I will argue that the difference comes not from the application of alternative techniques or new mathematics, but rather from fundamental differences in what questions are considered interesting and how one should go about solving them. This will be illustrated with a simple model for how systemic risks and extreme price movements are generated by the use of leverage (buying with credit). The current financial crisis illustrates that the economy is indeed a complex system, and that new approaches are needed that properly take this into account.