|Graduate Seminar: The Philosophy and History of Automated Decision Making
|Soda Hall 510
Co-taught by Shamik Dasgupta (Philosophy) and Ben Recht (EECS)
This course will survey automated decision making, focusing on its historical development during the Cold War and its philosophical foundations. We will consider the parallel development of modern optimization, game theory, randomized clinical trials, and machine learning between 1944 and 1970 and how seminal data science was forged on the first computers. This history will let us ask what one needs to assume about the future in order to automate decisions based on past experience. Along the way, we will examine various philosophical issues that arise in deploying these decision-making systems, including idealization in scientific models and rational choice theory, the role of chance and probability in guiding decisions, and the nature of pattern recognition. Through this philosophical lens, we’ll probe why these particular decision-making frameworks have become so entrenched over the past 50 years.