Predicting the Future is Risky: A Panel Discussion
Effective policies, wise investments, informed decisions, and smart designs all require some consideration of possible future conditions. This may take the form of a statistical forecast, model simulation, projection of trends, or simply an educated guess. However, the future is inherently uncertain, so perfectly accurate predictions are impossible. How can we characterize and plan for what we do not know? Does a detailed and honest characterization of forecast uncertainty lead to better decisions? Is it possible to choose a course of action with confidence, while still planning to learn and adapt as the future is revealed? Our multidisciplinary panel is planning to address these questions (but who can predict what will really happen?).
James Berger, Duke Statistical Science
Nita Farahany, Duke Science & Society
William Poundstone, Author, The Doomsday Calculation
Jonathan Wiener, Duke Law School
Mark Borsuk, Pratt School of Engineering
Lunch will be served at noon. Please register at http://duke.is/IZvffa to reserve your lunch.