X-Events and Human Progress, or Why the Trend is Not Your Friend

John L. Casti

Čtvrteční seminář CTS

Anotace

At any random moment in time, the current trend whatever it may be, will almost surely continue to the next time period. This is a mathematical fact that underlies the fondness of fortune tellers, corporate planners, government policy-makers and futurists of all stripes to engage in trend-following as the method of choice to forecast the future. But trends always end, and the usual way they end is through a huge surprise, what I call an X-event. These surprises are rare, have high-impact and are essentially impossible to forecast in the way that term is used in the natural sciences, especially physics, engineering and astronomy.

In this talk, I consider how to anticipate such extreme, trend-shattering events. In particular, we look at how to characterize and measure risk for events that have never happened before, or at least for which we have no database of past occurrences to draw upon to talk about the "likelihood" or "probability" of such an event. So the usual tools of probability and statistics are powerless to help in such situations. What to do?

Here we argue that a judicious blend of complexity science and group psychology can help address the risk of unknown unknowns. The methodology is illustrated by many events in the political, social and economic domains.

A consequence of the tools outlined here is that the trend-changing X-events are as much an opportunity as a problem, and that human progress itself requires such shocks in order to free up space for innovation, invention and new trends to gain a foothold.

Přednáška se koná za laskavé podpory společnosti Biomedica ČS.