They are both right! Here's why: Taleb describes two imaginary countries. The first is "Mediocristan" where predictive math models dominate and people and even

They are both right! Here's why: Taleb describes two imaginary countries. The first is "Mediocristan" where predictive math models dominate and people and events are nicely characterized by Gaussian statistics. The second is "Extremistan" where events are difficult, if not impossible, to predict. Here's the bad news: we think we're living in Mediocristan but we're really living in Extremistan. Taleb says the models of Mediocristan do not predict "Black Swans" - the "unknown unknowns". Their surprising occurrence isn't captured by Mediocristan's predictive math models. So in Mediocristan the Black Swans are characterized as outliers. Should a Black Swan occur, Mediocristan's math modelers sweep it under the rug by either forgetting about it or by extending the math models to include it (for "next time"). Of course next time a _different_ Black Swan alights. Note that there are good and bad Black Swans depending on your viewpoint. Bueno predicts something more specific - the outcomes (e.g., wars) of political conflict. Bueno has a mathematical model of such conflict based on game theory. So in Taleb's parlance, Bueno's model is in Mediocristan - the land where mathematical models prevail and no Black Swans exist. Bueno's model sometimes fails. And sometimes events occur whose parameters fall outside the domain of Bueno's model or any other model. Those would be more Black Swans.