What I thought to be the key point:

"Social science research suggests that humans tend to glom onto the most alarmist perspective even if they are told how unlikely it is, he adds. We inflate the danger of things we don't control and exaggerate the risk of spectacular events while downplaying the likelihood of common ones. We are more afraid of terrorism than car accidents or street crime, even though the latter are far more common. Statistical outliers like the Sept. 11 terrorist attacks are viewed not as anomalies, but as harbingers of what's to come."

Mike