akshayub

Well 9/11 would have seemed a perfect fantasy on 9/10/2001.

But within matter of few hours the rate of american killed by terrorists went from same as those killed by furniture to few thousand times. The central argument is that a comparison with death by furniture is irrelevant, since the risk profiles are significantly different. Finally one cannot minimize risk of proliferation of nuclear weapons in isolation from agents such as rogue states and terrorists.

I would like to present a very good counter argument put forth by Taleb on his Facebook page:

""" (Heuristic: go to Pinker). He promotes there a WSJ article to the effect that "Terrorism kills far fewer people than falls from ladders"; the article was written by a war correspondant, Ted Koppel and is very similar to his Angels thesis.

Now let's try a bullshit-detecting probabilistic reasoning.

A- Falls from ladder are thin-tailed, and the estimate based on past observations should hold for the next year with an astonishing accuracy. They are subjected to strong bounds, etc. It is "impossible" to have, say, >1% of a country's population dying from falls from ladders the same year. The chances are less than 1 in several trillion trillion trillion years. Hence a journalistic statement about risk converges to the scientific statement.

B- Terrorism is fat tailed. Your estimation from past data has monstrous errors. A record of the people who died last few years has very very little predictive powers of how many will die the next year, and is biased downward. One biological event can decimate the population.

May be "reasonable" to claim that terrorism is overhyped, that our liberty is more valuable, etc. I believe so. But the comparison here is a fallacy and sloppy thinking is dangerous. (Worse, Koppel compares terrorism today to terrorism 100 years ago when a terrorist could inflict very limited harm.)

"""

Another such case is Nuclear Weapons, Millions of dollars are spent on controlling proliferation of nuclear weapons, even when no one died last year because of them. However this does not implies that we should abandon strict controls, since the risk characteristics are completely different.