Overconfidence and crashes in Knowledge Overconfidence

In New Scientist an article claiming we have no clue about tail-event probabilities for some catastrophic events. Reading the first few lines I would though it was Nassim Taleb article, but it was someone else. The article start out pointing how horrible things went in finance where people supposedly had very sophisticaded models. The article use this as a warning signal that we possibly also are under estimating and have little clue about the exact probabilities also for catastrophes.

I fully agree with the author we have no clue about exact probability calculations in catastrophes. Still finance and catastrophe theory people are possibly among the most knowledgeable in applied probability theory, and still most of them are horrible in abusing the wrong models and assumptions.

Other scientific fields are just even more horrifying when it comes to use abuse and misunderstandings when it comes to probabilities, statistics etc.. (more concrete and horrifying examples will be given over time).

The cover story in the same magazine is about how Darwin's Tree of Life theory probably was incorrect, or at least very incomplete. In every scientific and not so scientific field people from time to time and over and over again seems to get big surprises how bad their models actually work. Then the models are replaced with something else, something better(?) and again people get very confident that they know almost everything. Until again the real world proves the model wrong. Or at least partly wrong on a little assumption that make the model fail big time just when the model is most needed. In economy the consequences is billions and billions and billions lost, in catastrophe theory the possibility of the lots of dead and suffering people that potentially could be avoided etc.

Personally I think finance, physics, cosmology, medicine, biology... still at best are at very primitive stages of what is possible to actually understand. No I do not have the answers myself (well possibly some of them ;-). The more I study the more I understand I do not understand and also that others do not understand. And even if we should get close to the ultimate truth in some scientific disciplines there are good reasons to believe there sooner or later will be a massive crash in also knowledge. Even mountains die. If you look for the exact probability for such a knowledge crash you have missed the main point. Well hopefully such a knowledge crash will not happen for many more generations to come, well that bad models crash is good, hopefully new and better ones will replace them, and hopefully the generations to come will get a glimpse of the ultimate models before it all crashes again.