Have you read The Black Swan? It's not a pep talk. No, but I've read the wikipedia article, and absolutely despised the tone of this article. That said, callin

Have you read The Black Swan? It's not a pep talk. No, but I've read the wikipedia article, and absolutely despised the tone of this article. That said, calling it "pep talk bullshit" was an emotional response - it's been a long day. What I mean to say is that Taleb is giving bad advice with a lot of confidence and bad "proofs" based on idiotically simple models. For example: Taleb says is to do things that will reward you if the unpredictable happens That sentence seems to either (a) have a huge gap of basic probability, or (b) contradict itself. First, if "unpredictable" is to mean "low probability of happening", there are several orders of magnitude more "unpredictable" things than "predictable" ones (and that's why they have such a low probability of happening). A better model of "events" and "probability" would consider the key variable correlating to probability of events is specificity - it's highly probable that some planes will crash in the next 10 years. That relatively general event can be partitioned into less probable (but more specific) events. There are more specific events than general events, and each is less probable in proportion to its specificity. If you could know which unpredictable thing were going to happen, they wouldn't be unpredictable anymore. Of course, reality is stranger yet - events aren't all neat subsets and partitions of each other like they are in that model, and predictable varies with probable based on some measure of the accuracy of the world model of the predictor (which, I would argue, is usually pretty bad). Hence, I would argue that predictability doesn't matter much at all (after some point) - evolution and the ability to let go of one seemingly great ideas matters more.