Take a photo of a barcode or cover
thefriedone 's review for:
The Black Swan: The Impact of the Highly Improbable
by Nassim Nicholas Taleb
Solid coverage of an interesting subject, but it's a slog at times. As I've come to expect with Taleb, you must withstand a barrage of anecdotes and character assassinations to come away with a relative few key points (all anointed with ham-fisted "novel" names, such as Extremistan vs Mediocristan). I'm glad I ended up with the later expanded edition because some of the more interesting arguments were added in a long (100+ pages?) essay addendum.
He writes with extreme arrogance and delights in trash-talking as many people as possible. I find that somewhat annoying but he's a colorful personality so at least it gives him a unique voice.
The general points/techniques I took away were:
- Identification of which things are/aren't scalable (body weight and height live in Mediocristan because they are confined to a relatively narrow range of values, while net worth or books sold live in Extremistan because they scale almost infinitely)
- Importance of how unknown unknowns cause Black Swan events, both negative (stock market collapse) and positive (runaway bestselling novel)
- Error in assuming most randomness can be modeled by Gaussian distributions (the concepts of standard deviation and variance, for example, are used in way too many situations where they make no sense)
- Futility of modeling the future only on past history (excludes possibility of Black Swans, leads us to believe that something that has never happened can't happen in the future, or is so unlikely as to be safely excluded)
- Ludic fallacy: our over-simplification of most randomness/uncertainty to reference points in games (dice, etc)
All told, some good new ideas. Worth the work but not a purely enjoyable read.
He writes with extreme arrogance and delights in trash-talking as many people as possible. I find that somewhat annoying but he's a colorful personality so at least it gives him a unique voice.
The general points/techniques I took away were:
- Identification of which things are/aren't scalable (body weight and height live in Mediocristan because they are confined to a relatively narrow range of values, while net worth or books sold live in Extremistan because they scale almost infinitely)
- Importance of how unknown unknowns cause Black Swan events, both negative (stock market collapse) and positive (runaway bestselling novel)
- Error in assuming most randomness can be modeled by Gaussian distributions (the concepts of standard deviation and variance, for example, are used in way too many situations where they make no sense)
- Futility of modeling the future only on past history (excludes possibility of Black Swans, leads us to believe that something that has never happened can't happen in the future, or is so unlikely as to be safely excluded)
- Ludic fallacy: our over-simplification of most randomness/uncertainty to reference points in games (dice, etc)
All told, some good new ideas. Worth the work but not a purely enjoyable read.