10 minute recap: The Black Swan
My own summary of the key points of Taleb’s vindictive invective, which I will shorten to “TBS” for convenience.
Prologue
- TBS k’teristics: rarity (lack of prospective predictability), extreme impact, and retrospective predictability
- The inability to predict outliers implies the inability to predict the course of history
- We tend to learn the precise, not the general: e.g. French Maginot Line vs. Germany
- Platonicity is our tendency to mistake the map for the territory and focus on constructs instead of messier realities – you don’t know where the map will be wrong, and mistakes can lead to consequences.
Part One: How We Seek Validation
- Umberto Eco’s Antilibrary – contains books you have not read, as the value is higher – don’t treat your own knowledge as a treasure.
- Three facets of TBS: The error of confirmation, the narrative fallacy, and the problem of silent evidence.
- History and the Triplet of Opacity: The illusion of understanding (things coming out from the History black box), the retrospective distortion (assess matters after the fact), the overvaluation of factual info and learned pple (esp. when they Platonify/’make categories’)
- Speculator and Prostitute: Distinction between scalable and nonscalable jobs -> nonscalable is great for discovering TBS -> 2 types of randomness: Extremistan, Mediocristan
- The Turkey Problem -> the same hand feeds you for 1000 days, and then kills you
- The Skeptic/Erudite -> can be dissatisfied with your own knowledge – genuine curiosity
- Human Tendency to “Tunnel” -> focusing only on a few risks
- Our activities map (psycho- and bio-logically) into: System 1 – Heuristics and Biases (Shortcuts and Mistakes) and System 2 – logical, progressive deduction. We mistake S1 for S2.
- Characters: Fat Tony and Dr. John and their attitude towards the coin toss game – wonder about the texture of reality instead of sticking to your assumptions: thus the ludic fallacy -> the idea that you can learn about reality from a game, in a reality where the odds are unknown and you have to discover them.
- Recommendations: Denarrate: avoid newspapers, avoid system 1 for impt decisions, bear in mind how shallow we are with probability, above all, avoid tunneling
- Prediction, not narration, is the real test of our understanding of this world
Part Two: We Just Can’t Predict
- Three most impactful inventions: laser, computer, and internet – were all unpredicted.
- Our track record in forecasting – Relation between gains in knowledge and offsetting gains in confidence – The limits of prediction
- We are demonstrably arrogant about what we think we know -> too confident of our own knowledge; implications of this arrogance
- type 1: Arrogance in the presence of some competence (the “expert” problem), type 2: arrogance mixed with incompetence (the “empty suit”)
- The Defense mechanisms of prediction failure: the “different game” card, the “outlier” card, the “almost right” defense
- We do not attach an error rate to our predictions, and do not review predictions after the fact. The problem of humans having free will, irrational actions.
- We have Solutions that are looking for the problem – lasers, computers, internet invented first and then found a use. Viagra meant to be a hypertension drug. etc.
- We do not know what we will know (otherwise, we will have the means to obtain it).
- Poincare’s chaos theory -> limits that nonlinearities put on forecasting
- An epistemocrat is someone who has epistemic humility – the guts to say “I don’t know.” -> a regime based on this is an epistemocracy.
- Another problem is that we also don’t know much of the past either.
- Recommendations: We have to deal with this human problem as humans – accept that there will be bias, and don’t always try to withhold judgment, just be a fool in the right places. But beware dependence on large-scale harmful predictions. Maximize the serendipity around you. Use the “barbell” strategy – cap downside, maximize upside. Specifically, 1) make a distinction btw positive and negative contingencies, 2) dont look for the precise and the local (invest in preparedness, not prediction), 3) seize anything that looks like opportunity, 4) beware precise plans by governments, 5) don’t waste time trying to fight forecasters.
- Pascal’s wager: I don’t know whether God exists, but I have plenty to lose if he does exist and nothing to gain if he doesn’t. Max upside?
- Summary of section: 3 reasons why we cant predict: epistemic arrogance and future blindness; platonic categories; flawed tools of inference.
Part Three: Gray Swans of Extremistan
- 1. the world is moving deeper into exremistan
- 2. the Gaussian bell curve as a contagious and severe delusion
- 3. the concept of mandelbrotian randomness
- 4. the ideas of philosophers who focus on “phony uncertainty”
- 1. the world is moving deeper into exremistan
- the Matthew Effect: “take from poor – give to rich” otherwise known as cumulative advantage. Zipf’s Law – the more you use a word, the less effort to use it again.
- Memes are not like genes – the best dont necessarily survive: the carrying agents are self-interested in distorting their propagation process.
- Nobody is safe in extremistan, but nobody is threatened with extinction either.
- Concentration of networks around central nodes -> fewer but more severe crisis. Perhaps more failures needed to make system more resilient.
- 2. the Gaussian bell curve as a contagious and severe delusion
- Gaussian variations face a headwind that make probabilities drop faster further from the mean. Law of large numbers make groups of it safe.
- Measures of uncertainty ignore discontinuities. We should take exceptions as starting point and work towards ordinary.
- Quetelet’s average monster - bell curve’s misues was not Gauss’ doing.
- What humanist does not want to minimize the discrepancy between humans?
- Two key assumptions in Gaussian approximations: independent trials (clashes with idea of cumulative advantage) and no wild jump (clashes with TBS)
- 3. the concept of mandelbrotian randomness
- Fractals, recursiveness
- 4. the ideas of philosophers who focus on “phony uncertainty”
Part Four: End (How to Get Even with TBS)
Didn't See that one coming, eh?
.
