Skin in the Game - Nassim Taleb

Amazon link

Reading SITG felt a little repetitive and not particularly coherent. Then again to do Taleb justice, he does mention that the book can be read in any order, with sub-chapters of 1-2 pages supposedly able to standalone, which positions more as a coffee table / toilet book, bite-sized knowledge at your own pace. Anywhomst, probably after reading his Incerto trilogy, the once unique and interesting tone Taleb employs (sardonic AF) didn’t make up for the lack of real content.


 (Intro)

  • Intellectualism: one can fix complex problems with top-down solutions

  • Scientism: over confidence in scientific results, not viewing it correctly as a process

  • Book is about (1) bullsh*t detection, (2) fairness/justice responsibility, (3) information sharing in transactions, (4) rationality in complex systems (2)

    • when one has skin in the game (SITG), these are related

  • Focuses on surprising implications, not explaining why these topics are important (5)

  • And it's about justice, honor, and sacrifice, things of existential importance to humans (6)

 (1)

  • Contact with earth, learning through experience is superior to via abstract reasoning (8), tinkering (Antifragile)

    • Pathemata mathemata: "learning through pain" (7)

  • Problem with policymaking: inability to understand complex problems, which do not have simple one-dimensional cause-and-effect mechanisms; meaning that intervention fails to envision second-order and interaction (9), i.e. lack of systems thinking

    • Crucially when sh*t hits the fan, policymakers are separated from consequences of their actions (10)… NT's definition of bureaucracy: "a construction by which a person is conveniently separated from the consequences of his actions"

  • Bob Rubin Trade: $120M from Citibank in decade pre-08, didn't write any check when it was bailed out later (13)

  • People learn by direct experience -> bureaucrats/leaders are increasingly better at explaining than understanding/doing; also systems learn by removing/mitigating what doesn't work (via negativa), (15)

  • Hammurabi's Law: if a builder builds a house that collapses and kills its inhabitant, the builder should be put to death (16)

    • Prevents builder's risk transfer + establishes risk symmetry, causes him to have skin in the game

    • Lex talonis, eye for an eye

    • NT doesn't want bankers to face the guillotine, but at least make Bob Rubin less attractive

  • Silver Rule is more robust than the Golden Rule (20), does not require deciding what’s "good" for others, indeed First Amendment = Silver Rule symmetry

    • Problem with Kant/universal behavior, is that we are local creatures that learn from direct experience, the general/abstract, i.e. universal, is not intuitive (21)

  • SITG is about living life with honor, some categorical do's and don'ts (33), soul in the game

  • Learning is rooted in repetition/convexity, rereading the same thing is better than reading two different things once (44)

  • "Books should be organized the way the reader reads, or wants to read, and according to how deep the author wants to go into a topic, not to make life easy for the critics to write reviews." (44)

    • Book reviews are judged according to how plausible and well written they are, never in how well they map to the book.

 Risk sharing / scale (2)

  • Regarding asymmetry, to what extent can people in a transaction have an informational differential? (55), what is ethical will always involve more transparency than what is legal

  • Not possible in practice to be both universalist and ethical (59)

  • Problem of scale: there's a group size sweet spot, larger than you, but narrower than humanity (61), for which risks can still be shared

  • Risk sharing, SITG, 'compericlity,' Rhodian Law "where merchandise is thrown overboard for the purpose of lightening a ship, what has been lost for the benefit of all must be made up by the contribution of all." (62)

  • We think conflicts of interests re: investment advice is more benign than impunity for bad advice, but in the absence of SITG journalists imitate other journalists -> monoculture + collective mirage (63)

 Minority rule (3)

  • Minority rule = asymmetry + minority: everyone can use disabled toilets, but disabled people can't use normal toilets -> more disabled toilets… kosher foods, automatic transmission, etc.

  • Moral values too, not via consensus, but the imposition of the most intolerant person (83)

  • It only takes a 3% minority for “Merry Christmas” to become “Happy Holidays” (88)

Employment (4)

  • Organizations deprive employees of certain freedoms by giving them SITG (96), a system that's evolved naturally because it's too costly to negotiate every contract/transaction

  • Company man has become the companies person, who fears upsetting potential employers (100), similar to Reid Hoffman’s idea of a gig vs. career 

Risk-taking (5)

  • Middle-class Americans resent high-paid professionals (no SITG) but not the rich (129)

  • Usually we consider static inequality, but dynamic inequality takes into account past/future, e.g. 10% of Americans will spend a year in the top 1% income bracket (131)

  • Perfect ergodicity: samples are representative of the whole; opposite: absorbing states, e.g. rich staying rich (132)

  • Better Angels diversion,

    • Pinker's response, "Taleb’s other allegations of statistical malfeasance are also products of dyslexia."

    • interpretation of Taleb, assume no trend, fit trendless model, fail to reject it… fallacious to "accept a hypothesis in response to nonrejection of that hypothesis with the test predicated on assuming the hypothesis is true"

      • Richardson: war generating process is Poisson, size is power-law

  • Revolving door, unethical, public servants should pledge to have any future private salaries capped

  • Lindy effect, Broadway shows that lasted for X days were likely to go on for another X days (141)

  • Grandma's advice is perhaps right 90% compared to anything from psychologists, it's been tested by time; things that hold in social science have to have antecedents in the classics (150)

6

  • Conditioned on having success/being qualified despite not looking the part, not looking the part suggests competency… between two equally competent consultants, choose the one that doesn't look like a consultant (154)

    • Also, given equal skills, hire the person with the least label-oriented education

  • Green lumber fallacy: what one needs to know in the real world isn't necessarily the same as what one can perceive through intellect; fellow make money on green lumber without knowing what the green actually means while another went bankrupt while knowing every detail about it (157)

  • Tendency to over-intellectualize: people believe they catch a ball based on solving equations regarding trajectory, instead they use a simple gaze heuristic (161)

  • Wealth and constructed preferences… Being rich should not limit your options; poison is drunk in golden cups: especially avoid advisors/consultants, people who are not incentivized to find simple solutions

8

  • Survival first, science last (217)

  • Rationality (and religion) survived because the populations that followed them survived (221)

  • Ensemble (independent) vs time (dependent) probability

    • In a strategy that entails (systematic) ruin, benefits are never worth it…  e.g. repeated Russian roulette