Book Review of “Fooled By Randomness” Part I of II
I’m writing this book review of Nassim Taleb’s “Fooled By Randomness” to an audience of people who’ve also read it (myself). So if what I say doesn’t make sense, go buy it and read it. I’m writing this so I have a very brief, very inaccurate synopsis for future reference. Although really, it’s not so much a synopsis of the book as much as a synopsis of my thoughts on random, sometimes disjunct parts of the book. To keep it somewhat cohesive, I’ll go chapter by chapter.
Part I: Solon’s Warning — Skewness, Asymmetry, Induction
“Solon, you were right!” Thus spake Croesus when he was on the pyre and presumably poor. What is Solon’s warning? Concisely, he said nothing’s certain but change, and to be truly happy you must have the assurances of the goddesses.
In Taleb’s first chapter, he illustrates the unintentional, unconscious, and subtly obvious tells people exude to broadcast their social pecking order. We hear of Nero Tulip, the cautious rich man, and John, the “High-Yield Trader.” Like Croesus, John experiences his eventual downfall, while Monsieur Tulip looks on, exuding subtly obvious tells that he is now higher than John in the social pecking order.
In Taleb’s second chapter, he illustrates the “Barrel of Reality.” Reality, he says, is a much more vicious game of Russian roulette because we do not understand the barrel in the game of reality. In Russian roulette, we know there is a one in six chance of losing the game; in reality, we have no clue. He terms this the “black swan problem” and promises to discuss it in chapter 7, which apparently is not a comprehensive chapter since the next book of his on my reading list is “The Black Swan” and much thicker. The “black swan problem” is encompassed by his argument for “A Bizarre Accounting Method” (the title of chapter 2) which is to account for what didn’t happen, not what did. In the Russian roulette example, if you did not lose the game after the first trigger pull, you would immediately account for what didn’t happen (the bullet did not exit that chamber). Now you don’t have a one in six chance of losing, but a one in five. And so forth. Taleb argues that we should judge actions by the cost of alternative histories; what would have been the cost of losing WWII? I’m very curious to see if he extrapolates this technique to the future (i.e., climate change). I speculate two things before reading on: 1) he does not extrapolate to the future and, 2) he says that since we cannot see the Roulette Barrel of Reality everything is random, and therefore we’re fooled by randomness. And here’s my favorite quote so far: “I thus view people distributed across two polar categories: On one extreme, those who never accept the notion of randomness; on the other, those who are tortured by it.” (p. 28)
In the third chapter, Taleb explains what a Monte Carlo simulation is. I’m a little vague on this idea halfway through the chapter, but perhaps upon rereading I’ll better understand. My poor comprehension of a Monte Carlo simulation says that it is a computer-generated simulation of all possible sample paths. The key phrase being “sample paths.” A sample is an observation, or a point in time. A sample path is how you get to that observation, all the possible paths to that observation, and all the possible results from that observation (if it’s hypothetical or abstract). I’m not sure this is accurate, but so far, in my mind, this is what a Monte Carlo simulation is. My new favorite quote from this chapter is from page 58 where Taleb is eviscerating journalists. “To be competent, a journalist should view matters like a historian,” he says, “and play down the value of the information he is providing… Not only is it difficult for the journalist to think more like a historian… the historian … is becoming more like the journalist.” And another great quote: “The problem with information is not that it is diverting but that it is toxic.” (p. 60)
Chapter four is Taleb describing aesthetics generated by Monte Carlo simulations. He argues that poetry and the like (things that are emotionally moving) can be generated by randomness (but that should not diminish their esteem in our eyes, as long as they’re evocative). He further argues that science, or reason, cannot be generated randomly. He suggests Alan Sokal’s Fashionable Nonsense to see how scientific buzzwords are laughably appropriated by literary intellectuals. He also proposes using the Reverse Turing Test. The Turing Test says that “a computer can be said to be intelligent if it can (on average) fool a human into mistaking it for another human.” Taleb argues that the converse should also be true. We can call a human unintelligent if we can replicate her speech by a computer and fool a human into mistaking it for another human.
Chapter five is all about Carlos the Idiot. Carlos looks the part of a trader, but he does poorly (very well at first, but then he blows up). Taleb defines blowing up as the state of traders who never expected the unexpected to happen to them. Taleb recently published an article here on Medium about looking the part; it’s worth a read. Anyways, I learned from Carlos the Idiot that I should always expect that the unexpected will happen to me (I’m guessing this is a lead-in for Taleb’s Antifragility hypothesis). He introduces Stephen Jay Gould’s idea of “negative mutations,” or “genetic noise,” which are mutations that survive despite being worse than their predecessors. Of course, these negative mutations do not last long in nature, due to temporal aggregation. Taleb does not explain temporal aggregation and a quick google search revealed nothing but financial-based results. I assume, based on the name, that it means that mutations, either “positive” or “negative” aggregate over time, and negative mutations will aggregate to the point where they’ll kill the organism or system in which they reside. Taleb’s main point here is that society, especially the markets, can aggregate negative mutations with little to no consequences or death to the system in which they reside. In his own words: “Owing to the abrupt rare events, we do not live in a world where things ‘converge’ continuously toward betterment.” I assume he’s speaking of human society here, and not nature, although nature and evolution are the metaphors he keeps using. He does clarify, with regard to temporal aggregation, that “time aggregation eliminates much of the effects of randomness; things (I read noise) balance out over the long run, as people say.” This seems to link in to what he said in the previous chapter about daily news being mere noise; we (he) can safely ignore it due to the law of temporal aggregation.
Chapter six begins with a discussion of how the median is not the message, per Stephen Jay Gould’s experience with cancer. I was listening to one of Taleb’s speeches on a podcast (perhaps at Berkeley, perhaps at Harvard, I don’t remember) and he talked about how one outlier can weight the mean, throwing off the usefulness of descriptive statistics. The example he uses in this chapter is a jibe at Robert Nozik, a philosopher, who said in his book The Nature of Rationality, “Since not more than 50 percent of the individuals can be wealthier than the average.” Of course, this is nonsense, if only you and I have the brains to call him out, and Taleb does. For example, if you “take a population of 10 people, 9 having a net worth of $30,000 and 1 having a net worth of $1,000” the average is $27,100 and 90 percent of the people are above average.
I am now about halfway through the book so I’m going to stop here.