Critical Insights on Nassim Nicholas Taleb's 'Antifragile

Be ultra conservative with 90% of your risk taking, then take high risks with high payoffs with the 10% you can afford to lose.

While I agree with a number of things Nassim Nicholas Taleb says they're so buried in a mountain of pretentious convoluted crap upon which a fun house of mirrors, located in the center of a cherry picking orchard, tended by straw men who are fighting and contradicting one another with an excess of real and made up pompous words in a desperate attempt to prove they're all the smartest that it's all but completely unusable and designed to fulfill his insecure need for self validation.

The above MASSIVE run on sentence is my impersonation of this author and his writing style:-)-

To be fair I want to list what I agree with.

1 ) The world is complex and there's much we can't know or don't yet understand.

2 ) ALL ideas must be subjected to falsifiable empirical testing.

3 ) Absence of evidence isn't evidence of absence. Here's the thing, it's also useless and unactionable. You can't make ANY use of anything lacking evidence and unprovable existence.

I have no evidence that the FSM ( Flying Spaghetti Monster ) exists, but my lack of evidence isn't evidence of its absence. You can't can't prove the FSM doesn't exist:-)-

It's completely and logically valid, but also useless and unactionable for changing ANYTHING.

In other words, with a few barely or unusable ( and interesting ideas ), Antifragile is just as useless as the Black Swan was. The rest is all the academic stuffing he criticizes everyone else of using while himself doing so to a greater degree than ANYONE else I have EVER READ.

I can't comment on subjects that I don't know, so I'll leave economics, math, finance, languages and history he appears to know more about than I. I know virtually nothing about those topics. So I don't know enough to know if he actually knows WTF he's talking about, or not.

I will openly criticize the many logical, evolutionary, fitness, metallurgy, science and other things I do know that he's either clueless about, or knowingly twists to prove himself right rather than arrive at any REAL scientific and logical truth.

Let's start with weights. A 300 ish pound dead lift he's so proud of is fuck all, unless you're a 150 pound woman doing it. If he's going to brag about how impressive he is, he should at least know what One REPetition MAX ( 1RM ) HIT ( High Intensity Training ) is, and that 300 pounds would only be impressive if he weighs between 75 to 150 pounds. This was his first fail, proving how unimpressive he is and how ignorant of the topic he was using to prove his point about acute stress in strength training as an illustration of Anti-Fragility.

Antifragility, is one of the dozens of made up words he uses when real, common words would do. He's so desperate to prove his vocabulary is superior to the reader, all other academics and Nobel prize winners , that he needs to add long ridiculous sounding words that don't exist to compensate for...something.

For some reason in his search for a word that means the opposite of fragile, he didn't find ADAPTIVE, which fulfilled every definition in his book for what he calls "antifragile". I'm completely open to ANYONE who can prove me wrong on this point, including and especially the author.

Back to strength training. While "proving" acute infrequent stressors cause what anyone who knows WTF they're talking about would call adaption or "supercompensation" he misses the point that maximum strength is gained by doing your 1RM on a CONSISTENT, PREDICTABLE, schedule that follows the peak of your super compensation curve. In other words, NOT random, NOR unpredictable. so by his own definition, super compensation ISN'T antifragile. But it seems he doesn't know that:-?

Strength adaptions would quickly disappear or fall off unless the stimulation was PREDICTABLE and regular, in an environment which was also predictable. For example, where gravity and other things remained relatively, or completely constant. If the stress was TRULY unpredictable or random, it would average out to ZERO strength gains, or injury from over training due to insufficient recovery.

Strength is adaptive, but maximally so ONLY under conditions that are predictable and optimally distributed over time. For me that's every eight days. Nassim, if you're going to prove your theory, at least know WTF you're talking about when you make analogies that DISCONFIRM your own point:-?

He makes SO many mistakes or willful misrepresentations of biological adaptions and other thing "proving" that nature is "antifragile" that his 16 hour 14 minute book would probably take weeks if I stopped to record every mistake he made:-?

He doesn't understand that caloric intake is and has always been the best predictor of weight gain or loss. He pontificates about the irrelevance of calories and the foolish naivety of applying thermodynamic models of energy to complex biological systems, but somehow didn't understand that an ACTUAL expert and leading international researcher Jennie Brand-Miller, author of The New Glucose Revolution said in that book " Caloric intake remains the best predictor of weight gain or loss." ( Paraphrased ) So even one of the leading GI ( Glycemic Index ) researchers in the world says calories matter.

My point here is that Taleb, throughout this book and the Black Swan basically says everyone but myself, and those that agree with him are idiot's, "fragelistas"...regardless of the fact that he's in no way qualified beyond an armchair / book academic ( like myself ) to express medical opinions. So while I'm no more qualified than him, we both read books, he's far less qualified than many people he dismisses and condescends to in THEIR area of expertise.

He doesn't understand evolutionary biology. Evolution is a process of elimination, NOT an antifragile adaption. While I sometimes anthropomorphize nature, he does it constantly and seems to not understand that nature doesn't like, hate, strive or aspire to ANYTHING any more than a rock "prefers a sunny location". Rocks aren't conscious, nor do they make moral judgments or decisions, they're AMORAL. That is, devoid of morality, not evil, kind, good nor bad and yet Taleb continuously anthropomorphizes nature to prove his idea of "antifragility".

Nature doesn't make decisions, it's more like a a drunk monkey with a machine gun, except that monkeys HAVE true intelligence, "Mother Nature" does NOT. The monkey fiddles with the gun, and IFF it finds the trigger, pulls it and is lucky enough to not accidentally kill itself, it may eventually live to use it effectively for its own purposes. More random fiddling, more loud scary sounds, and eventually it notices the effect it has on the environment. It gradually learns to aim it. First it sprays bullets everywhere. IFF by chance it shoots a bunch of bananas out of a tree, it MAY learn how to do it intentionally...

Nature uses an automatic approach. Evolution isn't adaptive as such. It throws out ALL possible variations of genetics, behaviors and organisms at random, and whatever survives, reproduces. Taleb talks about natures "reliance" on error to create antifragility through variations, that's just dumb. Transcription errors just happen in DNA, those mutations either survive to the next generation or don't. Those that do, are the starting point for the next generation, evolution isn't antifragile intelligent design, it's unlimited blind random attempts, and the ones best suited to their environment continue, PERIOD. Refinement happens because failures DIE, NOT because "mother nature" consciously designed an antifragile system. It APPEARS antifragile because only the best suited organisms survived. It's very similar to the mythologies about entrepreneurship. We only see the winners, because the rest work at McDonald's or elsewhere.

When he speaks about autophagy he omits ANY discussion about how it's possible that the self eating mechanism may "mistakenly" eat healthy cells and material, not just the damaged materials in the body. The little about the topic ( which I only recently learned ) suggested that autophagy processes selectively favors consumption of diseased cells, cellular waste and the like that could be converted to energy to keep healthy cells alive. That doesn't necessarily mean that process is without error and doing potential harm. Taleb conveniently avoids that contradiction between transcriptions errors, "creating" antifragility through genetic variability, and errors during autophagy that may cause harm aka create fragility by NOT addressing the possibility. If the errors necessary for genetic variability occur, isn't there an reasonable probability that there are autophagy errors that cause harm? Almost certainly, but in the Talebverse errors only occur when they support, not refute his hypothesis.

It's especially offensive since he points out that disconfirmation is the most powerful means to arrive at the truth of something. But he avoids any and all discussion of events or scenarios that would disconfirm antifragility. I can't think of a SINGLE case in which he does anything but swim in the confirmation bias he ALSO condemns. In a word, he's a hypocrite because throughout this book he uses everything he claims invalid, philosophy, logic and science especially that which he condemns in others yet uses to support his own points. But he goes much further engaging in every single specific practice used by "charlatans" in a desperate attempt to validate all that he thinks while using the same to discredit everyone else by name. He's the only person who calls Daniel Kahneman "Danny" when mocking, criticizing or disrespecting him, and "Daniel" when praising him in this book.

Evolutionary biology creates astoundingly effective designs by trying ALL designs. The best ones survive, reproduce and therefore refine themselves automatically...or die. Like iterative computer design software. It's NOT antifragile, it's DIVERSE and so the best suited to survive an unpredictable event do, everything else DIES. So if by antifragile he means anything that HAPPENS to survive is antifragile, then yes, in that specific environment and circumstance it's antifragile, but otherwise no. Taleb has defined antifragile in a way that's unfalsifiable. Luckily for him, he claims all sciences, save ( some ) physics and a few rare exceptions are nonsense, same for philosophy, except when he invokes it. In the same breath he engages in everything he criticizes by using the science and philosophy he claims are completely invalid:-?

The underlying theme of This book and Black Swan is some version of "nobody can know anything", exept me ( Taleb ), and anyone who agrees with me.

Other things he doesn't understand, but uses to illustrate why he's right are annealing of metal, and fractal geometry / chaos / non-linear dimensional geometry.

Let's start with annealing as proof of antifragility, and just by chance he ALMOST get's it ( sort of ) right. Metal can be made brittle ( fragile to sudden impact ), and annealing will make the metal softer and more plastic. Normally in industry it's simply called "tougher" and tempering, the opposite process to annealing makes metal "harder."

His own definition is laughably one dimensional, either fragile, OR antifragile which is why metallurgy and annealing are the perfect disconfirmation. When I worked at National Oilwell I often saw the welders heating the ends of pipes, then allowing them to cool slowly ( annealing ). It made the ends softer so they could cut threads into them while leaving the rest of the pipe hard to remain more wear resistant.

If annealing produces antifragility, why didn't they anneal the whole pipe? Simple, to cut the threads more easily annealing made the ends of the pipe fragile to thread cutting, but antifragile to shattering or breaking when the threads where cut. Wait, do you mean annealing is simultaneously fragile AND antifragile to different stressors, DAMN RIGHT I do! It's just that Taleb doesn't know enough about metallurgy to know that. and his theory is brittle / fragile to reality, metallurgy and the universe in general which aren't as one dimensional or fragile as his theory.

So annealing the ends made cutting the threads possible and leaving the rest of the pipe tempered ( through heating then suddenly cooling ) making it harder, and more resistant to wear, yet more brittle and susceptible to shattering or fracturing. So the most antifragile state for those pipes in that application was to have soft annealed ends, and a tempered center. I don't recall if they later tempered the ends to make them more wear resistant AFTER threading them or not. The point is, tempering OR annealing could increase their antifragility or robustness DEPENDING ON how they were used and for what application.

BTW, did you notice that by his definition anything that wasn't made stronger or better by random stressors, "up to a point" was fragile? So wait, aren't all metal pipes fragile by his definition of antifragile? Yes, yes they are, and there he was "proving" his model of antifragility on a material that is fragile by that same definition:-)- You'll have to ask HIM HTF that contradiction supports his antifragility hypothesis.

Speaking on one dimensional thinking, let's move onto fractal or ( fractional dimensional thinking ). I know just enough about the topic to know that Taleb doesn't know WTF he's talking about yet again, or has omitted any details which might undermine his model.

To be fair, I don't recall any specific mention of how antifragility related to fractal geometry directly. But he repeatedly invoked it, I think to prove he's smart and knowledgeable about a sexy fashionable branch of math. Sadly, I who failed second year calculus TWICE ( DO YOUR HOMEWORK! ) will easily mop the floor with a guy who's supposedly an expert in advance mathematics:-)- This isn't bragging, but illustrating how VERY little he actually understands or conveys his understanding of this topic in his book.

In my fourth year neuroscience class I wrote a term paper called "This is your brain on Chaos". The course was all 4th year neuroscience geeks, and an equal number of faculty. Almost the ENTIRE University of Lethbridge neuroscience faculty at that time. Each of us had to select a topic, then spend the semester presenting / teaching academic papers and research on that topic as it related neuroscience.

My entire "education" in fractal geometry was reading Chaos: Making a New Science years earlier, and a few papers about its application to the structure and function of neurons or collections of neurons. I can't event write a single equation of any known fractal figure, and yet it was obvious Taleb knew or expressed less understand than I 22 years after that single semester of study about fractal geometry in neuroscience:-?

At the beginning of the book he said trees "resemble" fractals. That's kinda true, in the same way a single stick of spaghetti resembles a flag pole, but barely more:-? To test whether something is fractal, rather than has a spaghetti / flagpole resemblance you can apply mathematical tests to see if they're genuinely fractal, or just resemble one. I think there were three test, but I only remember two. Still, they're more then enough to kick his ass on this topic.

The local flow test, I think;-) It measures the direction something moves in or grows compared to where it is, or was in the past. If it's local flow is fractal, you can mathematically create a function which will accurately predict the shape, growth pattern or movement.

Whether it's dendrites on a neuron, or branches on a tree, you can accurately predict the "flow" or direction of movement of growth of their branches. If the growth is fractal, there's a fractal that can match it, and proves the pattern of growth is fractal. If not, it's NOT fractal.

In case you care, dendritic growth / propagation ISN'T fractal, it's space filling. The same is almost certainly true of trees because both resemble one another closely. Space filling is just like it sounds. The branches just grow toward the middle of the largest available open space. It's like a music festival, people find the biggest open space available, if it's large enough, and they're selfish enough, they sit down in the very middle of the open space. Same thing.

For trees and dendrites alike it a simple heuristic, fill as much space as we can within the available volume. For dendrites, it minimizes the average distance to any other dendrite or neuron so they have less distance to grow to connect with any of their neighbors and form a new neural pathway. It's an adaptive heuristic for an adaptive biological computer, NOT antifragile, but adaptive via the evolutionary process of ELIMINATION / extinction. Trees are likely ( though I don't know for a fact ) the same. By growing toward the middle of the largest available empty space they can fill it more completely with leaves to collect sunlight to grow, thrive and survive.

Self similarity across scales. Fractals by definition should be impossible to estimate the scale of, if the edges of the object are outside the viewing frame. Whether you zoom in 1, 2, or 10X, a true fractal will look identical. Does a tree, or a neuron look exactly the same viewed at different distances or scales? If you zoom in with a camera from 1X, 10X...100X, does it look the same? If not, it's not fractal.

Put to this test, dendritic branching isn't fractal. Trees are unlike to be. To be fair, self similarity across scales has a limit in the real world of physical non-mathematical constructs, but it SHOULD be self similar across at lest some scales and trees and dendrites AREN'T. My favorite example of a REAL fractal in nature is green cauliflower aka broccoflower or romanesco broccoli.

Taleb takes almost the entire book to get to his self proclaimed "Magnum Opus"/ "Philosopher's Stone" which is nothing more than saying that power functions go up faster than root functions. That's obvious, and unusable to anyone reading his book. HOW can I use this obvious and artificial example to make the best stock picks because I didn't need you to tell me it's worth trying a risky drug if I'm already dying and it might save me:-?

Here's the main thing Taleb is COMPLETELY wrong about, there IS NO universal antifragility, or fragility because any property, or system of properties are adaptive, robust, resilient... to one event, but simultaneously made more fragile, to the opposite or different events.

He may be right in some specific domains, though little nothing about this book that isn't intuitively obvious is presented or explained in a usable way. But I get the feeling his goal was self glorification, NOT education or the civic good.

So my main takeaway is the same ONE practical thing I learned from the black Swan:-? And little else:-?

Be ultra conservative with 90% of your risk taking, then take high risks with high payoffs with the 10% you can afford to lose.

tim #bgreen??

P.S. What USEFUL, applicable things have you learned, if any from Nassim Nicholas Taleb?

P.P.S. My Next Article: Somebody Not Agreeing with You Doesn't Mean They Don't Understand.