A Summary of Antifragility - Personal Antifragility

Antifragile by Nicholas Nassim Taleb

Antifragility is a quality found in dynamic systems, they are everywhere in nature. It is about the “opposite of fragile”, the things that benefit from stress, disorder and volatility. And it allows systems to grow and improve in an unpredictable and volatile world. Yet, we are trying to avoid volatility and uncertainty in our constant fear of the unpredictable we have lost ourselves in controlling as much as we can, but in vain since the world out there is utterly unpredictable.

This book explains why we need to do exactly the opposite: to embrace the volatile and uncertain environment that is essential for antifragility qualities to emerge. Otherwise, through our compulsive need for order and control, we are making ourselves more vulnerable, more fragile to shocks in a world that is utterly unpredictable and uncertain.

Fragile vs. Antifragile

“If everything top-down fragilises and blocks antifragility and growth, then everything bottom-up thrives under the right amount of stress and disorder.” N. N. Taleb, Antifragile.

“you can state confidently that something is fragile, but you can’t predict that a particular shock or event is going to happen.” N. N. Taleb, Antifragile.

The Fragile is what hates problems and mistakes.

A glass won’t withstand a shock encountered with the floor, especially if dropped from the highest shelf in your kitchen as you were trying to pick it from the tip of your toes. Fragile systems try to avoid all failure hence the need for control and prediction, they can’t let go and see what will happen. This behaviour leads fragile systems to make rare but large errors that are costly, to the point of no return, putting themselves at a high risk of debt.

Fragile

“Sword of Damocles” situation
No skin in the game
Academic
Scientific theory
Directed research
Produce large but rare errors
Waterfall process
Centralised
Nation-state
Concentrated source of randomness
Via positiva (add/complexify)
Efficient optimisation
Rules
Tourist (follow plan)
Corporate employment
Bureaucrats
Statutory law, legal code
Ideology
Post-traumatic stress
Chronic stressors
Large
Separable

The Antifragile is what loves problems and mistakes.

If I keep the analogy of a glass, you would have to imagine a glass that creates another brand new glass as you smash it on the floor, and the harder you smash it the more glasses it produces. It’s like the mythological Hydra when cutting one head leads to 2 more heads to grow. Antifragile systems like failure as long as they are small and incremental since it’s an opportunity for learning. This behaviour pushes antifragile systems to seek small, benign, quick to overcome stresses and errors.

Antifragile

Hydra
Soul in the game
Erudit
Heuristics
Stochastic tinkering
Produce small and reversible errors
Agile process
Decentralised
Collections of city-state
Distributed source of randomness
Via negativa (remove/simplify)
Functional redundancy
Virtue
Flâneur (free to let it happen)
Freelance practices
Entrepreneurs
Common law, equity
Mythology
Post-traumatic growth
Acute stressors with recovery
Small and not specialised
Holistic

Antifragile: system’s properties

Although the topic seems quite undefined, what Taleb is philosophising with is the nature of complex dynamic systems, finding out that there are properties that make a system fragile, and by opposition thinking of the properties that constitute antifragile systems. Here are what I think Antifragile system’s properties are:

They are resilient to change, and can recover from disruptions.
They get stronger in response to stressors, shocks, and volatility.
They are complex and nonlinear, and exhibit feedback loops.
They typically have a high degree of redundancy and flexibility.
They are self-regulating, and can adjust to changing conditions.
They are able to self-organise and adapt in response to changes in their environment.
They are designed for failure, and can tolerate and learn from mistakes.
They are decentralised, and have many small, independent parts.
They are open and constantly evolving.
They are unpredictable and chaotic, and can generate surprise and novelty.

The properties of Antifragile systems are very similar to system thinking for any dynamic system to be and stay “healthy“.

Antifragile: design principles

Therefore, there are certain design principles and mindset that emerge from Taleb’s work that can help reduce the fragility of the systems we design.

1. Keep it simple

Complex systems do not require complicated rules, the simpler the rules the better! We must resist the urge to control and the temptation to respond to complexity with complex rules, as they have a tendency to produce more dependencies and therefore generate a cascade of unintended consequences.

Don’t respond to complex issues with complex regulations (Complex systems require simplicity):
Rights, values, principles, good deeds, and true acts of virtue can’t be taken away, they open-up and guide, rather than reduces the playground and control.
Under-regulate and provide simple laws or guiding principles (the simpler regulations and policies, the better and transparent it is):
Complex administrations and over-regulations can tip a system where its productivity is becoming less since it has to maintain the bureaucracy involved.
Complications leads to multiplicative chains of unanticipated effects (inter-dependencies and nonlinear behaviours):
Simplicity is complexity, making thing simpler doesn’t mean we remove nuances, just let the organic take place rather than trying to control everything (e.g. DNA with only 4 amino-acids).

2. Decentralise

Decentralised systems are better to be able to learn from randomness because adverse impacts are contained and prevent error propagation. A system with decentralised autonomous sub-units can watch and learn from each other as each unit improvises in response to unexpected events. A small, decentralised system proves to be far more antifragile than a very large, centralised system.

Have small and various sub-units (helps having redundancy, splitting your eggs across baskets, multiplying opportunities for upsides):
Everything bottom-up thrives under the right amount of stress and disorder. About everything top-down fragilise or block antifragility and growth.
Simplify to its smallest, autonomous self-efficient unit for a given function:
The larger and more complex a system is, the more it becomes exposed to volatility.
Give the freedom to collaborate and trade naturally from unstructured environments (behave like a super-additive function)
The freedom to explore options and opportunities, multiplied by the number of autonomous sub-units, added to the cross-collaboration, equals the unlocking of open-ended pay-offs (maximise innovation by serendipity).

3. Think holistically

This has in many respects some links to the second principle. Taleb points out that the antifragility of a system often comes at the expense of the fragility of its components, such as the failed 70% start-ups’ that drives the overall entrepreneurial success or the death of individual organisms that contribute to the antifragility of nature. Antifragile systems contain adverse impacts and increase the potential for learning from what happens to constituent units in the lower layers.

Tend at the system level, not at the event level:
Let it go, it can improve without direct intervention (sometime, some things need to break for the system to improve).
Take in consideration the risks that will compromise a prompt recovery:
Take care of the worst and the rest will take care of itself.
Respect the past and have curiosity about the historical record to learn from the elder’s wisdom:
If things have been around for 100 years, they will likely still be there for another more 100 years.

4. Build in redundancy and overcompensation

Redundancy in systems is a key to antifragility. As Taleb suggests, nature loves to over-insure itself, whether in the case of providing each of us with two kidneys or excess capacity in our neural system or arterial apparatus. Overcompensation is a form of redundancy that helps systems to opportunistically respond to unanticipated events. Perceived as inefficient or a waste of resources, it can actually prove to be enormously helpful, not just to survive unexpected stress, but to be able to turn the turmoil into opportunities.

Have redundancy, margins of safety, buffers and contingency plans:
Uncertainty and nonlinearity of the world means there are risks out there so, redundancy is a way to avoid bankruptcy in case the worst happen, it is the exact opposite of debt.
Safeguard enough to recover from a downside, and take small bets to maximise your upsides:
Don’t spread thin, add contingency when we know we may need a reach from time to time (it’s the difference between efficiency and effectiveness).
Reduce or remove asymmetry in risks, decrease the downside and increase the upside:
Get 90% of very conservative bets, and 10% of high risk bets (downside is to lose 10%, but you can recover from it, however if you win, you will have a big upside).

5. Embrace randomness

Taleb warns against the tendency of planners to try to eliminate volatility or unpredictable disruptions to a system. He stresses the paradox that efforts to eliminate this randomness will only intensify the vulnerability of systems to damage from disruption. Randomness is the root cause of serendipity, the source for most of the great discoveries that have moved our society forward. Without randomness, there can be no serendipity.

Give freedom to explore option, and allow for things to go wrong so it can learn from mistakes:
You get pseudo-order when you seek order. You only get a measure of order and control when you embrace randomness.
Keep things small and slow enough to recover and integrate information, so that distribution of randomness is high in volatility:
Random mutations and ‘happy’ accidents (serendipity) increase the chances of discovering something great.
Adopt a flâneur approach, open your system for discoverability, and nurture it for diversity:
Touristification is the systematic removal of uncertainty and randomness for the sake of efficiency and comfort.

6. Ensure participants have skin in the game

Participants within a system must face the consequences of their actions, to endure failure and enjoy success, so that they are motivated to learn and to not take any unwarranted risks. A system is fragile when a few selected, particularly those with an abundance of resources or power, are able to capture the benefits for themselves while exposing others to the risks, losses or harm. Bail-outs are highly dysfunctional.

Use downsides as feedback loops, it has to be a risk:
At no point in history have so many non-risk takers, that is there is not personal exposure, exerted so much control!
Don’t allow people to weight-in if they haven’t got any skin in the game:
People voting for, or being in favour of wars, should have a least a relative exposed to combat.
Be loyal to people (relation and moral commitment), but rationally opportunistic with things and ideas:
Thou shall not have antifragility at the expense of others’ fragility.

7. Value practice over theory

Taleb is pretty vocal about his contempt for theoreticians and his admiration for practitioners. He believes that we overestimate the role of research and analysis and underestimate the role of practice and experimentation when it comes to advance knowledge and well-being. Most suggest that theory and research lead to new insights that in turn shape our practices but most of our significant knowledge breakthroughs came from experimentation and tinkering.

Let failure happen, so it can learn by itself. But support buy making failure acceptable:
We should celebrate mistakes in the pursuit of learning and treat with respect and dignity those who failed in order to advance society.
Try the things you don’t know if it will work in order to discover something (riskiest assumption):
Innovation benefits from error, gains from uncertainty (like disorder and volatility).
Transform fear into prudence, pain into information, mistakes into initiations, desires into undertaking:
Information gives ideas. Ideas get ‘nourished’ from ‘attacks’. Ideas grow into something tangible and valuable, producing antifragility.

Strategies for antifragility

In Taleb’s view, the end goal for any antifragile strategy is to achieve convexity. Taleb draws a core contrast between concave and convex strategies. The question is to know whether there is more upside or downside. If the upside increases, you have positive asymmetry and a convex strategy. If the downside increases, you have negative asymmetry and a concave strategy — something to be avoided at all costs.

1. Take the via negativa

“We know a lot more about what is wrong than what is right… Negative knowledge (what is wrong, what does not work) is more robust to error than positive knowledge (what is right, what works). So knowledge grows by subtraction much more than by addition…” N. N. Taleb, Antifragile.

It’s like in medicine, you shouldn’t take the risks of side-effects if you are slightly or mildly ill and rather let your immune system do its job, but you should take a lot more risks and see 10 doctors if you have a terrible illness.

Remove the complex and the unnecessary hurdles, simplify the regulations and policies, don’t over-optimise nor over-structure:
Complex administrations and over-regulations can tip a system where its productivity is becoming less since it has to maintain the bureaucracy involved.
Reduce to the smallest functional units, decentralise the system, multiply or reproduce rather than growth centrally:
The larger and more complex a system is, the more it becomes exposed to volatility. There are limits to growth. In opposition, it’s often the small things that impact the most (less is more).

2. Filter signals out of noise

“The more data you get, the less you will know what’s going on.” N. N. Taleb, Antifragile.

Be suspicious of the noise, filter out the futile or the inconsequential, listen to the right signals, look behind the single tree that is hiding the full forest.

Focus on the “one” thing that is the most beneficial and ignore all the rest:
More data doesn’t solve problems, like adding more noise, it will make the signal weaker.
Procrastinate, do only what you feel you have to do, and let time show you more information to make a wise move:
Procrastination helps filter out the noise, the unnecessary, the futile. It subtract and refine down to the essential signals, the very needed, the most important.

3. Focus on options

“An option is what makes you antifragile and allows you to benefit from the positive side of uncertainty, without a corresponding serious harm from the negative side.” N. N. Taleb, Antifragile.

An option allows you to take the upside if you want but without the downside. Optionality, meaning the availability of options, reduces the need to understand or know something.

Wherever possible, seek out options with open-ended pay-offs rather than closed-ended ones, invest in people rather than in plans:
Optionality – the availability of options – reduces the need to understand or know something.
Domesticating uncertainty is having options:
Options allow for benefiting the positive side of uncertainty (without serious harm from the negative side).

4. Pursue barbell approaches

“Play it safe in some areas to mitigate the potential impact of negative unpredictable events while at the same time taking a lot of small risks in other areas to increase the benefit of positive unpredictable events.” N. N. Taleb, Antifragile.

Pursue a bimodal strategy, don’t play in the middle, be both aggressive and paranoid in carefully selected areas while avoiding the complacency that the deceptive middle produces.

Focus on decreasing the downside first rather than increasing upside:
Get 90% of very conservative bets, and 10% of high risk bets (downside is to lose 10%, but you can recover from it, however if you win, you will have a big upside).
Benefiting from volatility over time is about making more when right, hurting little when wrong:
Avoid playing in the middle, play it safe in some areas to mitigate the potential impact of negative events, while at the same time taking a lot of small bets in other areas to enhance the benefit of positive events.

5. Be curious and seek challenges to learn

“To this day I still have the instinct that the treasure, what one needs to know for a profession, is necessarily what lies outside the corpus, as far away from the centre as possible … Curiosity is antifragile and it is magnified by the attempts to satisfy it.” N. N. Taleb, Antifragile.

Curiosity and its close cousin, discovery, like disturbances. Disturbances create unexpected opportunities to learn more and help us to grow stronger in the face of challenges that we had not anticipated.

Aim at learning to learn how to learn and foster curiosity:
Curiosity is magnified by attempts to satisfy it. As the circle of my knowledge expands, so is the perimeter of my ignorance. Learning open new fields for our curiosity to explore.
Encourage people to be autodidact, eliminate curriculums, and let them learn by and for themselves:
Education should be the formation of character and personality with the acquisition of true and practical knowledge.

6. Tinker and experiment

“Focus on building and accessing tacit knowledge rather than rationality and explicit knowledge … Carefully pick the areas for experimentation where there is potential for significant upside.” N. N. Taleb, Antifragile.

Value experimentation and tinkering over theorising, structure them so that they are small in potential harm and so that you can pursue many of them.

Have feedback loops to coincidentally tinker the system, to improve and push the edges:
Random tinkering is a way to multiply the option to find something great (serendipity).
Try things out, follow a fast trial and error approach, be willing to pivot and change radically direction:
The mistakes and errors are safe as long as they remain confined (don’t risk of spreading errors). And the good thing is that it gets better overtime (easier, faster, more enjoyable, etc.).

7. Seek heuristic rather than logic

“Antifragility implies … that the old is superior to the new. What survives must be good at serving some (mostly hidden) purpose that time can see but our eyes and logical faculties can’t capture.” N. N. Taleb, Antifragile.

Be deeply suspicious of abstraction, theorising and rationality.

Study for life, not for grades or diplomas: Focus on building and accessing tacit knowledge rather than rationality and explicit knowledge.
Be deeply suspicious of abstraction, theorizing and rationality. Use them as tools to exercise truth and don’t mistake them for the truth itself:
“Logic excludes — by definition — nuances, and … truth resides exclusively in the nuances.” Ernest Renan

In a seed-shell

Let go, don’t try to predict or control.
Sense and respond where the system wants to go.