If you prefer to listen rather than read, this blog is available as a podcast here. Or if you want to listen to just this post:
February 2020, the last month of normalcy, probably feels like a long time ago. I spent the last week of it in New York City. Which was already ground zero for the pandemic—though no one knew that yet. I was there to attend the Real World Risk Institute. A week-long course put on by Nassim Taleb, who’s best known as the author of The Black Swan. The coincidence of learning more about black swans while a very large one was already in process is not lost on me.
(Curiously enough, this is not the first time I was in New York right before a black swan. I also happened to be there a couple of weeks before 9/11.)
Before we go any further, for any who might be unfamiliar with the term, a black swan is an unpredictable, rare event with extreme consequences. And, one of the things I was surprised to learn while at the institute is that Taleb, despite inventing the term, has grown to dislike it. There are a couple of reasons for this. First people apply it to things which aren’t really black swans, to things which can be foreseen. The pandemic is actually a pretty good example of this. Experts had been warning about the inevitability of one for decades. We had one in 1918, and beyond that several recent near misses with SARS, MERS, and Ebola. And that was just in the last couple of decades. If all this is the case, why am I still calling it a black swan?
First off, even if the danger of a pandemic was fairly well known, the second order effects have given us a whole flock of black swans. Things like supply chain shocks, teleworking, housing craziness, inflation, labor shortages, and widespread civil unrest, to name just a few. This is the primary reason, but on top of that I think Taleb is being a little bit dogmatic with this objection. (I.e. it’s hard to think of what phrase other than “black swan” better describes the pandemic.)
However, when it comes to his second objection I am entirely in agreement with him. People use the term as an excuse. “It was a black swan. How could we possibly have prepared?!?” And herein lies the problem, and the culmination of everything I’ve been saying since the beginning, but particularly over the last four months.
- People excuse their unpreparedness because they don’t understand how random the future really is.
- Technology and progress have given us the illusion of predictability, but it has actually made black swans more common.
- Because of their impact, the future is almost entirely the product of these black swans.
- And more terrifying than that, technology has made apocalyptic black swans possible.
Accordingly saying “How could we possibly have prepared?” is not only a massive abdication of responsibility, it’s also an equally massive misunderstanding of the moment. Because preparedness has no meaning if it’s not directed towards preparing for black swans. There is nothing else worth preparing for.
You may be wondering, particularly if black swans are unpredictable, how is one supposed to do that? The answer is less fragility, and ideally antifragility, but a full exploration of what that means will have to wait for another time. Though I’ve already touched on how religion helps create both of these at the level of individuals and families. But what about levels above that?
This is where I am the most concerned. And where the excuse, “It was a black swan! Nothing could be done!” has caused the greatest damage. In a society driven by markets, corporations have great ability to both help and harm by the risks they take. We’re seeing some of these harms right now. We saw even more during the 2007-2008 financial crisis. When these harms occur, it’s becoming more common to use this excuse. That it could not be foreseen. It could not be prevented.
If corporations suffered the effects of their lack of foresight that would be one thing. But increasingly governments provide a backstop against such calamities. In the process they absorb at least some of the risk. Making the government itself more susceptible to future, bigger black swans. And if that happens, we have no backstop.
Someday a black swan will either end the world, or save it. Let’s hope it’s the latter.
One thing you might not realize is that donations happen to also be black swans. They’re rare (but becoming more common) and enormously consequential. If you want to feel what it’s like to have that sort of power, consider trying it out.
Are black swans really Black Swans? Every swan in Europe is white, Europeans never saw a different swan so assumed all swans are white. Hence finding black ones outside of Europe was, well probably not an earth-shattering shock but a bit of a surprise. I mean if tomorrow if they announce deep in the Amazon they discovered a rare group of blue swans…. that would be surprising a bit I suppose but it’s not like my mind is shattering. It’s not like they argued it was a law of nature black swans didn’t exist, just that none had never evolved. I’m wondering if perhaps the concept is overplayed.
The core of his insight to me would be that we are inclined to assume some things will follow a normal distribution and sampling will bear that out in many cases but in reality, the distribution is fat tailed. The problem with this is when tails are fat, you will start getting impacts like the once in a century flood actually comes around every few years. In terms of finance, assets that normally have no correlation do correlate under extreme cases and at that point the normal distribution of returns no longer holds. Hence normally if you are writing insurance against mortgage default, you are pretty safe if you spread your policies out over the whole nation. Individual regions may suffer downturns and increased foreclosures, but never all at once. But then in an economic collapse this diversification no longer matters.
I’m pretty critical of how we handled this pandemic, but I’m less so from the idea of ‘preparedness’. Preparedness reminds me of the movie Signs where Mel Gibson’s daughter was constantly leaving glasses of water all over the house for reasons she couldn’t explain…but when the aliens invaded they discovered water burned them so smashing the glasses with a bat saved the family. Yes great, that sort of thing is a black swan but absent supernatural hints, you’re not going to think of that or do it Practically we would be ‘prepared’ by being flexible rather than trying to guess things like keeping mass stockpiles of N95 masks molding away in warehouses just in case.
And we were actually pretty flexible. We realized pretty fast that we needed to shut down. We also pivoted pretty fast towards masks as a way to re-open yet still reduce risk. Schools moved to online learning fast, workplaces moved to work at home. The vaccines were developed rapidly and while all long the way there was huge blunders (forcing meat packing plants to reopen, for example), we actually adjusted pretty fast and set in place a lot of practices that will likely make a new virus of this type harder to get hold.
Where I think we are confronting an uncomfortable new normal is not the collapse in authority but the collapse in respecting authority. Previous cultural regimes had limited ways to verify information in dispute. Someone may pull out a dictionary or an encyclopedia. You may have called a librarian but more likely to ‘settle a bet’ you might just grab some third person and asked him to declare who was right.
The explosion in information has been matched now by an explosion in do-it-yourself epistemology. A lot of people I think have cast this as a ‘decline in trust’ in authority but it’s more like a “decision to distrust”. Pick any topic you want and you can quickly get an intellectual toolset ready made for you to pick up and start swinging around. Just google “everything you think you know about X is wrong” to get started. Now your mileage on these tools will vary, you can get a screwdriver from Snap-on or from the dollar store after all. Perhaps we have a Sophistry Enablement Age (SEA) in play here.
Does SEA, though, mean we become more fragile? I don’t think so in terms of reaction to things like pandemics. Like it or not when the shit hit the fan, red states closed down too, shut their schools, pleaded with people to vaccinate. They created monsters and caused more death for their people but in the big picture that’s a blip.
SEA might mean we have to recalibrate what it means to make a collective decision. The population needs more to ‘get on board’ and very basic ideas need to be rearticulated for the simpler minded who have now joined the discussion. It may be helpful to make challenging old orthodoxies more common. In the past doctors when to their graves refusing to believe things like washing hands prevented deaths in childbirth or ulcers were caused by bacteria that could be treated with antibiotics rather than ‘too much spicey food’ and removing people stomachs.
Can we engage in advancement with SEA? I don’t know. It seems we can privately. Musk can get his people onboard with a bunch of impossible ideas and sometimes they work and sometimes they don’t. Sophists are free to quite and spend their afternoons tweeting insults at Tesla. (and the SEA might be good in that it’s the ultimate way to distract and occupy the time of those who would otherwise be inclined to be unhelpful or troublesome. Kind of like dangling keys in front of babies or sitting them down in front of an ipad). We might be unbalanced, though. We are good at private goals but can’t accomplish collective ones well and we’ll get worse and worse at that. For example, ignoring climate change year after year but tripling down on refining the most killer apps imaginable. Existential threats that can’t be incorporated into the private sector will get ignored hence increase the fragility in the long run.
These days Taleb does seem to use the term “fat-tailed” more often than black swans these days, for what it’s worth.
As far as prepardness I think there are asymmetries we could be exploiting that we’re not. Places where it’s cheap to be wrong and where it’s expensive to be wrong. It’s cheap to be wrong about the need for N95 masks in the direction of having them but not needing them, and it’s expensive to be wrong about N95 masks in the direction of needing them, but not having them.
As far as your thoughts on the SEA, that’s some of your best stuff. I’m stealing it. (Though in the process I’ll almost certainly turn it to my own evil ends…)
Thanks for the theft…. The mask issue is interesting because it contrasts with Asia a lot. 1918 was bad there, masks have always been part of the culture since. People tended to have masks at home and when Covid started it was no big deal to bring them out. One could quibble over the difference between having N95 masks and lesser masks, but in the US it didn’t even occur to us that you could make a mask out of cloth at home until a few weeks in. The N95/PPE consumption in hospitals was insane, thousands were consumed every day and staff started trying tricks like using UV lights to clean their N95’s, protecting them with multiple lesser masks on top, and so on. The official rule of tossing an N95 after each use would have meant even a large stockpile would have been exhausted quickly.
If the new norm going forward is that most Americans have a bag or two of N95’s at home that will mean there will never be such a shortfall again. Even though a surge in demand might cause some shortages (like toilet paper), the system will be in place to increase production. The revolutions that I think that need to come from this are:
1. Taking respiratory viruses more seriously, monitoring them and trying to develop rapid tests to measure what is going around and push for things like universal cold/flu vaccines that can provide immunity to whole families of viruses.
2. Taking ventilation seriously. Right now there is official HEPA and UV light systems. There are also a lot of ad-hoc affairs going on like UVC lamps (which can be harmful if people get exposed to them), CO2 monitors. When schools reopened I tried to offer to buy a few portable HEPA filters but they assured me they had purchased $600 air purifier machines for each classroom plus busses. I think going forward we need real research here and that should be thought about in building design and upgrade.
This is hard research. The reality is we never see infection happen, we can only infer it from facts like did someone touch a doornob the infected person use or were they on opposite sides of the room and never got within a few feet of each other? It’s hard to know exactly what happens when we exhale and virus goes out into the environment. I still get stupid people telling me masks can’t work because the holes in a mask are bigger than the virus.
Just showing how hard is 50 scientists pulled together a bunch of supercomputers to build a simulation of a single virus encased in a drop of water. 300m atoms are needed to simulate a single virus and about 1B more are needed for the drop of water it is encased in.
Keeping this all simple, though, it sounds like ‘black swans’ are:
1. Processes that have a fat tail rather than normal distribution.
2. The fact that many samples of data will be consistent with a normal distribution, people will assume they are working with a normal distribution assuming events in the tails will be thin (i.e. a bad storm every century only).
3. Ignoring fat tails is a road to ruin, but over a period of time…which could go on for years or even decades…everything will seem ok.
One podcast I follow (Decoding the Gurus) recently asserted that Taleb’s style is a bit of stage show here. His big concepts about math are not all that amazing to anyone who does real statistics for a living unless they’ve come out of 1957. But then unlike 1957 anyone today with Excel can do quick regressions on any data set they please often barely understanding the other data that is spit out like P-values or F-scores.