Thoughts of The Week #53

“Thoughts of The Week” is a review of the most interesting thoughts I encountered on Twitter and on other mediums during the last week. It is free, but if you like it, you can support me on Patreon.

1/5: Design for failure

“We need global circuit breakers.” – Francisco Amadeo (link)

In the current globalized words, every country borders China. The increased ease and speed of travel does not only apply to people but to pathogens as well. We need to design systems not only for success, but for failure too, especially when failure has multiplicative dynamics.

Design for failure mostly means two things: redundancy and circuit breakers. Redundancy protects the system from the second-order effects of shocks, circuit breakers prevent shocks from propagating.

The defining feature of circuit breakers is that the logic that makes them engage is local. There is no such thing as a centralized circuit breaker.

In my essay “What will we have learnt” , I argue that centralization is short-sighted (“centralization is only efficient when viewed from the center”) and that localism has larger long-term advantages (which become evident only with time spans large enough so that efficiency tends to reliability).

In his Spectator column, Rory Sutherland wrote: “Solve for the specific and you may also solve for the general. The reverse does not apply.” Another reason for solutions that work locally.

2/5: Conservation of instinct

“It is easier to build a strange new form of office politics than it is to get rid of office politics, easier to have idiosyncratic personal attachments than none.” – Matt Levine (link)

“A Conservation of Religiosity principle: human nature demands religion in some form, but the suppliers differ across epochs and cultures.” – Geoffrey Miller (link)

It is easier to change the way an instinct manifests, rather than to suppress it.

A practical example: vaccines. It is extremely difficult for a belief to reach 100% consensus, as the higher the adoption, the higher the incentives for individuals to profess an incompatible fringe belief to accumulate “loyalty points” with a fringe culture. For example, anti-vaxxers opposing government-mandated vaccines are rewarded with status within the anti-vaxxers community.

If this is the incentive for anti-vaxxers, no single vaccine will reach 100% adoption. The solution: to manufacture two government-approved vaccines for each disease but only recommend one. Anti-vaxxers will express their in-group loyalty by refusing the recommended vaccine and choosing the other one.

3/5: Managing tail events

“Managing the tail event must happen PRIOR to the event.” – Paul Portesi (link)

Flights to and from countries affected by the virus should be blocked before the outbreak is too large. Otherwise, carriers of the pathogen will already have begun spreading it to new countries.

Two implications:

• Decisions on tail event management have to be made on leading indicators, not on lagging ones. For example, counting the confirmed cases of infection is a lagging indicator, for symptoms only show after days of incubation.
• Tail event management is an inherently inefficient venture – the reaction has to be larger than the current hazard – and yet a necessary one. No one throws out a fire-extinguisher saying “it hasn’t been an efficient purpose, I didn’t get to use it”. Efficiency has to be on the do-not-measure list of metrics.

4/5: Comparisons

“You must never compare deltas when gammas are very different.” – Nassim Nicholas Taleb (link)

Translated in plain English, it means: you should never compare rate of changes, when the rates of change of rate of changes are very different.

In this case, it applies to comparing the coronavirus to the seasonal flu or to the 2003 SARS epidemic. The risk posed by the coronavirus derives from multiplicative dynamics. Under multiplicative dynamics, a small difference in parameters can produce large differences in outcome, and a small difference in rate of change can produce large differences after compounding is taken into consideration.

The problem of the coronavirus is not its mortality rate now, but its mortality rate when hospital beds are full, for example. Seasonal flus don’t fill hospital beds. The coronavirus can – as it proved the case in China. That’s a whole different scenario to prepare for – or worth taking large-scale measures to avoid in the first place.

5/5: Instincts

“I listen to my instincts because they are older than me.” – Joe Norman (link)

Instincts are the genetic memory of the problems our ancestors faced.

Too often, we make the mistake of believing that our problems are unique. Our instincts are there to prevent this, to prevent us from over adapting to ephemeral trends, to remind us that what worked for others is often the most likely solution to work for us too.

Bonus: essay of the week

This essay on monetary unions, by Conrad Bastable (found via Rory Sutherland). It is long, but it also has a summary at the top.

My own essay of the week

The most important question to stop the coronavirus.

…

If you liked this newsletter, please share it with your friends!
They can subscribe here: luca-dellanna.com/newsletter

Otherwise, you can support me on Patreon

Become a Patron!

…

(Here is my general disclaimer.)

Quotes are edited for punctuation and grammar. Eventual formatting is mine. Also text outside of italicized quotation marks is mine. The inclusion of quotes does not imply my endorsement; merely, that they gave me food for thought. I did not optimize this review for clarity, but for its ability to spark thoughts in the reader.