Sense out of Nonsense (part 4) - Butterfly effects and Pandemics

BUTTERFLY EFFECTS AND PANDEMICS

Prasad Chaphekar

The current pandemic has exposed the fragility of current human society. The more connected and interdependent we are, the more fragile we get, at least, this seems to be the case today. As Nassim Nicolas Taleb, a Lebanese- American scholar working on problems of randomness, probability, and uncertainty points out, systems have to be designed with antifragility in mind. Antifragile systems are those which thrive in times of disorder and uncertainty. He suggests that decision makers have to invest a small but significant amount of resources in developing systems which would thrive or benefit if the current system collapses. To give an example, it is counterintuitive, but anti-fragile for aircraft companies to invest in companies which develop software for video-conferences/ net meetings. An industry which thrives on physically moving people can only become anti-fragile by betting that one day, people would NOT be required to move physically.

Butterfly effects and Pandemics

Are pandemics or natural disasters nature’s way of punishing us? Not necessarily. To me, it seems like the higher order effects of our actions, hitherto unknown, are manifesting themselves.

When we are playing billiards, we hit the cue ball towards the target aiming to push another ball in the pocket either directly or through an intermediate ball. However, we do not bother about what happens to the cue ball or the intermediate ball after it hits the intended target. These balls move further and knock other billiard balls out of position. These collisions are higher order effects, which we choose to ignore. The billiards table is still a highly deterministic system in an extremely controlled environment. With physics, we can reasonably predict the result of the first few collisions. However, beyond first few collisions, the error in predictions will grow with every secondary collision.

This effect is called the “butterfly effect” where a small variation in input parameters leads to a very large variation in the output parameters. The effect is so named over a statement that “flapping wings of a butterfly at one end of the world may cause a hurricane at the other end”. The effect was discovered by chance when a scientist, who was trying to predict weather patterns by a computer simulation, entered some values manually to save time. However, the end result was widely differing from earlier simulations. After thorough analysis, the only reason he found for the huge difference, was that he had rounded off a few decimal places, when he entered values manually.

Weather has been notoriously difficult to predict precisely because of the presence of a high number of such higher order effects and the interdependence of various factors which create the weather.

The funny thing is, with the benefit of hindsight, everything is analyzed and written as if it was perfectly predictable. Though we can often foresee a general direction, we cannot presume to predict the exact course of events to follow, as even small changes will lead to large deviations in the long term (and sometimes in the short term). Long term predictions are thus, useless at best and harmful at worst. Let us take an example of a factory polluting a local water source. This can lead to many different outcomes. Let us analyze 2 possible outcomes:

1)    The marine life gets polluted. The population in the area stop using the water source. Water shortage forces people out of the area. The factory is closed down, leading to job losses and economic reversal.

2)    The marine life gets polluted. A concerned citizen in the nearby area raises a movement against the pollution. The government imposes strict pollution control measures. No factories are closed down, but they are given sufficient time to comply with the pollution norms. The pollution is stopped and everybody is a winner.

Both courses seem rational and in hindsight, perfectly predictable. However, given only the fact that a factory is polluting a water source, it is impossible to state which course would follow. The same is the case with natural and man-made disasters.

Nagarjuna, Buddhist philosopher propounded the theory of “shunyata”. In this, he has examined causation thoroughly. He states that cause and effect are not distinct. Effect physically depends upon the cause and the cause depends notionally on the effect (i.e. without any effect, no one would look for a cause). Further, each cause is an effect of some other cause and each cause leads to many effects. This leads to an ever-increasing chain of cause and effect traced either backward or forward.

(On a side note, individuals whom we call enlightened have/had probably raised their consciousness above most others so that they are/were able to perceive a far higher number of higher order effects than most of us can)

When you hit a rock 100 times with a hammer and it breaks with the 100^th stroke, did it break because of the 100^th stroke alone or also because of the conditions created by the previous 99 strokes?

Till humankind was physically isolated and living in small groups, everything was “local”. Epidemics came with cities and floods devastated populations because the people were living in large cities next to big rivers. So, it can also be argued that evolution is responsible for devastation caused by floods. It was evolution that lead to us walking erect, which lead to smaller birth canals in women, which lead to babies being born much before they matured (as compared to other mammals), which meant that humans had to live in societies to survive and care for their young, which lead to big cities and coupled with the need to have water, lead to settlements near big rivers and the destruction by floods seems inevitable. Thus, the chain of causation can be shown to be going as far back as you desire. As explained earlier, hindsight is always 20/20. The higher order effects seem “obvious” only when they manifest themselves.

So, it is difficult to say that we are causing anything beyond first order effects. We can alter our collective behavior to increase/decrease the odds. But how these behaviors would affect us, say 10 years in the future is impossible to predict.

Coming back to Nassim Nicolas Taleb, we would never know for certain what shocks awaits us, but we can design our systems presuming shocks are definitely going to happen. In general, increased fragility will lead to higher chances of large scale destruction and lower fragility (more robustness) and more antifragility will result in higher chances of us surviving/ thriving the catastrophes yet to come.