How to use doubts to understand the world.

This book is really a timely read for us that face a surrounding with rapid changes and a precarious future. Bombarded everyday with almost alarmist news on wars, potential joblessness due to teach advance, and natural disasters intensifying every year, it is exhausting for us to internalize so much of these uncertainties that may threaten our own sense of security.

The primacy of Doubt is a user manual on quantifying uncertainties and a guide on decision making for the future upon the uncertainties we have in the presence.

The author Tim Palmer walks us through the strategies to deal with doubts in weather modelling, quantum physics, and systems analyses. These branches of sciences translate uncertainties to probabilities of whether certain events would happen or not.

Sources of uncertainties

The first source of uncertainties stems from dealing with a chaotic system, a system highly sensitive to the initial condition and we have huge difficulties in precisely monitor such initial condition and are unable to predict the outcome due to incomplete information.

This is a common theme in weather forecast. In the pre-computation era, people tried to monitor the weather better by increasing the number of weather stations over an area, mapping initial weather condition to records that show similar patterns to predict how the weather would progress.

In the computer-aided weather forecast, scientists have proposed a physical based model for computer simulations. To capture the wind and rain patterns of the atmosphere, one have to build a 3-D model based on gridboxes, and we assume homogeneity within a gridbox. It is these small-scale errors from modelling each gridbox that generate a big error in the final model. To increase the precision of the weather forecast, rather than increasing the resolution of the gridbox that lead to a diminishing return, MCMC simulation was introduced. In such ensemble model, one rerun the simulation dozens of times with slightly different parameters and initial input. And the simulations collectively return the probabilities of an event happenstance.

The second source of uncertainties are the quantum uncertainties. It is from the famous Heisenberg’s uncertainty principle that one cannot determine (measure?) precisely both the position and the momentum of a particle. To address how the uncertainty arises, the author discuss deeply about the hidden-variable theories on the phenomenon of entanglement.

The third source of uncertainties comes from instable systems, a system that will act in a chaotic manner with certain parameters. One example the author provide is that population’s Rmax that describe a maximum growth rate a population could achieve before forming a chaotic cycle that inevitably lead a an extinction. Another example is about modelling the stock markets with individual agents acting not too rationally.

Discoveries along the chaos

The author believe that chaos actually underpin the behaviors of dynamic systems. Through examining various complex systems, the author touches on the properties of time irreversibility, nonlinear systems, and the underlining math of fractals and p-adic number system. These examples also point out the importance of stochastic modeling (spiking in randomness to the model) could amplify the signals and help us to see distinguish the pattern from noises.

Problems that chaotic theories tackle

One of the pressing questions on climate modelling is the consequences of climate change. The authors went through how the ensemble models predict and give us a informed answers of the likely trajectory of how many degrees warmer the earth will get, laying out clearly all the uncertainties and probabilities. The author also demonstrate that ensemble model prediction can be factored in cost-benefit analyses that better inform people on investing on disaster (storm) prevention and aid distributions. Such warning-impact matrix are now routine in weather forecasts for rain, hill fire, and cyclones.

Another interesting example is using chaos theories to predict locations that are prone to wars and conflicts. On top of prediction models for local conflicts, one can also use ensemble models to estimate the displacement of people upon regional unrest. The predicted migration routes from the network model were then be used for refugee supply allocations.

The authors also apply chaos theories and quantum entanglement on neuroscience to explain how our brain work in such energy efficient mode and the possible cause of the eureka moment.

Conclusion

I think this book has so much to offer, densely packed knowledge presented in such a friendly way. I am so happy to learn more about the history of climate science, quantum physics and see how these seemingly hard subjects tie in to our everyday life!

It is definitely useful to gain a better understanding on how we can use these theories to assess our own uncertainties, converting an illegible concept of doubts to probabilities and actionable plans. It is also humbling to understand our limits on the uncertainties that is uncomputable and uncountable, and live with that.