#35: Patterning, herding, programming
Why seeing and recognizing uncertainty is so hard.
Greetings to readers old and new. This is yet another of Vaughn Tan’s weekly attempts to make sense of the state of not-knowing. The first issue explains the project; you can see all the issues here. You can now also pre-order The Uncertainty Mindset, a book about uncertainty, innovation, cutting-edge cuisine, and other good stuff—it comes out officially on July 28.
This last week, I learned that “doomscrolling” is the neologism for my nightly survey of the apparently inexhaustible sea of bad news on the internet.
On the ground, here in London, there is an increasing sense of the surreal. The vast buckets of emergency support and government stimulus poured into the economy in the last few months have kept things from falling off a cliff—for now.
This conceals, but very imperfectly, how the future is extraordinarily uncertain. The future is especially uncertain in countries like the US and UK which did not take serious enough action early enough, have not been as successful at controlling coronavirus infection, and lack a serious plan for managing spread after reopening.
Robust testing and tracing is foundational to any reopening plan because it is something like a substitute for a vaccine. Detecting, tracing, and isolating infected people aggressively keeps them away from the rest of the population so everyone can carry on with life as normally as possible.
Unless there are only a small number of new cases each day, even robust testing and tracing won’t be enough to keep infection rates down when economies reopen—and if infection rates don’t stay down, further disruption to daily life seems unavoidable. We have known this for months; now we have evidence from the delayed reopenings and the returns to lockdown in the US.
There are probably levers to pull to control coronavirus, but we don’t know how hard to pull, and on which levers. We don’t even have a very clear understanding of the factors that influence the spread of coronavirus from infected persons.
The most realistic evaluation of the situation is there is enormous uncertainty in the future in the US and UK. To the extent that the following make very little sense in the near and medium terms:
Long-term commitments to densely populated cities
Investing in business models built around people being physically close together
Expecting casual, inexpensive, convenient international travel to return quickly
Hoping that economies will not undergo profound structural adjustments
The prevailing sense of the future—for me, at least—is of speeding toward the cliff’s edge or of already being suspended above a gulf of air, possibly by a fraying thread. A nameless dread is now always in the background.
What’s surreal is that there are many other signals to indicate that other people in the US and UK don’t share this nonspecific sense of impending doom.
A friend who works at a quite unusual property agency says that home sales across the UK have surged and prices are up. US stock markets are near all-time highs—even for hospitality and travel. People in both countries seem happy to form enormous crowds both outdoors and indoors. Friends tell me about their plans to go on holiday soon. People seem to believe that they can see the future and that things will be okay.
Behaving as if the future is knowable makes people do things which will probably be maladaptive when the future turns out to be something unexpected.
Why do we find it so difficult to recognize a situation of tremendous uncertainty even when, as now, it has come up and essentially punched us in the collective face? I have three half-baked theories:
Patterning
Herding
Programming
Patterning.
Patterning takes past events and creates connections between them. Even the most basic kind of causal inference is patterning: “If [I touch that hot pan without an oven mitt] → [ [it will hurt a lot] → [my finger will blister soon after] ].” To take any intentional action at all, we need to have theories of causation—so we need to pattern, whether we do it consciously or not.
This means patterning is both inherently retrospective and a strong imperative. Pattern-creation and pattern-matching are always based on history, whether directly or vicariously experienced. We use the patterns from the past to pattern events as they happen and to anticipate the patterning of events to come.
Patterning allows us to make sense of what has happened, what is currently happening, and what will happen. The events of the past (and the patterns derived from them) contaminate how we respond in the present and what we expect of the future.
The past is much more certain than the future, so the patterning imperative tends to assume the future will be more certain than it actually is. Herding.
Being in a collective requires members to have some kind of agreement about the rules of the game—it doesn’t matter if the collective is the size of a family or the size of a nation).
These rules are often unwritten and unspoken, but they nonetheless establish how members can expect others to behave and react to each other. This is how members can act within the collective. “I’m going to buy a house that costs a bit more than I want to pay each month, because I expect to stay in my job and be promoted so my employer will pay me enough eventually for the mortgage to be comfortable.” Buying the house depends on expectations driven by rules: stay at a job long enough and get promoted; get promoted and get paid more.
Agreement about these rules is essential for complex, interdependent collectives to function—and the more complexly interdependent, the more agreement about the rules there needs to be. Agreeing on rules becomes harder when you expect the future to be unpredictably different from the past.
The larger and more complex the collective, the stronger the pressure on each members to expect that the future is certain—otherwise acting and being in the collective becomes impossible.
If most people around us believe things will continue more or less in the same way (because this is essential for people to live together, work together, and depend on each other) as individuals we are disincentivised to believe that the future will be uncertain. Programming.
Daily life habituates us to think in terms of risk (understandable, quantifiable unknowns) instead of true uncertainty. This happens through education, entertainment, and through the broader economy and polity we are exposed to.
Schools teach frequentist probability by default (Bayesian probability is for the advanced courses). Nearly all the games we play have explicit game dynamics based on risk (even if the non-explicit game dynamics may be based on uncertainty). Governments and businesses apply risk management as the default mode of action for anything from deciding where to put streetlights and police stations to how much to charge for travel insurance—and as customers of governments and businesses, our lives are shaped by this risk mindset.
The reason for this is probably quite simple: risk is much easier than uncertainty to conceptualize, model, communicate, and implement. We have a robust vocabulary for risk but not one for uncertainty. And without a vocabulary of uncertainty that is as robust as our vocabulary of risk, it may not even be possible to clearly perceive the difference between a future that is risky and one that is uncertain.
Because we are habituated to living and thinking risk (vs uncertainty), risk becomes the default and dominant lens through we see the future and act in it—even when the future is both risky and uncertain. Patterning, herding, and programming each push us away from acknowledging the uncertainty of the future.
This is problematic because defaulting to interpreting the future as risky means not knowing how to react if the future turns out to be different from what’s expected and planned for—either being paralyzed or being stuck with maladapted or counterproductive actions. (This should sound familiar.)
But the other problem with not actively developing the capacity to recognize uncertainty is that this stifles the ability to innovate. Those who are unable to interpret the future as uncertain are unable to see that the future can take unexpected shapes—and this means being much less able to imagine and create unexpected futures.
It’s just an ordinary story about the way things go.
Find me on the web at www.vaughntan.org, on Twitter @vaughn_tan, on Instagram @vaughn.tan, or by email at <uncertaintymindset@vaughntan.org>.
Two thoughts about herding.
1. To the extent that reality is what we agree on -- and when it comes to economics and behavior I believe this rounds to true -- a solid group can herd so effectively that uncertainty is little more than a low background noise.
2. There's more than one herd and people can be persuaded to move from one that might be a natural or first home for them to one that promises more. This dynamic is all over capitalism and politics.