Thread by @DellAnnaLuca

EXPOSURE DYNAMICS:
why people take risks and how they adjust risk-taking as their circumstances change
(thread)

1/ People take risks in an activity to grab rewards to use to mitigate risks outside that activity.

For example…

2/ …in normal circumstances we might think that the right speed to commute is 40mph, but the day we are late to the meeting we might decide that driving at 50mph is better even if it increases our chances of incident above what was considerate appropriate the day before

I.e, …

3/ we drive faster taking risks (inside the activity of driving) in order to increase a payoff (in this case, available time) to manage risks outside that activity (in this case, the risk of being fired for being late to the meeting).

4/ For this reason, the appropriateness of risk-taking can never be evaluated considering only the risks inherent to the activity at hand.

5/ The consequence of the phenomenon described above is risk homeostasis: often, when given the choice between changing an activity they were already performing and making it a little safer or a little more effective, people choose the latter, keeping the risk level constant

6/ The point above is not absolute, of course: participants adjust their risk-taking to minimize their total risk exposure across all risks *as they subjectively perceive them*.

7/ To estimate risk, people mostly use damage and pain.

(Pain is a signal of vulnerability, i.e. of risk exposure; more info in my post luca-dellanna.com/pain-signal-vu…)

In some circumstances, damage vicariously experienced through emotions is also used (eg, seeing an injured friend)

8/ All the above are lagging indicators, so risk-taking is mostly adjusted reactively.

(we can imagine risk, but unless we experienced its consequences at least emotionally, we rarely do adapt behavior to imagined risk)

9/ With some approximations, the risk model in our mind adapts to the environment like an antifragile entity comprising a multitude of mental patterns subject to stressors whenever an external event causes pain or a deep emotion.

10/ A phenomenon emerging from the above:

People use the frequency of the more common low-magnitude stressors to estimate the frequency of the rarer high-magnitude ones.

For ex., we often use the number of scares (eg emergency braking) to estimate the risk of driving on a road

11/ This has an undesirable consequence in fat-tailed activities - the Fence Paradox:

Protecting adaptive entities from low-magnitude stressors or shielding them from their consequences causes them to expose themselves to fat-tailed risk, decreasing their overall survival chance

12/ For the same reason, suppressing pain from stressors causes an entity to take unnecessary risks and makes it more fragile, thereby decreasing its overall survival rate.

Unless pain suppression gives advantage to manage *other* risks, such as being able to flee while injured.

13/ The contents of this thread were first published (with more detail) in the first section of my paper "The Dynamics of Risk-Taking"

luca-dellanna.com/the-dynamics-o…

Tomorrow, a thread on the fifth section: "Ergodicity and how to reap the gains of antifragility"