one life to live · metaphor 9 of 100
When does the turbulence in your life help you, and when does it grind you down? Whether you should pray for calm or for chaos depends on one thing: the curvature of your exposure.
Picture two people with equally unpredictable years. The first is a salaried specialist with a mortgage, a title, and fixed obligations sized exactly to her income. For her, every surprise is downside: the good years can't pay her more than her contract says, but a reorganization, an illness, a rate hike can take almost everything. The second is an inventor with small stakes in many things — half-finished prototypes, odd collaborations, tiny bets. Most of his surprises are duds that cost him a weekend; but every scrap of upside in his life arrives as a surprise, because nothing scheduled ever made him anything.
Measure their lives with a statistician's ruler and the variance is identical — the same jitter, the same unpredictability. Yet one is being ground down by it and the other is being fed. The difference is shape: the curve that converts how a year goes into what a year pays.
All numbers computed live: E[f(x)] by numerical integration of the bell curve the years are drawn from (clipped to the visible range, with the clipped mass parked at the edges — the raindrops are clipped identically), the running average from the years that actually fall.
The gap
Everything in the instrument hangs on one distinction that ordinary planning erases. The outcome of an average year — f(E[x]) — is what you'd get if life delivered its statistical mean, once, on schedule. The average of your actual years — E[f(x)] — is what you really accumulate, jitter included. Jensen's inequality says these two numbers agree only when the curve is a straight line. Bend the curve and they split: convex bends pull the lived average above the planned one, concave bends drag it below.
A curve that opens upward gains more from a good surprise than it loses from an equally sized bad one — so shuffling outcomes around the mean nets you something. A curve that opens downward loses more than it gains — so the same shuffle bleeds you. Variance is transferred through curvature into gain or harm. This is the arithmetic behind the two lives in the opening: volatility is neither good nor bad in itself. It is a raw quantity, and it gets multiplied by shape.
What to try
Fix the curve, sweep the variance. Press Antifragile — convex, then drag turbulence from dead calm to full storm. The mean of the years never moves — only their spread — yet E[f(x)] climbs and the gap turns a deeper green. Now press Fragile — concave and repeat: same weather, and the gap plunges red. Nothing about the storm changed. Only what it was falling on.
Fix the variance, sweep the curvature. Set a healthy storm, then slide curvature slowly from left to right. Watch the gap pass through exactly zero at the straight line — the one shape for which unpredictability is genuinely irrelevant — and change sign on the other side. Under convexity, more chaos is strictly better — every extra unit of turbulence adds to the lived average. If you can find or build a position like that, the storm is on your payroll.
Reading your own curvature
You can't see your life's payoff curve, but you can probe its second derivative with a single question. For any commitment — the job, the lease, the project, the relationship pattern — ask: if this year surprises me wonderfully, what do I gain? If it surprises me terribly, what do I lose? Bounded downside and open upside — the cheap experiment that might become a vocation, the introduction that might change everything — is convex: you should welcome noise there. Open downside and bounded upside — the maxed-out lifestyle, the single client, the reputation that can only be maintained or destroyed — is concave: noise there is pure erosion, however calm things feel today.
Run the test across a whole life and a portrait emerges. Most people discover they are a bundle: convex in their hobbies, concave in their obligations — and that they've been praying for calm in exactly the places they should have been rolling more dice, and vice versa.
Buying and selling curvature
Once you see exposure as a curve, a family of familiar transactions is one transaction. Insurance is paying to remove concavity: you accept a small certain loss to chop off the open downside, straightening the left tail of your curve. Experiments, options, and tinkering are paying a little to own convexity: the ticket price is capped, the discovery is not, and each small stake is a bet that variance will do your searching for you. And overcommitment is selling convexity without noticing: every fixed cost you take on, every promise sized to your best-case year, quietly caps your upside and opens your downside — you've written someone an option and pocketed nothing for it.
These purchases matter more the wilder your weather. In a dead calm, shape is nearly free — the gap is tiny at any curvature. In a storm, shape is almost everything.
The mapping
| Mathematics | Life |
|---|---|
| x | How the year actually goes — the raw, unchosen draw of events. |
| f(x) | What your arrangements convert that year into: money, standing, options, joy. |
| curvature f″ | The asymmetry of your exposure to surprise — whether a great shock gains you more than a bad one costs. |
| variance of x | How turbulent your life is: the sheer unpredictability of the years, before any verdict on it. |
| Jensen gap | What volatility silently gives or takes — the wedge between the life you planned on and the one the jitter actually pays. |
| convex position | A life where you want more dice rolled: bounded downside, open upside, chaos on the payroll. |
Where the metaphor tears
The instrument holds one bend across the whole range; few real exposures are so consistent. Many positions are comfortably convex near home and violently concave far out — the trade that "always works" until the one year it doesn't, the lifestyle whose small shocks are absorbed easily right up to the shock that isn't small. The asymmetry test must be run in the tails, where it's hardest to imagine, not just around the mean, where it's easy.
Jensen's inequality is a statement about averages, and you are not an average — you live through every draw, in order. A convex life may pay more over twenty years while feeling like anxiety the whole time: humans metabolize the jitter itself, not just its expected value. A curve that's mathematically superior can still be one your nervous system can't afford to hold.
The framing of "buying and selling" shape suggests a market anyone can enter. But cheap experiments require slack — savings, health, safety nets, second chances — and the poor are structurally short convexity: their downsides are open, their upsides capped, through no dial of their own. Read this page first as an observation about circumstance, and only after that as advice. Telling someone to "own more optionality" without asking what it costs them is how the metaphor curdles.