Why "You Get Out What You Put In" Is a Lie

Every coach and teacher who ever told you that effort and reward scale together was wrong, and it wasn't a small error. It was a category mistake that leaves you unprepared for how the world actually distributes outcomes — in business, fame, science, war, almost everywhere that matters.

The truth is that returns for performance are superlinear. Being twice as good doesn't get you twice the reward. It can get you ten times the reward, or zero. A product that's half as good as a competitor's doesn't win half the customers — it often wins none, and the company dies. This isn't a flaw in capitalism that better rules could fix. It shows up in fame, military conquest, scientific discovery, even historical influence on human welfare. Wherever you look, the rich get richer, and the gap between "good" and "best" is not proportional — it's a chasm.

Once you notice this, a lot of the world reorganizes itself in your head. There are, it turns out, only two underlying mechanisms that produce it, and almost every superlinear situation is one or both of them running underneath the surface.

The first mechanism is compounding. Some processes have the property that how well you're doing right now determines how fast you grow next — and that produces exponential curves, not straight lines. Bacterial cultures. Startups with real growth rates. Fame, where existing fans recruit you new ones. Scholarship, where the more you know, the faster you can learn more. The startling thing about this, once you sit with it, is that if reward compounds with performance over time, your payoff isn't proportional to how good you are — it's proportional to how good you are raised to the power of how long you sustain it. A small, sustained edge in performance, given enough cycles, produces an output that looks less like an advantage and more like a different universe. This is also why Y Combinator tells founders to obsess over growth rate rather than current revenue: growth rate is the dial that, turned even slightly, compounds into a wildly different company two years later. Absolute numbers lie to you early on. The rate of change is the real signal.

The second mechanism is thresholds. A sports match doesn't reward you in proportion to how much better you played — it gives the win to one side, full stop, whether the margin was one point or fifty. Theorems get proven or they don't. Products cross a usability bar or they get ignored. Thresholds turn continuous effort into discrete outcomes, and that discreteness is where a lot of the world's apparent unfairness actually lives. It has nothing to do with rigged competition; you can hit a threshold entirely alone, with no rival in sight, and still get a step-function payoff instead of a proportional one.

Here's the part worth sitting with: these two mechanisms tend to travel together and reinforce each other. Crossing a threshold often triggers compounding — the side that wins a battle usually takes fewer losses, which makes the next battle easier to win. And compounding often lets you reach a threshold you couldn't otherwise touch — a startup growing fast enough can lock out competitors before they even organize. Fame works this way too: it compounds because existing attention recruits more attention, but it stays scarce because there's only so much room on the mental A-list people carry around. Once you can see both mechanisms operating at once in a single situation, you understand why some gaps in outcome look wildly disproportionate to the underlying gap in effort or talent.

There's a historical reason this feels so unintuitive: humans have almost no inherited customs for reasoning about exponential growth, because for most of history there was barely any of it to reason about. Herds couldn't grow exponentially — grazing land capped them. The only real preindustrial examples were empires (too rare and remote to shape common wisdom) and scholarship (too abstract to feel consequential). It's only in the last couple of centuries that ideas started compounding fast enough, and colliding with enough thresholds, to reorganize the world in a human lifetime. We are running ancient linear-effort intuitions on a superlinear operating system, which is exactly why "you get out what you put in" still gets repeated in classrooms full of kids who are about to enter an economy where it's false.

If you're trying to actually use this rather than just admire it, the operating rule is: seek work that compounds, and treat learning as a form of compounding even when your immediate results look flat. This is the real logic behind Silicon Valley's tolerance for failure — it isn't sentimental, it's mathematical. A founder whose company stalls but who is visibly learning fast is still on an exponential curve, just one whose payoff hasn't crossed its threshold yet. The heuristic that follows is simple to state and hard to obey: always be learning, because if you're not, you're probably standing on a line, not a curve — and lines, in the world as it actually works, lose to curves every time.

Distilled from Paul Graham Essays