The Function You're Already Optimizing

Posted on Edited on In

Most people aren’t struggling with effort. They’re optimizing an objective function they never consciously chose.

The Puzzle

Some people execute well and still go nowhere. Others move slowly and accumulate over time. Execution quality alone doesn’t explain trajectory.

The common explanations fall apart. The hard worker often has better techniques, puts in more hours, may even have more resources. When direction is wrong, better execution only accelerates deviation.

I kept seeing this pattern around me: people grinding hard, doing everything “right,” and still ending up confused about why nothing was adding up. The real issue is optimization direction, not execution ability.

Life as Optimization

This article frames life as an optimization system. Not because that’s the only valid lens, but because it’s the most operable one for decision-making.

Whether you write it down or not, your decisions are optimizing something. You can model it as an objective function: f(x) = what you’re maximizing.

The simplest version is single-variable: maximize salary, maximize GPA, maximize interview pass rate. Simple, measurable, short-sighted.

Reality requires multiple variables: career decisions balance salary + growth + location + work intensity + visa + risk. Trade-offs emerge. Most people handle this implicitly through gut feeling rather than writing their function down.

The real world requires conditional functions:

1
2
3
4
5
6
7
def f(context):
    if survival_pressure:
        maximize(cash_flow)
    elif growth_phase:
        maximize(growth_rate)
    elif stable_phase:
        maximize(risk_adjusted_return)

Mature decision-makers don’t have one objective function. They have state-switching functions. This framing is useful because it makes decision direction explicit and debatable. You can’t debate what you can’t see.

Unconscious Optimization

Most people never write their objective function. They optimize inherited defaults from culture, parents, social expectations.

Student phase: optimize test scores. Work phase: optimize title. Anxiety phase: optimize security. These aren’t designed functions. They’re defaults.

Default functions share traits: unstable, inconsistent, non-compounding. You switch between them while believing you’re doing long-term planning.

The first step is making the function you’re already optimizing visible. Not writing the perfect function. Just seeing the current one.

Diagnostic Signals

These signals are runtime logs, not predictions. By the time they show up, the system has usually been drifting for a while.

Trajectory discontinuity. You work hard each phase, but your trajectory doesn’t connect. Three years optimizing for one thing, then switching to something unrelated, then switching again. You think you’re accumulating, but you’re changing games. The accumulation is an illusion created by effort without direction.

Non-compounding effort. Consistent effort, but results don’t build on each other. Each achievement stands alone. You’re not building on previous work. You’re optimizing locally but not globally.

Narrowing optionality. You’re achieving things, but your options are shrinking. You’re getting better at one narrow path, but that path is becoming less valuable externally. You’re optimizing for depth in a domain that’s losing relevance.

Function failure comes in three types: wrong from the start, expired (was right but context changed), or conflicting (multiple functions competing). The signals don’t distinguish these. They just tell you something is misaligned.

When Methods Stop Working

Methods don’t fail on their own. They fail when the function they were designed for is no longer active.

Most people don’t notice when their function has shifted. They keep using methods optimized for the old function.

Learning: If your function is “pass the exam,” optimal strategy is practice problems. If it shifts to “long-term understanding,” practice problems drop in weight. Systematic learning rises. The methods didn’t become wrong. They’re tuned for a different function.

Investment: If your function is “short-term returns,” you trade frequently. If it shifts to “30-year survival + compounding,” you move to index funds, diversification, risk control. Most trading techniques become irrelevant.

Career: If your function is “maximize title,” you take path A. If it becomes “transferable skills + global opportunity,” path B. Neither is wrong. They optimize different things.

The real trap is that people don’t realize their function changed, so they keep running methods optimized for a function that no longer applies. When you rewrite your function, accept that most previous optimizations become suboptimal. That’s evolution cost, not failure.

When to Rewrite

Objective functions need stability, but must be rewritable when proven wrong. Not frequent changes. Evidence-based rewriting.

Disconnected long-term trajectory. You’re changing games every few years but thinking you’re accumulating. Your trajectory doesn’t connect. Your function isn’t aligned with your actual goals.

Local optimum without amplification. You’re getting stronger in your domain, but external value is decreasing. Your function may be too narrow, or the domain is losing relevance.

Sudden risk exposure. Your function didn’t include a risk dimension, and now the system is stressed. Your function was incomplete.

Some functions aren’t wrong. They’re expired. “I used to be successful” doesn’t mean your function is still right. Accept the rewrite cost: most existing methods become suboptimal.

A Starting Exercise

Make your current optimization target explicit. Write it down. Then check:

  • 5-year validity? Will this still matter in five years?
  • Survival probability? Does this include risk management?
  • Compoundability? Do results build on each other?

Writing the function isn’t the goal. The ability to see it is. Most people optimize inherited functions, never designed ones. This exercise moves your function from implicit to visible. Not perfect, but visible.

Boundaries

Execution vs. function: If insufficient execution causes poor results, fix execution (discipline, systems, habits). If better execution leads you further from long-term goals, the function itself is broken. The distinction matters because the solutions are completely different.

Measurement vs. optimization: Tracking metrics doesn’t equal optimizing the function. Many people measure outputs religiously but never examine what they’re actually optimizing for. You can have a beautiful dashboard and still be optimizing the wrong thing.

“Wrong” means misaligned: Not a moral judgment. It means the current function can’t produce your expected long-term trajectory. Wrong relative to your own stated goals.

This article stays at the decision layer. If your objective function is bound to identity, rewrite cost goes up sharply. That’s a different problem.

You don’t rise to the level of your effort. You move along the gradient of your objective function. Most people never chose theirs.