Priorities and Salience

How to pass an hour and spend a decade

Imagine being handed an hour, with nothing else specified — just: go and do something about a problem that matters. Most people would reach for whatever feels most pressing that day. There's no time to build a framework; instinct has to do the work.

Now imagine being handed eighty thousand hours instead — roughly a working life — or two hundred billion dollars. The scale changes the question. An hour rewards whatever is closest to hand. A working life, or a fortune, rewards patience: time enough to ask which problems are getting worse quietly while everyone watches the loud ones, and which have a door that's closing rather than one that stays open, however expensive, indefinitely.

Both are versions of the same question — how effort, at any scale, should get prioritised among the problems worth working on. What looks obviously right with an hour to spend can look obviously wrong with a lifetime, and vice versa.

An earlier essay in this series named three such problems — conservation, world peace, and fulfilment — and stated plainly that they "aren't ranked." That's still true, as a claim about categories. What it doesn't answer, and was never quite trying to, is the harder question above: given a specific amount of time or money, right now, where does it actually go.

Three different questions dressed up as one

Say someone asks you to help fix something. Before you can decide whether to help, you're actually asking three separate questions, whether you notice it or not.

How hard is this to fix? That's effort. Some problems take an afternoon; some take a decade and a research budget you don't have.

What happens if nobody fixes it? That's a different question entirely, and it's the one people conflate with severity most often. A leaking tap and a cracked foundation can look similarly bad on the day you notice them. They are not similarly bad if left alone for five years.

Is anyone actually working on it? This is the one most people skip past, because it feels like a footnote rather than a real variable. But a problem that's severe and hard and completely unattended is a different animal from a problem that's severe and hard and already has a hundred well-funded teams on it. The first is where your hour or dollar does something. The second is where it probably doesn't.

Most conversations about priorities smash these three into a single word — "important" — and then argue about which problem is more important, as if that settles anything. It doesn't, because "important" is silently answering all three questions at once, often inconsistently, and nobody notices which one is actually carrying the argument.

Each of these deserves its own name. Once they can be told apart, a lot of confused priority-setting starts to resolve itself.

Effort, cost, and the question nobody asks cleanly

Call the first one effort to solve. Straightforward: what does it actually take — time, money, coordination, technology — to fix this.

Call the second one cost if unsolved. What accrues, quietly or not, while the problem sits there. This is the one worth pausing on, because it hides a distinction that changes everything downstream.

Most costs-if-unsolved are large but recoverable. A war is horrific and can go on for years, but history is, grimly, full of examples of societies that rebuilt afterward — sometimes badly, sometimes remarkably well, but the option to rebuild existed. An unfair institution can persist for generations and still, eventually, be reformed. The cost of leaving these unsolved is severe. It is not infinite, in a specific technical sense: the door to fixing it later stays open, even if it's expensive to walk through.

Some costs-if-unsolved are not like this. Past a certain point, a species doesn't come back. An ice sheet that's gone is gone. A tipping point, once crossed, doesn't offer a return trip. These aren't "very expensive to reverse" — they're not reversible at all; the option to solve the problem later simply stops existing. That's the honest meaning of "irreversible," and it's narrower than people usually intend. Most things called irreversible are just expensive and slow to reverse. The genuinely irreversible ones are rarer, and different in kind, not just in degree.

Here's why the distinction matters practically: if the cost of leaving something unsolved is heading toward "the option disappears forever," the value of acting now goes up sharply, no matter how expensive acting now is. Roughly: the value of solving something soon is what you avoid losing, minus what it costs to do it. Ordinarily that's a normal trade-off. But when what you'd lose is heading toward "everything, permanently," the trade-off stops being normal. Almost any effort becomes worth it, because the alternative isn't expensive — it's over.

This is not a reason to declare conservation the winner of a three-way contest against peace and fulfilment. It's a reason to say: whichever problem currently has cases nearing that kind of threshold deserves the marginal hour, right now — a moving target, not a permanent scoreboard. War can, in principle, produce its own point of no return — escalation past the level where anyone can walk it back. Entrenched systems of suffering can become self-perpetuating in ways that are functionally, if not literally, closed doors. Those thresholds haven't been mapped as carefully for peace and fulfilment as they have for the natural world, where scientists have spent decades identifying exactly where the tipping points sit. That's a gap worth naming, not a reason to assume conservation always wins.

The variable everyone forgets: is anyone looking?

Effort and cost-if-unsolved are both properties of the problem itself. The third variable is different — it's a property of us, of where our collective attention happens to be pointed.

Salience is whether a problem is getting attention in proportion to how much it actually matters. A lot of very real problems have low salience not because they're unimportant, but because they're boring, or slow, or don't produce a compelling image, or are hard to explain in one sentence. Soil degradation doesn't have a photogenic disaster moment the way a hurricane does. It's arguably at least as consequential. It gets a fraction of the attention.

This creates a specific and somewhat counterintuitive opportunity: a problem with low salience and real severity is often where a given hour or dollar goes furthest, precisely because almost nobody else is looking. It's the philanthropic and personal-effort equivalent of an unfashionable stock — the value is real, it's just not priced in, because attention hasn't found it yet.

But salience and cost-if-unsolved are not the same axis, and mixing them up causes real confusion. A problem can be highly salient (everyone's talking about it) while being nowhere near any point of no return. A problem can be almost invisible while quietly approaching one. Salience tells you what people are looking at. Cost-if-unsolved tells you what happens if they keep not solving it. These can point in completely different directions, and a good priority-setting instinct needs both readings, not just one.

Where the "not ranked" position actually lands

The three problems still aren't ranked as categories — nothing here concludes that conservation simply outranks peace and fulfilment forever, or vice versa. They compound each other, and treating one as generally more important tends to produce blind spots rather than clarity.

What "not ranked" can't do is answer the hour-versus-lifetime question posed at the start. The more honest position: priority is dynamic, set by proximity to a closing-door threshold, wherever one currently exists. Right now the clearest documented cases sit in conservation, largely because tipping points there have been mapped more precisely than anywhere else — a fact about where current knowledge is sharpest, not a permanent verdict. If peace or fulfilment turn out to have comparably sharp thresholds, priority shifts accordingly, without revising the underlying principle at all.

One piece remains genuinely unresolved: effort to solve and cost if unsolved may be more entangled than presented here — possibly two views of the same curve rather than separate ones. Treating them as distinct is the more cautious assumption, held loosely, pending real cases to test it against.

What this changes, and at what scale

Not much about the big picture. All three problems remain worth working on; none can be solved in isolation from the others; the status cycle running underneath all three is real, regardless of which problem gets the next hour.

What changes is smaller and more portable: three separate questions instead of one blurred one, at any scale. How hard is this to fix. What happens if it isn't. Is anyone actually looking. For an hour, the honest answer is often "there's no time to find out — act on instinct, and don't pretend otherwise." For a lifetime, or two hundred billion dollars, there's no such excuse. The answer to where it should go was usually hiding in the gap between those three questions the whole time, obscured only because they'd been crammed into a single word.

If you're contemplating how to direct a career toward the problems that need it most, 80,000 Hours has spent over a decade building out a version of this same allocation question — their framework of scale, neglectedness, and solvability maps closely onto cost-if-unsolved, effort-to-solve, and salience. Find it at 80000hours.org.

This essay is part of a series on tools, cycles, and the problems worth working on.

Written with AI. Version 1 — July 2026.