longtermism still needs a living present

Longtermism asks you to take seriously people who do not exist yet. Take it seriously and the first thing you notice is that every one of them will need to drink.

If you arrived here through Will MacAskill's What We Owe the Future, Toby Ord's The Precipice, or a donor-advised fund your employer matched, you already hold the view. This post does not argue with it. It points at the one term in the calculation that longtermism has treated as a constant, and asks what happens when it isn't.

longtermism, in its own words

Longtermism is the view that positively influencing the long-term future is a key moral priority of our time. That is MacAskill's sentence, and the argument under it has three steps he states plainly: future people count; there could be a lot of them; we can make their lives go better. If humanity lasts even as long as a typical mammalian species, the people not yet born outnumber everyone alive today by orders of magnitude. Distance in time, on this view, is no more a reason to discount a life than distance in space.

Ord's contribution is the ledger. The Precipice puts the total chance of an existential catastrophe this century at roughly one in six, and breaks it down: about one in ten from unaligned artificial intelligence, one in thirty from engineered pandemics, one in a thousand each from nuclear war, climate change, and other environmental damage. An existential catastrophe, in his sense, is not only extinction. It is anything that permanently destroys humanity's potential: an unrecoverable collapse, or a lock-in of values that can never be revised.

Two things follow for a donor. First, the movement's money went where the ledger pointed, toward AI alignment, biosecurity, and the institutions meant to catch a civilization-scale mistake before it lands. That is a coherent reading of the numbers, and this post will not ask you to move a dollar of it. Second, the framework has a precise vocabulary for the far future (expected value, trajectory change, lock-in, existential security) and almost none for the ground the far future stands on.

One sentence for the fiduciary who wandered in: this is the philosophy, not the portfolio. Investment long-termism, meaning patient capital and holdings that can wait decades, is a different question with its own guide: the forever asset.

the ledger is right about the risk

Say this clearly so the rest lands honestly. The climate line in Ord's table is one in a thousand, and the environmental-damage line is the same. Those are not dismissals. They are subjective credences that a warming planet or a thinning biosphere is unlikely to end humanity outright this century — contested, and not a dispute this post needs to settle. Humans are adaptable. A degraded world is survivable.

So the argument here does not run through extinction. If you want a nature-as-existential-risk case, you will not find it in this post, and you should be suspicious of anyone who offers you one with confidence.

The argument runs through the other term.

what the ledger multiplies

The expected value of the long-term future is, at its crudest, a probability times a quantity: the chance we make it, multiplied by how good it is if we do. Longtermist giving has concentrated almost entirely on the first term. The second term is where the living world sits, and it is not a constant.

Every one of MacAskill's three premises assumes a physical condition it never names.

what longtermism says	what it assumes underneath
future people count	there is a habitable world for them to count in
there could be a lot of them	that world can feed, water, and cool a lot of them
we can make their lives go better	the systems that make lives go well (rain, soil, forage, fisheries, a stable climate) are still running

None of those are given. The Global Commission on the Economics of Water reported in 2024 that the global water cycle is out of balance for the first time in human history, and that more than half the world's food production is at risk by 2050 on current trends. About a fifth of the planet's inland wetlands have been lost since 1700 (Fluet-Chouinard et al., Nature, 2023). Topsoil is built over centuries and lost in decades. None of this ends the species. All of it lowers the ceiling on what the species can be.

A future counted in trillions of lives, the scale longtermists are comfortable reasoning about, is not neutral about hydrology. It is downstream of it. You can build a very large civilization on a broken water cycle. You cannot build a flourishing one.

irreversibility is a longtermist word

Here is where the two vocabularies actually meet. Longtermism's sharpest concept is not the size of the future; it is lock-in, the idea that some choices, once made, cannot be revised by anyone who comes after, and so deserve a weight out of proportion to their near-term cost. MacAskill spends chapters on value lock-in. Ord's whole definition of existential catastrophe turns on the word unrecoverable.

Ecology has been living inside that word for a long time.

A species extinction is the paradigm case of an unrecoverable event, and the current rate is estimated at tens to hundreds of times the background rate of the last ten million years (IPBES, 2019). That is lock-in in Ord's strict sense: once the last individual is gone, no later generation can revise it.

A longtermist will say a millennium is short — peat can re-accumulate, an aquifer can refill, a forest can grow back. Fair. The lock-in that matches unrecoverable is extinction. The lock-in that matches a donor's century is path-dependence: peat oxidized this decade is gone on every human planning horizon, and the species that depended on it may already be gone. Those are not the same claim. Both deserve money. Only extinction is a lock-in in the strict sense.

Two places we already write about make the point concrete. The Okefenokee is a blackwater swamp on the Georgia–Florida line that has spent several thousand years building peat up to fifteen feet deep, and it feeds the Suwannee and St. Marys rivers from that store. A mine proposed on its eastern edge would have changed how the swamp holds water. In June 2025 The Conservation Fund bought that land and its mineral rights, ending that specific threat. Nothing about that story was an existential risk. Everything about it was path-dependence, caught in time. The purchase is also what a land trust and a conservation buyer already know how to do; ensurance does not replace that closing. It keeps the condition and the routing visible after the deed is recorded. Inland wetlands more broadly (marshes, bogs, fens, floodplains) are a named living stock on our own map precisely because the drained ones do not come back.

A longtermist who takes lock-in seriously should notice that the living world is full of it, and that almost none of the movement's money is pointed there.

what a longtermist would do with new money

Hold both. AI alignment and biosecurity are real; keep funding them. Then look at the quality term, and make three moves the framework already knows how to make.

Name the dependency. Every cause on the list (the lab, the clinic, the model) runs on water, soil, and a climate that still works. Write that down as a premise, not a footnote. Effective altruism forgot the factory walks the reclassification in EA's own terms.

Fund irreversibility first. Prioritize the living systems whose loss cannot be undone on any human timescale: peat, old forest, aquifers, species. That is the same logic that makes lock-in a priority for the far future, applied to the ground.

Hold a place, not a theme. A donor-advised fund is a parking lot. A living system is a place. Pick a named stock with a measurable condition (a basin, a wetland, a stretch of river) rather than a cause label, and hold it long enough for the condition to change.

The usual EA reply is that nature fails on tractability and neglectedness, not importance. Importance alone does not move the dollar. Tractability: named places already have stewards, deeds, and annual monitoring — the next post is the checklist, with no invented DALY. Neglectedness: the physical substrate is underfunded relative to its role as the factory every other cause runs on, which is a different claim than "conservation has no money."

That last move is where ensurance enters, and only here. A grant from a donor-advised fund can fund a named place; the recipient — a land trust, a fiscal sponsor, the place's own account — keeps exclusive legal control. The certificate is a record tied to that place, held by the recipient. It is not consideration, not a return benefit, and not contingent on the grant. A personal hold, if you want one, lives in a taxable account, not the DAF. Ensurance funds the living thing. It is not the living thing, and the people who steward the place govern it. We are early and say so: taking a grant directly from a DAF sponsor is a design target, not a shipped button, and today that gift usually routes through a land trust or fiscal-sponsor cousin. None of this is tax, legal, or investment advice.

If you have a place in mind, start a giving conversation. The next post looks at where the fortune itself came from — the basin a model is trained on — and why the people who built the lab are unusually well placed to see the dependency.

frequently asked questions

what is longtermism?

Longtermism is the view that positively influencing the long-term future is a key moral priority of our time. It rests on three claims: future people count morally, there could be an enormous number of them, and actions today can make their lives better or worse. Associated with the philosophers William MacAskill (What We Owe the Future, 2022) and Toby Ord (The Precipice, 2020), it motivates much effective-altruist funding of existential-risk reduction: AI safety, biosecurity, and nuclear risk.

is longtermism the same as long-term investing?

No. Longtermism is a moral philosophy about how much weight future people deserve. Long-term investing is a portfolio discipline about patient capital and holdings that can wait decades for a return. They share a time-word. For the investing question, read long-term investing still needs a living present. The duration / forever-asset math is still in the forever asset.

why would a longtermist fund nature?

Not primarily as an existential risk; Ord estimates climate and environmental damage at about one in a thousand each, and those estimates are probably right. A longtermist would fund nature because the value of the far future is survival multiplied by how good the future is, and the second term depends on water, soil, and a climate that still work. Extinction is irreversible in Ord's strict sense. Peat loss and depleted aquifers are path-dependent on a human timescale: cheap to prevent now, effectively impossible to undo later.

also in this series

The investing leftover of the same spine — same time-word, different job.

longtermism still needs a living present — you are here (the moral argument)
long-term investing still needs a living present
short-termism is why long money still acts short
future generations inherit a living system or they inherit a story
a long-horizon book still needs a living clock

The patient-capital math lives in who can hold nature for decades.

the series

Twelve posts for people about to have more money than plan, most of it headed for a donor-advised fund. Read in order or jump.