William MacAskill is an Oxford philosopher who wants you to give everything away. Not metaphorically—literally. Ten percent of your income, minimum, redirected to causes his framework identifies as maximally effective. Your career, ideally, subordinated to whichever profession generates the most donatable income or the most direct impact on ranked global problems. Your moral attention, permanently reoriented away from the particular people and projects that constitute your actual life, toward an impartial ledger of aggregate welfare spanning all sentient beings across all time.
This is Effective Altruism. It has the backing of some of the wealthiest people in the world, the credibility of Oxford and Harvard affiliations, and the aesthetic of rigorous empiricism. It presents itself as ethics finally made serious—no more fuzzy intuitions, no more feel-good giving, just evidence, calculation, and maximum impact.
The attraction is real. Take it seriously for a moment: most charitable giving is sentimental, inefficient, and self-regarding. EA identified that problem before most people had named it. If you’re going to give, why not give where it actually works?
That question deserves an honest answer. Here it is: because the framework built to answer it contains a premise so fundamental, and so wrong, that everything constructed on top of it inherits the error. EA doesn’t need better models or more rigorous cause-ranking. It needs a different starting point. The one it has doesn’t just produce bad outputs. It produces a specific kind of human diminishment—systematically, by design, dressed in the language of virtue.
The Framework
Utilitarianism, EA’s ethical foundation, holds that the right action is whatever maximizes aggregate welfare across all affected parties. Every sentient being counts equally. Impartiality is not merely preferred—it is the supreme moral virtue. Your family, your friends, your community, your own projects and pleasures: these carry no privileged moral weight. They are, in the framework’s terms, biases. Partiality toward people you actually know and love is the moral equivalent of racism—favoring one group arbitrarily over others equally deserving.
This isn’t a caricature. Peter Singer, EA’s philosophical godfather, has argued explicitly that spending money on luxuries while children die of preventable disease is morally equivalent to walking past a drowning child because helping would ruin your shoes. MacAskill’s work extends that logic into career choice, cause prioritization, and long-term resource allocation. The utilitarian premise is not incidental to EA. It is EA. Strip it out and nothing remains but a collection of spreadsheets without a reason to open them.
Stated plainly, the premise is this: your life is not yours to direct. It is a resource to be allocated. The question isn’t what you want to build, whom you love, what work calls to you—the question is where your inputs generate maximum aggregate output. You are a variable in someone else’s equation.
The Mortgage
Ayn Rand identified this structure with precision that still cuts. Writing in 1957, decades before EA had a name or a giving pledge, she described altruism’s core demand: that need functions as “the first mortgage on your life and the moral purpose of your existence.” Not a claim on your surplus. A mortgage—prior, structural, non-negotiable. You service it before you live.
That is exactly what EA asks. The Giving What We Can pledge doesn’t request whatever you can spare after living well. It establishes an obligation that precedes your choices. Your earning capacity exists, in this framework, primarily as a mechanism for servicing the aggregate. What remains after the mortgage is yours to spend on the bias of your own existence.
Rand called this treating man as “a sacrificial animal.” EA calls it earning to give. The vocabulary has been updated. The structure hasn’t moved.
Sam Bankman-Fried is the most instructive case—not for the reasons usually cited. The standard reading is that he was a hypocrite who used EA’s language to cover fraud. The more disturbing reading is that he wasn’t. The logic of expected value, applied without remainder, genuinely licenses high-variance bets on aggregate outcomes. If the math says risking everything for a probability-weighted gain serves the ledger, the framework has no governor to stop you. Individual moral integrity—honoring your obligations to specific people in the present—is precisely the “bias” utilitarian impartiality is designed to override. SBF didn’t corrupt EA. He ran it to its conclusion.
The Disguise
The machinery of rigor EA deploys—QALYs, cost-effectiveness analyses, cause-prioritization rankings—isn’t neutral scaffolding. It’s the disguise.
The utilitarian calculus requires that human experience be fungible: your suffering and my satisfaction entered on the same ledger and netted out. They can’t be. Pain is not outweighed by pleasure when the pain belongs to one person and the pleasure to another. The measurement problem isn’t technical. It’s a category error dressed as arithmetic.
EA extends this across species, across centuries, across hypothetical future populations numbering in the trillions. Tiny probability adjustments applied to cosmic-scale outcomes produce numbers that dwarf any present concern. The math says: a marginal reduction in extinction risk outweighs any amount of immediate human suffering you can name. The framework calls this rigor. It is overconfidence with better footnotes.
The Wrong Question
Here is what EA gets wrong at the root, beneath the epistemology and the ethics both: it asks the wrong question.
“How do we maximize aggregate welfare across all sentient beings?” is not a refinement of the moral question. It’s a replacement of it—one that makes individual human life answerable to a calculation it didn’t authorize and cannot contest. The individual appears in this framework only as a container of welfare to be summed. Their choices, projects, loves, and particular existence carry no independent moral weight. They are inputs.
The older question—what does a human life require to flourish?—has answers. It requires the freedom to pursue values chosen by one’s own judgment. It requires that one’s work, relationships, and moral commitments be genuinely one’s own, not conscripted into service of an aggregate the individual will never see or verify. It requires that the self not be treated as a rounding error in someone else’s model of the good.
That question EA cannot ask. Its framework dissolves the questioner before the question can be formed.
EA set out to do the most good across all beings and all time. The ambition was genuine. The error was structural.
By making the individual a variable in the aggregate—by treating impartiality as the highest virtue and the self as the primary bias to overcome—EA didn’t transcend selfishness. It just fancifully demands self-sacrifice to their models. It isn’t a philosophy for doing good. It’s a mortgage, dressed in a spreadsheet, calling itself liberation.
The giving pledge isn’t generosity. It’s conscription with good branding.
