A Critique of Effective Altruism

Framing: Written as a deliberate hypothetical apostasy (Bostrom’s exercise) by a committed effective altruist “in character.” Kuhn stresses the critical tone doesn’t reflect his actual stance — he isn’t abandoning EA — and warns EA readers that the piece may trigger defensive rationalization.

Central thesis

• EA is arguably the first broadly appealing movement built on a genuinely useful ethical toolkit (its principles: altruism, maximization, egalitarianism, consequentialism), which imply improvements over the social default (earning to give, working in the developing world, using evidence/feedback).
• But as a movement it fails to use those principles to reach correct, nontrivial beliefs about how to improve the world.
• Key distinction: “actually trying” (find the utility-maximizing action, then do it) vs. “pretending to try” (query your brain for something acceptable and roll with it). Kuhn summarizes much of EA as “pretending to actually try”: EAs noticed the pretending/actually-trying distinction but wrongly assumed knowing it confers the ability to actually try. In practice they still satisfice — pick options that don’t obviously violate EA principles, then choose among them on essentially random preferences — adding “epistemic inertia” that makes the consensus slow to update.

The specific critiques

• Philosophical difficulties: EAs deeply disagree on population ethics, which is crucial for prioritizing causes, yet little progress is made toward convergence — risking fragmentation.
• Poor cause choices: GiveWell’s top charities are all global health, not because it’s proven highest-leverage but because it had the most data. Heuristic arguments suggest x-risk, animal advocacy, or open borders could be more cost-effective. Also, since the “best known intervention” keeps improving, donating now (rather than using a donor-advised fund) seems perverse.
• Non-obviousness: EAs are smug rather than worried that such a simple idea arose so late. In a roughly efficient idea-market, that lateness should prompt them to ask what they’re missing.
• Efficient markets for giving: the “nonprofits are a market for warm fuzzies” line ignores effectiveness-minded money like the Gates Foundation ($60bn, 6M+ lives saved). EAs need to articulate what “edge” they have that Gates lacks — “if you don’t know what your edge is, you’re the sucker.”
• Inconsistent attitude toward rigor: extreme rigor on charity recommendations, but gut-feeling decisions on ancillary questions (give now vs. later, career choice, how “meta” to go).
• Poor psychological understanding: deep conflicts between EA and other goals get resolved via motivated reasoning (“it’s my comparative advantage to spread EA through poetry!”) or the easiest “honest” out. EA norms don’t proscribe this, so people don’t notice their own motivated cognition. Quotes John Sturm on the “commitment level” issue (sometimes people aren’t lazy EAs — they’re just not trying at all in some domains). EAs also show surprisingly little interest in self-modifying toward more altruistic intentions.
• Historical analogues: almost no research into outside views. The most similar prior movement is communism — also started by philosophers/technocrats who grasped the “actually trying” distinction (Aaron Tucker’s extended quote on Marx) and still failed abjectly. EA needs more awareness of the pitfalls of being a large movement.
• Monoculture: EAs are demographically narrow (white, upper-middle-class, developed-country, philosophically inclined, likely male-skewed) and humanistically unaware (from analytic philosophy, quick to dismiss critiques like cultural imperialism as postmodern fluff) — a problem given global-health beneficiaries are culturally very foreign.
• Community problems: discussion venues (e.g. the Facebook group) are being flooded by newcomers who haven’t absorbed the norms/standards, degrading the group truth-seeking process — and few seem aware of the severity.
• Movement-building issues: large movements and truth-seeking rarely coexist; growth brings epistemic inertia as social-pressure-driven members (slow to update) reach critical mass.

Conclusion

• Three recurring failure modes: (1) EAs stop thinking too early / satisfice; (2) EA norms don’t guard against motivated cognition under EA’s heavy demands; (3) EAs ignore “being a successful movement” questions (group epistemology, historical precedent, diversity).
• The deeper problem is lack of awareness of these issues. That no insider had done a substantial critique yet is “a huge red flag” — suggesting many EA beliefs rest on social rather than truth-seeking behavior. Without fixing this, the movement adds only epistemic inertia over what individuals could achieve alone.

Proposed fix

• Issue #3 (self-awareness as a movement) subsumes the others. Kuhn proposes adding a fifth principle beyond altruistic/maximizing/egalitarian/consequentialist: self-aware — deliberately thinking through the challenges of being a successful movement so EA can move past obvious applications toward nontrivially better ways to do good.
• Credits Nick Bostrom (hypothetical apostasy) and contributors including Michael Vassar, Aaron Tucker, Andrew Rettek, John Sturm.

Notable comment (Vaniver)

• Calls it a “we need two Stalins!” apostasy — the self-serving kind that “lands you a cushy professorship” (is “be more introspective” conveniently your comparative advantage?). Suggests a sharper prompt: imagine that in ten years you view your EA advocacy as a huge waste — why? — and works through how each of the four principles might individually fail.