Effective Altruism
Effective Altruism is the philosophical and practical movement that seeks to use evidence and careful reasoning to do the most good possible with one's resources, treating beneficence as a quantitative problem of maximisation under constraint. Its philosophical foundation is Peter Singer's essay 'Famine, Affluence, and Morality' (1972), which argued that if we can prevent serious harm at little cost to ourselves we are morally required to do so regardless of distance, and his later 'The Life You Can Save' (2009) and 'The Most Good You Can Do' (2015). The movement emerged organisationally in the late 2000s around Oxford-based philosophers including Toby Ord (founder of Giving What We Can, 2009) and William MacAskill, whose 'Doing Good Better' (2015) provided its popular manifesto, and around evaluator organisations such as GiveWell. Ord's 'The Precipice: Existential Risk and the Future of Humanity' (2020) and MacAskill's 'What We Owe the Future' (2022) extended the framework into longtermism — the view that positively influencing the long-run future is a key moral priority of our time. The movement has directed substantial philanthropic resources toward global health interventions (anti-malarial nets, deworming, direct cash transfers), animal welfare, biosecurity, and AI alignment research. The post-2022 reputational crisis surrounding the FTX founder Sam Bankman-Fried, a high-profile EA-aligned donor convicted of fraud, has provoked extensive internal reckoning over governance, epistemic culture, and the relation between maximising consequentialism and ordinary moral constraints.
Worldview
The effective altruist inhabits a world in which suffering is enormous, resources are finite, and the difference between mediocre and excellent charitable interventions can be two or three orders of magnitude in lives saved per dollar. The fundamental orientation is one of impartial, quantitative beneficence: take seriously the equal moral weight of every sentient being, and then think hard about how to convert your scarce resources into the most good. The longtermist extension adds a temporal dimension — the trillions of potential future persons whose existence depends on what humanity does in the coming centuries — and a corresponding focus on existential risks from engineered pandemics, misaligned artificial intelligence, and great-power conflict. The framework classifies this as None: effective altruism is a secular and broadly consequentialist tradition that locates moral motivation in rational concern for sentient welfare and grants no role to cosmic-ordering principles, personal deities, or spirit-relational powers. The framework reads this as Reason: moral authority rests on careful argument, empirical evidence, and expected-value calculation rather than on revelation, tradition, or unmediated experience, though the post-FTX reckoning has produced a substantial internal debate over whether such reasoning needs to be constrained by widely shared moral intuitions about honesty, integrity, and side-constraints on harming others.
Moral Implications
EA ethics is broadly impartial consequentialism: the right action maximises expected aggregate welfare, with all sentient beings counting in proportion to their capacity to experience well-being. This generates strong demands of giving (Ord and many other EAs pledge at least ten percent of lifetime income), of career choice (the 80,000 Hours organisation advises on high-impact careers), and of cause prioritisation (concentrate resources on interventions where the marginal dollar does most good). The movement has been criticised for moral overdemandingness, for naïveté about institutional and political context, for its willingness to override ordinary moral intuitions in the name of expected-value calculation, and — after the FTX collapse — for an epistemic culture that rationalised significant ethical failures by influential members.
Practical Implications
EA-aligned funding has reshaped parts of global health philanthropy (the Against Malaria Foundation, GiveDirectly), animal welfare (corporate cage-free campaigns), biosecurity policy, and AI safety research, redirecting on the order of tens of billions of dollars over the past decade. The 80,000 Hours career-advisory project has influenced a generation of graduates toward high-impact work in policy, research, and operations. The movement's analytical templates — cost-effectiveness in DALYs averted, probabilistic forecasting, explicit cause prioritisation — have diffused well beyond its self-identified membership. Its post-2022 crisis has triggered substantial governance reforms, the restructuring of leadership institutions, and a renewed argument about whether the movement can sustain ambitious longtermist projects without recreating the failure modes that discredited its most visible donor.
I. Time
Time is substantival, one-dimensional, linear, continuous, and non-deterministic. The longtermist branch of EA treats time as potentially extending across millions or billions of years of future human and post-human existence, and weights the well-being of future persons equally with that of present persons; the existential-risk literature (Ord, Nick Bostrom's 'Existential Risk Prevention as Global Priority', 2013) draws the policy implications. Time freedom is non-deterministic: humanity's long-run trajectory is open, and the present generation occupies a uniquely high-leverage 'hinge of history' in which catastrophic risks must be navigated.
Attributes
II. Space
Space is substantival, three-dimensional, flat, and local in the everyday sense, but EA's longtermist horizon takes the relevant spatial scale to be infinite — the accessible universe over astronomical futures, in which technologically mature civilisation might expand. Spatial distance carries no intrinsic moral weight: a child dying of malaria in sub-Saharan Africa has the same claim on the donor in San Francisco as a child next door. The flatness of moral space is one of the movement's most distinctive commitments.
Attributes
III. Matter
Matter is substantival, conserved, three-dimensional, and local. The longtermist literature treats the universe's available matter and energy as the resource base for possible future flourishing, occasionally invoking the astronomical scale at which intelligent action might eventually operate. In ordinary EA work the material world is taken for granted as the standard scientific physical world within which interventions are designed and evaluated.
Attributes
IV. Observer
The effective altruist is an embodied, finite, broadly impartial agent who treats her moral situation as a decision problem under uncertainty. Knowledge is mediated and partial: she relies heavily on cost-effectiveness analysis, randomised controlled trials, expert elicitation, and explicit probabilistic forecasting to identify interventions whose expected impact is large. Agency is active and quantitatively self-aware: career choice, donation strategy, and research direction are all decisions to be optimised. Plural observers populate the moral universe — every sentient being, present and future, counts equally in principle — and the species, including its possible descendants over astronomical timescales, is the relevant frame of reference. The observer is sceptical of moral intuitions when they conflict with explicit expected-value reasoning, though the post-FTX debate has pushed many in the movement to take ordinary moral constraints more seriously.
Attributes
V. Energy
Energy is substantival, finite, conserved, and irreversibly dispersible in the standard physical sense. The movement takes its physics from contemporary science without distinctive theoretical modification. In the longtermist register, the eventual heat-death of the universe and the available negentropy of the accessible cosmos figure as outer limits on the scale of possible future value, but for practical present-day work energy considerations enter chiefly through climate policy and decarbonisation as instances of existential or catastrophic risk reduction.
Attributes
VI. Information
Information is emergent and conserved at the social scale: EA institutions (GiveWell, Open Philanthropy, the Forethought Foundation, 80,000 Hours) constitute a self-conscious epistemic infrastructure dedicated to producing, refining, and disseminating evaluations of charitable interventions and global priorities. The movement places unusual weight on forecasting accuracy, calibration, and the public updating of beliefs. Personal informational conservation is denied: the individual does not survive death, but the longtermist strand of the movement places extraordinary moral weight on informational continuity at the species scale — on the preservation of the human (and possibly post-human) civilisational project across deep time.
Attributes
Works that name Effective Altruism in their embodiments
Foundational texts that draw on this school, with each work's declared weight.
How Effective Altruism resolves each dilemma
57 resolved positions across 4 dimensions, including 11 distinctive where the majority of schools go the other way.
Each dimension is sorted so minority positions come first. Mainstream positions are folded into an expandable list.
Time · 9 dilemmas · 3 distinctive
Persistence, the future, and the direction of becoming.
6 mainstream positions
Matter · 7 dilemmas · 4 distinctive
What stuff is — fundamental, relational, or appearance.
3 mainstream positions
Observer · 37 dilemmas · 4 distinctive
Mind, agency, and the knower's relation to the known.