# Effective Altruism A philanthropic movement, intellectual community, and increasingly a political force — effective altruism (EA) argues that charitable giving and career choices should be optimized using evidence and reason to do the most good possible. What began as a sensible idea (compare charities, fund what works) evolved into something far more ambitious and far more dangerous: a moral framework that justified massive risk-taking, diverted resources from near-term human suffering toward speculative future-oriented projects, and produced one of the largest financial frauds in history. **Core thesis for this KB:** EA is the "EA" in [[TESCREAL]] and the movement's trajectory — from evidence-based charity to longtermist AI obsession to FTX collapse — is a case study in how idealist frameworks get captured by elite interests. EA's institutional infrastructure now functions primarily as a talent pipeline and funding network for AI companies, defense tech, and the broader tech oligarchy documented in this vault. → Part of the [[TESCREAL]] bundle. See also: [[Longtermism]], [[MIRI]], [[Effective Accelerationism]] --- ## Origins: The Sensible Phase (2009–2015) ### The Oxford Genesis EA emerged from conversations between Oxford philosophers [[Toby Ord]] and [[William MacAskill]] in the late 2000s. The founding insight was genuinely powerful: most charitable giving is driven by emotion, proximity, and marketing rather than evidence of impact. If you actually measure outcomes per dollar, some interventions (distributing antimalarial bednets, deworming programs, direct cash transfers) are hundreds of times more cost-effective than others. Ord's 2009 "[[Giving What We Can]]" pledge — committing to donate at least 10% of income to the most effective charities — and MacAskill's 2011 "[[80,000 Hours]]" career advice organization (named for the approximate hours in a career) built the institutional foundation. [[Peter Singer]]'s utilitarian philosophy provided the moral framework: if you can prevent suffering without sacrificing anything of comparable moral importance, you're obligated to do it. ### [[GiveWell]] and the Evidence Phase GiveWell, founded by [[Holden Karnofsky]] and Elie Hassenfeld in 2007, became EA's analytical engine — rigorously evaluating charities and publishing transparent analyses of cost-effectiveness. This was EA at its best: unglamorous, evidence-driven, focused on measurable outcomes for the world's poorest people. The early EA movement funded concrete, provably effective interventions: the Against Malaria Foundation, the Schistosomiasis Control Initiative, GiveDirectly. These organizations saved and improved real lives with real evidence behind them. --- ## The Longtermist Turn (2015–2020) ### From Bednets to [[Existential Risk]] The pivot began when MacAskill and others argued that EA's focus on near-term, measurable outcomes was thinking too small. If you take seriously the idea that future people matter morally, and if you accept the astronomical numbers involved (potentially trillions of future humans), then the expected value of reducing existential risk dwarfs anything you can achieve with bednets. This was the [[Longtermism]] capture of EA. The movement's analytical framework — expected value calculations, cost-effectiveness estimates, utilitarian moral reasoning — was redirected from concrete present-day interventions toward speculative long-run problems, particularly AI alignment. ### The Pivot to AI By the mid-2010s, AI safety had become EA's dominant cause area. The reasoning: unaligned artificial general intelligence (AGI) could pose an existential risk; the probability of this is uncertain but not negligible; the number of future lives at stake is astronomical; therefore, even a tiny probability multiplied by astronomical stakes makes AI safety the most cost-effective intervention available. The logic is seductive, and not entirely wrong — AI risk is real. But the practical effect was to redirect EA resources away from proven interventions for the world's poorest and toward speculative research that happened to be centered in exactly the same institutions and social networks as the tech industry's most powerful players. ### [[Open Philanthropy]]: EA's Central Bank In 2017, GiveWell's Holden Karnofsky and Facebook co-founder [[Dustin Moskovitz]] spun off Open Philanthropy as an independent foundation with a mandate to go beyond GiveWell's evidence-backed charity evaluations. Open Phil became the EA movement's dominant funder — and its strategic orientation was increasingly longtermist. By 2023, Open Phil was spending approximately $46 million annually on AI safety alone, making it arguably the world's largest funder of AI safety research. The foundation's total giving exceeded $4 billion by mid-2025 across global health, AI safety, pandemic preparedness, and animal welfare. **Rebranding:** In November 2025, Open Philanthropy renamed itself **Coefficient Giving**, signaling a shift toward operating multi-donor funds that other philanthropists could join. The rebranding reflected both the organization's expansion and its desire to distance itself from EA's post-FTX reputation. --- ## The FTX Catastrophe (2022) ### [[Sam Bankman-Fried]]: EA's Philosopher-King Sam Bankman-Fried (SBF) was EA's ultimate success story — and then its ultimate cautionary tale. A MacAskill protégé who embraced "earning to give" (the EA principle that making as much money as possible in order to donate it is morally superior to direct service), SBF built the FTX cryptocurrency exchange and its associated trading firm Alameda Research into a multibillion-dollar empire. SBF was EA's largest individual funder. His FTX Future Fund distributed hundreds of millions toward longtermist causes. FTX and Alameda co-founders' wealth was estimated to constitute **35% of all funding available to EA**, making the movement dangerously dependent on a single source. SBF publicly identified as a utilitarian and longtermist. He told interviewers he would take a bet with a 51% chance of doubling the world and a 49% chance of destroying it — because the expected value was positive. This was longtermist logic taken to its unhinged extreme, and it was exactly the reasoning he applied to Alameda's finances. ### The Collapse In November 2022, FTX collapsed, revealing that SBF had been using customer deposits — roughly $8 billion — to cover losses at Alameda Research and fund his philanthropic and political activities. It was one of the largest financial frauds in history. The EA movement's response was chaotic. MacAskill tweeted that he was "shocked" — but reporting by *Time* magazine revealed that EA leaders had been warned about SBF's unethical behavior as early as 2018. Multiple people had raised concerns about SBF's honesty, his risk-taking, and Alameda's operations. The warnings were not acted on. SBF was convicted of fraud in November 2023 and sentenced to 25 years in federal prison in March 2024. ### What FTX Revealed The FTX collapse wasn't an accident — it was the logical endpoint of EA's moral framework pushed past its breaking point: **"Earning to give" as moral hazard:** If making money to donate is the highest moral calling, and if longtermist expected value calculations say the stakes are astronomical, then virtually any financial risk can be rationalized. SBF's fraud wasn't a betrayal of EA principles — it was EA principles applied without constraint. **Concentration risk:** A movement that allows 35% of its funding to come from a single source controlled by a single person has built a structural vulnerability indistinguishable from a scam. **Utilitarian blind spots:** Pure utilitarianism — maximizing expected value regardless of rights, consent, or process — will reliably produce monstrous outcomes when wielded by people who believe they're smarter than everyone else. This is the oldest critique of utilitarianism, and EA walked straight into it. **The guru problem:** EA's social structure concentrated enormous authority in a small number of philosopher-leaders (MacAskill, Ord, Karnofsky) and mega-donors (Moskovitz, SBF, [[Jaan Tallinn]]) with minimal accountability mechanisms. The structure rewarded deference to high-status figures — exactly the dynamic that enabled SBF to operate without scrutiny. --- ## The Post-FTX Reckoning (2023–Present) ### Institutional Restructuring The EA movement underwent significant organizational changes: **Effective Ventures (EV)** — the umbrella entity that housed CEA, 80,000 Hours, and Giving What We Can — agreed to a **$26.8 million settlement** with the FTX bankruptcy estate. Starting in late 2023, EV announced plans to decentralize, spinning off its subsidiary organizations into independent legal entities. The process was expected to take 1–2 years. **FTX Future Fund** — shut down immediately after the collapse. Its staff resigned en masse. **CEA and 80,000 Hours** — both transitioned to independent status, attempting to rebuild credibility outside the FTX shadow. ### Reputational Damage The movement's public image took a severe hit. EA went from being celebrated in mainstream media as a sophisticated approach to philanthropy to being associated with fraud, hubris, and intellectual arrogance. The UK Charity Commission investigated Effective Ventures. Recruitment to EA organizations declined. Several prominent EA figures publicly distanced themselves from the movement or left entirely. ### The [[AI Safety]] Pivot Continues Paradoxically, the FTX collapse did not slow EA's pivot toward AI safety — if anything, it accelerated it. With near-term EA funding decimated, the remaining infrastructure concentrated further on AI as a cause area. Open Philanthropy / Coefficient Giving continued ramping up AI safety grants. [[Anthropic]] (co-founded by former [[OpenAI]] researchers with EA-aligned safety concerns) became one of the most prominent AI companies in the world. The result: EA's current state is a movement that talks about "doing good" but functions primarily as an intellectual and financial ecosystem for AI development — the exact convergence that [[TESCREAL]] describes. --- ## EA's Institutional Infrastructure ### Funding - **Coefficient Giving** (formerly Open Philanthropy) — Dustin Moskovitz/[[Cari Tuna]]. ~$4B+ total giving. Dominant EA funder post-FTX. - **[[Survival and Flourishing Fund]]** — Jaan Tallinn. Longtermist and rationalist-aligned funding. - **FTX Future Fund** — Defunct. $26.8M clawed back via bankruptcy settlement. ### Movement Organizations - **[[Centre for Effective Altruism]] (CEA)** — Coordinates EA movement strategy, runs EA Global conferences. Now independent. - **80,000 Hours** — Career advice organization; the talent pipeline. Directs people toward longtermist-aligned careers. Now independent. - **Giving What We Can** — Pledge-based giving community. The surviving remnant of EA's original near-term focus. - **GiveWell** — Charity evaluator. Still funds evidence-backed near-term interventions, but increasingly overshadowed by AI-focused organizations. ### Research - **[[Global Priorities Institute]] (GPI)** — Oxford. Academic research on EA priorities. - **Rethink Priorities** — EA-aligned research organization covering animal welfare, AI governance, and global health. - **[[MIRI]]** — [[Machine Intelligence Research Institute]]. Rationalist AI safety research, Thiel-funded. ### AI Companies with EA DNA - **Anthropic** — Founded by Dario and [[Daniela Amodei]] (ex-OpenAI). EA safety philosophy is core to its mission and branding. - **OpenAI** — Originally EA-aligned nonprofit; pivoted to for-profit under [[Sam Altman]]. EA's biggest success story and biggest betrayal, depending on who you ask. --- ## The Critiques ### The Ends-Justify-Means Problem *Fortune* identified EA's "fundamental flaw" in a January 2024 analysis: the movement's utilitarian framework has no principled way to prevent the logic of "the ends justify the means." If the stakes are high enough (existential risk, trillions of future lives), virtually any present-day action can be rationalized. SBF's fraud is the extreme case, but the same logic operates at lower intensities throughout the movement. ### The Technocratic Capture Critique EA's evolution from "compare charities" to "AI safety is the most important cause" tracks perfectly with the material interests of its richest members. Dustin Moskovitz (Facebook), SBF (crypto), and the broader tech industry all benefit when EA directs resources toward AI research rather than, say, wealth redistribution or labor rights. The movement's analytical framework produces conclusions that align with its funders' interests — a pattern that's either a remarkable coincidence or a predictable outcome of who gets to set the parameters. ### The Torres-Gebru Critique [[Émile Torres]] and [[Timnit Gebru]] argue EA is not a standalone philosophy but one component of the [[TESCREAL]] bundle — and that its primary function within the bundle is to provide **moral legitimacy** for the concentration of power. EA's rhetoric of "doing good" and "saving the world" gives the tech oligarchy a philanthropic veneer that makes their self-interested power accumulation harder to criticize. See: [[TESCREAL]]. ### The Democratic Accountability Gap EA operates entirely outside democratic accountability. Its priorities are set by a small number of philosophers and mega-donors. Its institutions are private. Its decision-making is opaque. When EA decides that AI safety is more important than malaria prevention, there's no mechanism for the people affected by that decision — the children who won't receive bednets — to contest it. --- ## Connections to the Right-Wing KB **TESCREAL node:** EA is the moral legitimacy engine of the [[TESCREAL]] bundle. While other components provide the vision (cosmism), the urgency (singularitarianism), and the politics (longtermism), EA provides the framework that makes it all sound altruistic. **The Unholy Alliance:** EA alumni and institutions are embedded throughout the tech wing of [[The Unholy Alliance]]. EA-aligned people hold senior positions at Anthropic, OpenAI, and numerous AI safety organizations that interface with government policy. **Defense tech pipeline:** The EA-to-AI-safety pipeline feeds directly into the defense tech thesis: if AI is an existential risk, then AI supremacy is a national security imperative, then defense contracts for AI companies are morally justified. See: [[Transhumanist Defense]]. **Crypto-political ecosystem:** The FTX collapse exposed the EA-to-crypto pipeline — how EA ideology funded and legitimized the crypto bubble, and how crypto wealth was recycled into political influence. The FTX story connects directly to the broader [[Crypto-Political Machine]] documented in this vault. **The talent pipeline:** EA's most durable contribution to the tech oligarchy may be its recruitment infrastructure. 80,000 Hours and EA university groups funnel ambitious young people into AI companies, longtermist nonprofits, and eventually the political operations of the tech elite. --- ## See Also - [[TESCREAL]] — The bundle framework; EA is the moral legitimacy engine - [[Longtermism]] — The philosophical direction EA was captured by - [[Effective Accelerationism]] — The accelerationist fork that split from EA but shares conclusions - [[MIRI]] — The rationalist bridge between EA and AI safety - [[Transhumanist Defense]] — The defense tech endpoint of the EA-to-AI pipeline - [[Russian Cosmism]] — The deepest philosophical root of the long-term orientation - [[Dark Enlightenment]] — Shares EA's skepticism of democratic governance, through different reasoning - [[Techno-Optimist Manifesto]] — Andreessen's public declaration of the accelerationist position EA's safety wing ostensibly opposes