Effective Altruism
Doing the most good with what we have.
Effective altruism asks how to use evidence and reason so limited time and money create as much human well-being as possible. The starting obligation is familiar: when serious suffering can be prevented at small personal cost, looking away is hard to defend. That idea supported charity evaluators, earn-to-give career paths, and large directed donations. The wider goal is to raise total well-being for friends, neighbors, and strangers, and to aim help toward a greater good for people as a whole when emotion alone would pick a narrower target.
What is effective altruism?
Peter Singer's shallow-pond example is the usual entry point. Someone who can easily pull a drowning child from water in front of them is expected to do it, even if shoes or clothes get ruined in the process. Singer's extension is that geography should not erase the duty. Preventable deaths from disease in another country still count as preventable deaths, so people with surplus resources are pressed to give, and to give where each dollar prevents the most harm or creates the most benefit that can be tracked.
Close relationships still matter in ordinary life. Family and friends will usually, and should, come before distant strangers when loyalty and care are on the line. Caring about humanity in general then means using what is left to do the most good available. That habit is separate from suicidal empathy, where endless acceptance of destructive behavior produces more damage. The useful discipline is noticing what you choose and what those choices are for.
Comparing impact
Effective altruism tries to compare good outcomes the way careful people compare investments. Programs get scored with tools such as cost per life saved or quality-adjusted life-years, so two charities can be set side by side. If one reliably helps many more people per dollar than another, sending money to the weaker option on purpose looks careless. The standard is maximizing impact under scarcity.
Careers get sorted with similar tools. Earn to give tells some high-skill people to take lucrative work, including Wall Street or quant trading, and donate heavily, on the theory that the donated dollars can outrun what they could do in a lower-paid helping job. Cause prioritization sorts problems by how large they are, how ignored they are, and how solvable they look, then sends effort toward the winner of that ranking. On that scale, a child who needs an organ transplant can outrank someone who will miss one meal, because the lasting damage is not comparable.
Feelings and inherited custom are treated as weak guides next to trials, metrics, and open calculation. The downside is obvious: justice, dignity, and solidarity are harder to put in a spreadsheet, so they can get ignored by people who only trust numbers. The opposite failure is real too, when raw feeling runs the show and slides into suicidal empathy. Good use of the method keeps measurement attached to actual human welfare and leaves room for judgment about what the numbers miss.
Hard utilitarian examples show how far some people push consequences. Imagine that stealing one loaf of bread somehow unlocked a path to fund major endometriosis research and treatment for many women. Stealing remains wrong, the owner is harmed, and trust takes a hit. In that invented chain, the later medical payoff still helps a large number of people living with severe pain. Not every effective altruist would accept that trade. The broader habit in the movement is still to ask which choice helps more people when money, time, and attention cannot cover everything.
Near-term aid and long-horizon bets
The movement grew out of the rationalist world around LessWrong, and that background helped shift attention from only present suffering toward huge future populations. Helping people alive today remains valuable. Once future people are counted equally, and their possible numbers are enormous, shaving down a small chance of human extinction can look larger, on expected value, than many malaria or poverty programs. Longtermism grew as the most influential expression of that shift inside effective altruism.
Into the 2020s, some of the movement's most visible career guidance put AI risk ahead of global poverty, while climate work often sat lower than people outside the community expected. Tools first popularized around mosquito nets and deworming were applied to bets on machine superintelligence. For a stretch, Sam Bankman-Fried was celebrated as a major donor who talked like an effective altruist while running a crypto empire that collapsed into one of the largest financial frauds on record.
Choice over control
The misunderstanding
Émile P. Torres, who spent time inside these circles, and Timnit Gebru's broader critique of related TESCREAL ideas, treat effective altruism as a pipeline into AI catastrophe culture, a story that can launder bad behavior, and a style of ethics that turns moral life into calculation tilted toward whoever holds the spreadsheet. Torres presses the point that hard utilitarianism can strip acts of built-in wrongness, so murder, fraud, lying, or cheating get judged only by whether the expected tally looks good afterward. Other attacks lean on looser language. Philanthropy gets called colonial whenever the people funding or running it fit certain demographics. Sam Bankman-Fried gets held up as if one crypto fraudster were the face of effective altruism.
The positive reality
Bankman-Fried committed a serious crime, and that fact should not be softened. He was a major donor who used effective-altruist language, not the definition of the movement, and treating him as its face flattens thousands of people working on health, poverty, and careful giving. Effective altruism still has a practical center: put scarce resources where they stop the most suffering and grow the most well-being. Family and friends can hold first claim in daily life while distant strangers still matter in the budget of surplus care.


