Effective Altruism

Effective Altruism is an intellectual movement with a simple core question: If you want to do good, how do you do as much of it as possible? Its answer: with evidence, numbers, and comparisons. Anyone donating to an aid organization should know how many people are actually made better off as a result. Anyone planning a career should weigh where, over the course of their life, they can achieve the greatest positive effect. The movement emerged in the early 2000s within the academic environment of Oxford and Princeton universities, shaped above all by the philosopher Peter Singer. Today it has offshoots in dozens of countries and is widely spread within the tech industry, especially in Silicon Valley.

Influence on AI safety research and tech philanthropy

Effective Altruism would have remained a fringe phenomenon without the AI boom. But many adherents of the movement are convinced that uncontrolled Artificial Intelligence could pose one of the greatest threats to humanity in the coming decades. This argument — known as “AI Safety” — has channeled billions of dollars into research institutes and nonprofit organizations.

Well-known organizations such as the Machine Intelligence Research Institute (MIRI) or the Center for Human-Compatible AI (CHAI) were substantially funded by donors close to EA. The nonprofit company Anthropic, which develops the chatbot Claude, was also founded by people close to the EA milieu. The scandal surrounding FTX founder Sam Bankman-Fried brought the movement into the headlines in 2022: he was a prominent EA representative and justified his actions with the movement’s principles.

How EA adherents make decisions

The core of the method is a comparison: Which problem is large, neglected, and solvable? Large means: many people or beings are affected. Neglected means: hardly anyone is working on it. Solvable means: additional resources would actually change something. Malaria prevention in Sub-Saharan Africa is considered a textbook example in EA thinking — a mosquito net costs only a few dollars and demonstrably saves lives.

Applied to AI risks, the argument runs: If a superintelligent AI were ever to become truly uncontrollable, the damage would be so immense that it is worth working on today — even if the probability is low. Critics counter that this logic is too speculative and neglects concrete problems such as poverty or climate change. Within EA, this is a serious debate, not a fringe opinion.

In practice, EA thinking manifests itself in certain career decisions. The organization 80,000 Hours — named after the average working time of a professional life — advises young people on which careers have the greatest social impact. Well-paid jobs in the finance industry are not vilified there, as long as the money earned is consistently donated. This is called “Earning to Give”.

Where the term appears in tech news

Anyone reading reports on AI regulation, start-up funding, or safety research regularly encounters EA. Many founders and investors in the AI field have an EA background or at least refer to its ideas. This explains why debates about “existential risks from AI” carry so much weight in the tech industry — even though they are more contested in mainstream academic research.

The term is also becoming politically relevant. In the US and the UK, EA-aligned lobby groups have influenced early attempts at AI regulation. Some journalists criticize that a small, wealthy movement is thereby disproportionately setting the agenda — pushing other risks, such as algorithmic discrimination or data misuse, into the background.

Effective Altruism is not a technical term in the narrower sense, but without it the culture of the AI industry can hardly be understood. Anyone wanting to know why certain research topics receive so much funding, who is behind certain organizations, or why some developers ponder the end of humanity will find in EA an important key.