What is Effective Altruism — and why is Silicon Valley obsessed with AI doom?

Effective Altruism traces its intellectual origins to Australian philosopher Peter Singer, who argued in 1972 that moral obligations to save lives should extend beyond geographical distance. This idea resonated with wealthy donors who sought a rational, evidence-based method to maximize the positive impact of their charitable contributions. The movement formalized in the late 2000s and early 2010s, with initiatives like GiveWell, co-founded by Holden Karnofsky, providing data-driven analyses of charity effectiveness, and "Giving What We Can," launched by Oxford philosophers William MacAskill and Toby Ord, encouraging significant income donations to combat global poverty.

The movement's emphasis on data and reason has attracted followers within American tech and finance sectors, where such decision-making frameworks are common. This approach aimed to shift the focus from traditional metrics, such as overheads, to an organization's actual impact. Richard Chappell, a University of Miami professor and EA proponent, identifies global poverty, factory-farmed animals, and "global catastrophic risk" as EA's three primary causes, with the latter directly linking to concerns about artificial intelligence.

EA's engagement with AI risks is further shaped by "long-termism," a philosophy championed by figures like MacAskill that prioritizes the distant future of humanity, and the rationalist movement, which has warned of existential threats from superintelligent AI since the 2000s. These overlapping ideas and memberships underscore a shared conviction that preventing human extinction should be a top priority, drawing figures like Sam Bankman-Fried, an early investor in Anthropic, who stated he entered finance to generate wealth for philanthropic giving.