NYT profile traces how rationalist circles shaped Dario Amodei's conviction that only people like him should build AI

dotey · x · 2026-11-08

The New York Times published a long profile of Anthropic CEO Dario Amodei on Nov 5, timed ahead of the company's IPO this month. The reporter interviewed over twenty classmates, friends, teachers, and colleagues, and reviewed early emails and internal OpenAI messages; Dario declined to be interviewed via a company spokesperson.

Core narrative: young Dario was steeped in the "rationalist" community (formed around Eliezer Yudkowsky's blogging, where many believe AI will destroy humanity) and effective altruism. Together they convinced him AI was so dangerous that only he and like-minded people should be trusted to build it — friends and colleagues say this drove him to race ahead of people who don't care about safety, and to seek control over the technology.

Other details: raised in San Francisco's Mission District by a leatherworker father and a librarian-restorer mother; attended Lowell HS while taking UC Berkeley classes, already pondering utilitarianism. Famous sushi-restaurant thought experiment: asked how many chickens someone would have to kill to justify killing that person, Dario smiled and said "there's always some number of chickens that outweighs a person." He started a Princeton physics PhD in 2006 (biophysics and computational neuroscience) the same year his father died of hepatitis C; his ex-girlfriend recalls he cited his father — who never saw the drugs that now cure most HCV — as proof that tool-builders beat single-disease researchers.

Original post →

NYT profile traces how rationalist circles shaped Dario Amodei's conviction that only people like him should build AI

More from AGI Musings