saved
Inside the Biggest Feud in Artificial Intelligence
Hraness wrote this summary from a saved copy of the source. Quotations are taken word for word from the source.
gist
Kevin Roose traces Anthropic’s split from OpenAI into a rivalry shaped by safety convictions, leadership distrust, and the commercial race toward artificial general intelligence. The article follows Dario and Daniela Amodei from internal OpenAI disputes through Anthropic’s growth from a small safety lab into a major competitor. Roose reports that the companies still cooperate on some safety measures while people close to their leaders expect the personal feud to continue.
ideas
- Dario Amodei’s scaling theory steered OpenAI toward large language models. His 2017 “Big Blob of Compute” essay argued that intelligence grew with scale and helped direct research toward artificial general intelligence.
- Amodei’s objections combined mission, safety, and trust. He opposed commercial choices, the countries plan for raising money against future AGI rights, and what he saw as Sam Altman’s misleading or reckless decisions.
- The 2020 departure turned an internal faction into Anthropic. Seven co-founders formed a safety-focused company after negotiations over veto power, employee departures, research documents, and a non-disparagement agreement failed to keep them at OpenAI.
- Anthropic outgrew its founding design. The company moved from a planned 50-person, safety-focused lab to thousands of employees, high-security facilities, major books and essays, and a race with OpenAI toward AGI.
- The rivalry remained personal after the companies became peers. Roose reports that leaders on both sides privately described the other as hypocritical, power-hungry, or evil even as the companies cooperated on some safety measures.
quotes
“The air is heavy with excitement and dread.”
“We’ll make the government pay for it”
“It’s stressful on every dimension”
“For now, it seems, the “machines of loving grace” will be built by humans of bitter spite.”