Anthropic CEO Dario Amodei Cites Lack of Trust for OpenAI Departure
Anthropic CEO Dario Amodei has elaborated on his 2020 departure from OpenAI, stating that a fundamental lack of shared vision and trust with CEO Sam Altman was the core reason. Amodei, who co-founded Anthropic with his sister and nine other OpenAI employees, expressed that he is "at peace" with the ongoing rivalry between the two major AI companies. He believes that the market and public opinion will ultimately determine the outcome of their differing approaches.
Dario Amodei, CEO of Anthropic, has publicly addressed his decision to leave OpenAI in 2020, attributing it to a lack of trust and differing visions with OpenAI CEO Sam Altman. Amodei stated, "Why argue with someone when you don't have the same vision and you don't trust them?" He indicated that he finds other figures in the AI sector, such as Google DeepMind CEO Demis Hassabis, to be trustworthy.
Amodei articulated that the resolution for such disagreements is for each party to pursue their own path. He expressed being "completely at peace" with Anthropic and OpenAI developing independently, suggesting that the ultimate judge of their approaches will be the market and public opinion. The exit of Amodei, his sister Daniela, and nine other OpenAI employees in 2020 to establish Anthropic has become a notable event in Silicon Valley.
The context of Amodei's departure has received renewed attention following a report by Ronan Farrow in The New Yorker, which examined questions regarding Altman's trustworthiness and cited Amodei's contemporaneous notes. Following the report, Altman's home was attacked, an incident he partly attributed to critical discourse within the industry, including how "certain other labs" and Anthropic discuss OpenAI.
Amodei also commented on a widely noted incident at an India AI summit, where he and Altman notably did not join other industry leaders in a show of unity. He attributed this to the "extreme disorganization" of the event, where Indian Prime Minister Narendra Modi was present. When discussing global cooperation on AI safety, Amodei suggested that "trustworthy actors" should unite to establish standards that compel others in the industry to conform. (Source: Business Insider)
