Sam Altman's Leadership Style and Conflicts of Interest

Key Insight

Despite public assurances in June 2023 that 'no one person should be trusted here' and that the board could fire him, Sam Altman reportedly called the shots behind the scenes, frustrating the supposedly independent board majority. His attempts to add an AI safety expert to the board after ChatGPT's release stalled due to his and Brockman's foot-dragging, leading him to propose candidates the board felt were beholden to him. This indicated an early 'power struggle' within the governance structure.

Altman's leadership was marked by perceived conflicts of interest and a lack of transparency. He claimed a board member's involvement with Quora's Poe chatbot was a conflict, arguing for their removal despite board consensus that Poe, as a wrapper, didn't meet their definition of a frontier model competitor. More significantly, the board discovered, months after initial inquiries, that Altman personally owned and managed the OpenAI Startup Fund, which used scarce company resources like early access to products, rather than the returns going to OpenAI investors. This raised concerns given his repeated claims of no stake in OpenAI, a status crucial for his board service.

Further concerns arose from Altman's extensive extracurricular activities, which he eventually claimed were on behalf of OpenAI. These included raising money for a venture to expand global microchip supply, taking his nuclear fission company Oklo public via a blank check company, Microsoft signing a deal to buy energy from his nuclear fusion company Helion (pending its invention), and meeting with Jony Ive about an AI device. The board often learned about these high-stakes activities from the press, fueling anxieties about potential use of OpenAI intellectual property for projects outside the national interest and a general lack of a 'whole story'.